a. Photomicrograph of thin (lO^m) section of the Andura meteorite, an olivine bronzite-chondrite (H6) . Note the strong
contrast between the relatively coarse, equigranular, "metamorphic" texture of the matrix surrounding the chondrule, lower
left, and the fine-grained, unrecrystallized texture of the chondrule itself and other parts of the matrix, a contrast that indicates
a nonmetamorphic origin. Silicates, mostly olivine and pyroxene, show light pastel colors, metal (NiFe) brown, troilite (FeS)
dark brown. This section also illustrates the extreme difficulties in performing a modal analysis optically. Length of section
2.3 mm. b. Photomicrograph of thin (10/tm) section of the Pulsora olivine-bronzite chondrite (H3 to H7?) . Upper left
part consists of skeletal and zoned olivine (Fai3 to Faie) in silica-rich glass (H3) . The normal matrix, lower right, containing
numerous chondrules and fragments, is unsorted, mostly fine-grained and has the same (and constant) olivine-pyroxene
composition as the chondrules. Some lithic fragments, also with constant olivine composition, show an "achondritic" (H7?)
texture. Silicate minerals show varying pastel colors, the glass is pinkish, and the opaques (metal and troilite) are dark. Note
the eutectic metal/troilite spherule, top left. This section clearly demonstrates that the Pulsora chondrite never suffered
thermal metamorphism. Length of section 3.6 mm. (Fredriksson et al., "The Pulsora Anomaly.")
SMITHSONIAN CONTRIBUTIONS TO THE EARTH SCIENCES ? NUMBER 14
Mineral Sciences Investigations
1972-1973
George S. Switzer
EDITOR
SMITHSONIAN INSTITUTION PRESS
City of Washington
1975
ABSTRACT
Switzer, George S., editor. Mineral Sciences Investigations, 1972-1973. Smith-sonian Contributions to the Earth Sciences, number 14, 88 pages, 29 figures,
and frontispiece, 1975.?Thirteen short contributions from the Smithsonian'sDepartment of Mineral Sciences for 1972 and 1973 are gathered together in this
volume. Scientific contributions include new data on some mercury mineralsfrom Terlingua, Texas; a description of dashkesanite from St. Paul's Rocks; a
note on high-alumina basalt from the Aleutian Trench; descriptions of samplesfrom the Apollo 15 and 16 lunar missions; chondrule composition of the Allende
meteorite; the Pulsora meteorite and metamorphic equilibration in chondrites;the possible survival of very large meteorites that encounter the earth's surface;
data on eight observed-fall chondritic meteorites; chemical analyses of two micro-probe standards; and a technological note on the preparation of multiple micro-
probe samples. A history of mineral sciences in the Smithsonian Institution anda list of meteorites in the Smithsonian collections complete the volume.
OFFICIAL PUBLICATION DATE is handstamped in a limited number of initial copies and is re-corded in the Institution's annual report, Smithsonian Year. SI PRESS NUMBER 5138. SERIES
COVER DESIGN: Aerial view of Ulawun Volcano, New Britain.
Library of Congress Cataloging in Publication DataSwitzer, George S., comp.
Mineral sciences investigations, 1972-1973.(Smithsonian contributions to the earth sciences, no. 14)
Supt. of Docs, no.: SI 1.26: 14"Thirteen short contributions from the Smithsonian's Department of Mineral Sciences."
1. Meteorites. 2. Mineralogy. 3. Lunar petrology. I. National Museum of Natural History.Dept. of Mineral Sciences. II. Title. III. Series: Smithsonian Institution. Smithsonian
contributions to the earth sciences, no. 14.QE1.S227 no. 14 [QE395] 55O'.8s [549] 74-10590
For sale by the Superintendent of Documents, U.S. Government Printing Office
Washington, D.C. 20402 - Price $2.50 (paper cover)
Contents
HISTORY
Page
MINERAL SCIENCES IN THE SMITHSONIAN INSTITUTION, by Brian Mason ... 1
MINERALOGY
NEW DATA ON SOME MERCURY MINERALS FROM TERLINGUA, TEXAS, by
Roland C. Rouse 11
DASHKESANITE: HIGH-CHLORINE AMPHIBOLE FROM ST. PAUL'S ROCKS,
EQUATORIAL ATLANTIC, AND TRANSCAUCASIA, U.S.S.R., by Sara S. Ja-
cobson 17
PETROLOGY
NOTE ON HIGH-ALUMINA BASALT DREDGED NEAR THE ALEUTIAN TRENCH,
by William G. Melson 21
LUNAR STUDIES
COMPOSITION OF THREE GLASS PHASES IN AN APOLLO 15 BASALT FRAG-
MENT, by George S. Switzer 25
PETROGRAPHIC ANALYSIS OF APOLLO 16 SAMPLES 66083,1 AND 67943,1, by
Brian Mason 31
METEORITES
THE ALLENDE METEORITE: CHONDRULE COMPOSITION AND THE EARLY
HISTORY OF THE SOLAR SYSTEM, by Andrew L. Graham 35
THE PULSORA ANOMALY: A CASE AGAINST METAMORPHIC EQUILIBRA-
TION IN CHONDRITES, by Kurt Fredriksson, Ananda Dube, Eugene
Jarosewich, Joseph A. Nelen, and Albert F. Noonan 41
IMPACT SURVIVAL CONDITIONS FOR VERY LARGE METEORITES, WITH
SPECIAL REFERENCE TO THE LEGENDARY CHINGUETTI METEORITE,
by Robert F. Fudali and Dean R. Chapman 55
PRELIMINARY DATA ON EIGHT OBSERVED-FALL CHONDRITIC METEORITES,
by Roy S. Clarke, Jr., Eugene Jarosewich, and Albert F. Noonan 63
LIST OF METEORITES IN THE NATIONAL MUSEUM OF NATURAL HISTORY,
SMITHSONIAN INSTITUTION, compiled by Brian Mason 71
STANDARDS FOR CHEMICAL ANALYSIS
CHEMICAL ANALYSES OF TWO MICROPROBE STANDARDS, by Eugene
Jarosewich 85
POLISHED SECTION TECHNIQUES
PREPARATION OF MULTIPLE MICROPROBE SAMPLES, by Grover C. More-
land and Richard Johnson 87
iii
MINERAL SCIENCES INVESTIGATIONS
1972-1973
Dedicated to
EDWARD P. HENDERSON
Curator Emeritus, Department of Mineral Sciences
Smithsonian Institution
Mineral Sciences Investigations
1972-1973
Mineral Sciences in the Smithsonian Institution
Brian Mason
ABSTRACT
The Smithsonian Institution's interest in mineralsciences extends back to the founding of the Insti-
tution, whose founder, James Smithson, publisheda number of mineralogical papers. This account
traces the development of the mineral sciences (in-cluding all of geology, except paleontology and
stratigraphy) in the Smithsonian from its foundingin 1846 to 1963.
Introduction
This history has been compiled to record the
origin and development of mineral sciences as a
part of the Smithsonian Institution. Mineral sci-
ences are interpreted in a very broad sense as in-
cluding all of geology, except paleontology and
stratigraphy. This is essentially how the subject has
evolved in the Institution, and the present Depart-
ment of Mineral Sciences is responsible not only
for mineralogy and petrology (both terrestrial and
extraterrestrial?meteorites and lunar rocks), but
also for most aspects of physical geology, including
the exhibition hall "Our Restless Planet" and the
answering of many queries from the public on
general geology. The sources of this history include
published information from annual reports and
Brian Mason, Department of Mineral Sciences, National
Museum of Natural History, Smithsonian Institution, Wash-
ington, D.C. 20560.
research papers, excerpts from manuscripts in the
department's files (including a lengthy historical
account written by Dr. G. P. Merrill near the end
of his long curatorship), and verbal reminiscences
by staff members and others. The account has been
carried up to 1963, when the Department of Geol-
ogy was divided into two departments, Paleobiology
and Mineral Sciences. This seems to be a convenient
point in time to stop, since subsequent develop-
ments are well known both from published reports
and from the personal knowledge of the present
staff.
Although the Department of Mineral Sciences
was established as recently as 1963, the Smithso-
nian's interest in this field extends back to the
establishment of the Institution?or even before, if
we consider James Smithson's own contributions
to mineralogy. He published a number of minera-
logical papers, the most notable being one in the
Philosophical Transactions of the Royal Society
(1803), in which he showed that the material then
known as calamine comprised two distinct miner-
als, zinc carbonate (now known as smithsonite) and
zinc silicate (hemimorphite). Along with Smithson's
monetary bequest, the Institution received his
cabinet of minerals, described (Goode, 1897:305)
as "a cabinet, which . . . proves to consist of a choice
and beautiful collection of minerals, comprising
probably eight or ten thousand specimens. The
specimens, though generally small, are extremely
perfect, and constitute a very complete geological
and mineralogical series, embracing the finest vari-
ties of crystallization, rendered more valuable by
SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
accompanying figures and descriptions by Mr.
Smithson, and in his own writing. The cabinet also
contains a valuable suite of meteoric stones, which
appear to be suites of most of the important mete-
orites which have fallen in Europe during several
centuries." This collection was unfortunately lost
in the disastrous fire at the Institution in 1865.
The original charter of the Smithsonian Institu-
tion provided not only for a museum, but also for
a chemical laboratory. The chemical laboratory
was established by 1853, since in his report for 1854
the then Secretary, Joseph Henry (1854:18), re-
corded: "The laboratory of the Institution during
the past year has been used by Professor J. Lawrence
Smith in the examination of American minerals.
. . . He also made a series of analyses of meteorites,
among which were fourteen specimens belonging
to the cabinet of James Smithson." This work was
published as "Re-examination of American Miner-
als" (1853-1854) and "Memoir on Meteorites"
(1855). Unfortunately, the specific meteorites from
the Smithson collection are not identified and
presumably were lost in the 1865 fire. J. Lawrence
Smith moved to Louisville in the fall of 1854, after
giving a series of lectures during the winter of 1853-
1854 (for which he was paid $550, according to the
financial records). Apparently, however, he worked
in the Smithsonian chemical laboratory during
several subsequent summers.
After the brilliant beginning provided by Smith,
the mineral sciences languished for the next quarter
century. Joseph Henry was unwilling to commit the
Institution's limited income for the development
of a museum. Indeed, he appears to have discour-
aged the accumulation of large collections. In 1861
Professor Thomas Egleston of the School of Mines
at Columbia University, a prominent mineralogist,
was employed in the Smithsonian laboratory ex-
amining and classifying the collections representing
the acquisitions of the various land office surveys,
exploring expeditions, and other sources (Henry,
1862:38). It is further recorded (Henry, 1873:51)
that during the interval 1861-72 "a large number
of specimens in mineralogy and geology were also
packed up and transmitted to Professors Egleston
and Newberry of the School of Mines of New York
under the existing arrangements with those gentle-
men to select and label a perfect single series for the
U.S. National Museum and to exchange the dupli-
cate specimens in its interest." This may not have
been altogether in the interest of the Smithsonian
Institution; Merrill comments:
It is impossible, however, from the available records
to state the full number in the collections at any one time,
owing to the manner in which sets were made up to send
out under the administration of Messrs. Egleston and New-
berry. There is here shown a singular and inexplicable
discrepancy between manifested intentions and actual ac-
complishment since under the administration of Egleston and
Newberry these collections were largely, and in some cases
almost completely dissipated or depleted.
In this connection the following excerpt (Merrill,
c.l 925) of a letter from W. P. Blake to G. P. Merrill
dated 12 October 1899 is significant:
I was also astonished in looking through the museum
collection at Columbia College, New York to see the series
of very remarkable concretions I collected on the Colorado
desert in the cases. Wondering how they got there I inquired
of Prof. Baird and he told me that an arrangement had been
made with Newberry to look over the mass of material which
had been accumulated with permission to him to take such
duplicate material as he liked. [Merrill (c.1925) comments:]
"Evidently Prof. N. made the most of his opportunity!"
In 1873 F. M. Endlich was appointed chemist and
mineralogist in the Institution and served until
1879 when he resigned. During this period he was
engaged during the summers as a geologist on the
Hayden surveys in Colorado. Endlich reported in
1873 that the geological collections all told consti-
tuted 6300 specimens of which 3000 were minerals,
representing 230 species; 2300 rocks and 500 ore
specimens, with about 800 samples on hand await-
ing classification. In his final report Endlich (1880:
333-335) listed 343 species and varieties of minerals
in the collection.
The museum operations of the Smithsonian In-
stitution were reorganized and greatly expanded in
1880, following the great accretion of collections
from the Centennial Exhibition of 1876 and the
provision of a large new building to house them
(the present Arts and Industries Building). In a
handwritten manuscript Merrill commented:
It was a square squatty affair of red, blue and yellow
brick, exteriorly an architectural horror, interiorly a barren
waste. It presented but one redeeming feature?space; and
as it was space that Baird was after I presume it may at
first thought have been considered a success. One can here
but be reminded of the reply made by a high official after
being shown thru the then newly finished Pension Office
building. "Well," it was asked, "have you any criticism?"
"Yes," was the reply, "it is fireproof."
NUMBER 14
FIGURE 1.?George P. Merrill (1854-1929).
The expansion and reorganization of 1880 found
expression in the creation of several departments
and the appointment of salaried officials to take
charge of the various collections. Dr. George W.
Hawes was appointed curator of geology and min-
eralogy. In 1881 he was given three aids: F. P.
Dewey for chemistry and metallurgy, G. P. Merrill
for petrology (Figure 1), and W. S. Yeates for min-
eralogy. The organization thus formed was short-
lived, being terminated by the death of Dr. Hawes
on 22 June 1882. Upon the death of Dr. Hawes,
Mr. Dewey was appointed an assistant curator and
shortly afterwards full curator of a separate Depart-
ment of Metallurgy and Economic Geology. In
December 1883, Dr. F. W. Clarke of the U.S. Geo-
logical Survey was made honorary curator of the
Department of Mineralogy, with Mr. Yeates as
assistant. In the same year Dr. Merrill was made
assistant acting curator in charge of the newly
created Department of Lithology and Physical Ge-
ology. In 1889, after the resignation of Mr. Dewey
to become assayer at the Bureau of the Mint, the
two departments of Metallurgy and Economic
Geology, and Lithology and Physical Geology were
united under the title of Department of Geology,
Systematic and Applied, and Dr. Merrill appointed
curator. A complete reorganization in 1897 created
a single Department of Geology, with three divi-
sions: Physical and Chemical Geology, Mineralogy,
and Stratigraphic Paleontology. Merrill was ap-
pointed head curator of the department, and was
also curator of physical and chemical geology.
Clarke remained as honorary curator of mineralogy,
with Wirt Tassin assistant curator and Dr. L. T.
Chamberlain as honorary custodian of gems and
precious stones. The Department of Geology so con-
stituted survived with minor modifications for over
sixty years, until the creation of separate depart-
ments of Mineral Sciences and Paleobiology in
FIGURE 2.?William F. Foshag (1894-1956).
SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
1963. Throughout this time, however, the divisions
of Physical and Chemical Geology, and of Miner-
alogy actually functioned essentially as a single
unit, although they appear as separate units on the
table of organization until 1942. Merrill was effec-
tively curator of both divisions until his death in
1929. He was succeeded as head curator of the De-
partment of Geology by Dr. R. S. Bassler, a paleon-
tologist, and Dr. W. F. Foshag (Figure 2) was
appointed curator of both divisions. Dr. Bassler re-
tired in 1948, and Foshag succeeded him as head
curator, a position he held until his death in 1956.
Dr. G. A. Cooper, also a paleontologist, then be-
came head curator, and Dr. G. S. Switzer, curator
of mineralogy, an arrangement that continued until
the eventual formation of separate departments of
Paleobiology and Mineral Sciences.
Functions of the Department
With this introduction, it is convenient to de-
scribe the operations of the different segments of
the department under the headings of mineralogy,
petrology, meteorites, and gems and precious stones.
MINERALOGY
William S. Yeates, the first appointee specifically
in mineralogy, graduated M.A. in 1881 from Emory
and Henry College, Virginia, and immediately
thereafter became an assistant to Dr. Hawes. After
Hawes' death in 1882, Yeates remained as the
salaried officer in charge of the mineral collections
until May 1893, when he resigned to become state
geologist of Georgia. He was apparently more an
administrative officer than a scientific investigator,
and his work at the Smithsonian Institution con-
tributed little to the published record. Yeates was
succeeded for a short period by O. C. Farrington,
whose career is essentially linked with the Field
Museum of Natural History in Chicago, where he
was curator of geology from 1894 to 1930. In 1894
Wirt Tassin became assistant curator, with F. W.
Clarke continuing as honorary curator.
Tassin worked in the Division of Mineralogy
from 1894 to May 1909, when he resigned to go into
private business. He left an interesting and in-
formative manuscript account of the history of the
division up to 1906:
FIGURE 3.?Joseph E. Pouge (1887-1971).
In 1883 Mr. F. W. Clarke was appointed honorary curator.
This marks the real beginning of the present Division of
Mineralogy. To Mr. Clarke more than any other is due the
present standing of the collection. His appreciation of good
minerals, his knowledge of their value, his enthusiasm, his
widespread acquaintance and his position on the Geological
Survey and on exposition boards has been of incalculable
value to the collection. In fact it would be almost impossible
to say too much concerning the value of his influence in
securing fine specimens.
Tassin gave an account of the growth of the
mineral collections from 1859 to 1906. The total
number grew slowly from 793 in 1859 to 3551 in
1880, and then increased rapidly, evidently due to
the reorganization, reaching 45,291 in 1895; growth
then leveled off, the total in 1906 being 48,987.
While Tassin evidently devoted much of his work
to his curatorial and exhibition responsibilities, he
also did some research. His name is associated with
Dr. Merrill's in the investigation of several meteor-
ites, the first such research in the Institution since
that of J. Lawrence Smith half a century earlier.
Tassin was succeeded in 1909 by Dr. Joseph E.
Pogue (Figure 3), of whom Merrill (c.1925) re-
marked: "For the first time in the history of the
Department the collections were administered by a
NUMBER 14
FIGURE 4. Edgar T. Wherry (1885- ) . Photograph taken in his garden in Chevy Chase, D. C, about 1917.
,-?J.-,**
trained mineralogist." Pogue worked as assistant
curator of the Division of Mineralogy until October
1913, when he resigned to join the U.S. Geological
Survey. Much of his time must have been occupied
in moving the collections and exhibits into the pres-
ent Natural History Building, which was opened
in 1910. During his term in office Dr. Pogue visited
many European museums, studying collections and
displays. He also published several papers involv-
ing original research, among them a monograph
on turquoise published by the National Academy of
Sciences in 1915.
Pogue was followed by Dr. Edgar T. Wherry
(Figure 4), who carried on the work of the division
along the lines laid down by his predecessors, but
who was able to devote more time to research. He
published a number of papers, including one in
1917 giving the name "merrillite" to a new calcium
phosphate mineral that Dr. Merrill had recognized
as a meteoritic constituent in 1915. He was one of
the founders of the American Mineralogist and its
first managing editor (1916-1921), and president
of the Mineralogical Society of America in 1923.
He resigned from the Smithsonian Institution in
August 1917 to join the Bureau of Chemistry of the
Department of Agriculture.
When contacted in April 1972, Dr. Wherry was
kind enough to provide the following reminis-
cences:
In 1913 I was teaching mineralogy at Lehigh University,
Bethlehem, Pennsylvania, and spending summers as field
assistant to Professor Florence Bascom of Bryn Mawr College
in the preparation of geologic folios of quadrangles in
southeastern Pennsylvania, for the U. S. Geological Survey.
I was also collecting the minerals of the state, and writing
articles about them. Finding these activities somewhat over-
taxing I was glad to accept an invitation from Dr. George
SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
P. Merrill, head curator of Geology at the Smithsonian, to
join his staff as assistant curator of Mineralogy and Petrology.
Some years before this position had been held by a Mr.
Wirt Tassin, who had arranged the mineral collection as a
sort of textbook, accompanying each specimen with an elabo-
rate descriptive label, the study of which would take far
more time than most visitors could spare. When he retired,
his place was taken by Joseph Pogue, who rearranged the
collection in accordance with standard classifications, re-
placing the textual labels with simple ones bearing name,
composition, and locality. He resigned to go into business,
and I was taken on to complete his reorganization of the
collection.
My duties included answering correspondence about
minerals, including especially requests for identification of
specimens submitted, usually with reference to their potential
commercial value. Sometimes my reports would be ques-
tioned, but to make me feel at ease, the Smithsonian office
showed me a bit of pertinent correspondence: this comprised
a report on a specimen made by Dr. Merrill, a response from
the sender questioning its correctness, and the marginal
annotation addressed by Merrill to the Secretary of the
Smithsonian Institution, "If what this man says is true,
Merrill ought to be fired."
The most worth-while specimen I received was a bit of
clay, and instead of an inquiry as to its value, the question
was: "Is this of igneous or sedimentary origin?" Finding it
to be free from quartz grains, which make up the large bulk
of ordinary sediments, I guessed that it represented altered
volcanic ash. I Was promptly contacted by a patent-lawyer,
who explained that his client held a patent on the use of
igneous material for water-softening, and wished to claim
royalties from the organization marketing this clay for that
purpose, if it was indeed of igneous origin. I was accord-
ingly invited to go to South Dakota where the material
occurred, and study it fully. I was thereby enabled to
establish the nature of "bentonite" as representing altered
volcanic ash, which has had important geological applica-
tions.
In the course of that trip I stopped off at several notable
mineral localities, and collected multiple specimens of such
things as geodes, and lead-zinc ore-minerals. These were
packed in barrels, which could then be shipped under
government frank, and sent back for adding to the educa-
tional collections which the Smithsonian Institution was
distributing to schools.
In those days the whole field of mineralogy was handled
by one man, and I never had a technical assistant or under-
study. Records were kept by a pleasantly efficient secretary,
Miss Margaret W. Moody. She had scratched with a diamond
point a catalog number on the back of every gem-stone; and
when I resigned, it was her duty to check the gems against
the catalog, to see that I had not lost any of them. It was
pleasant to hear that only one ruby appeared to be missing,
but that turned out to have been transferred to a special
exhibit of North Carolina specimens.
Dr. Wherry was succeeded as assistant curator by
William F. Foshag, and at about the same time
Earl V. Shannon was appointed assistant curator
in the Division of Physical and Chemical Geology,
Merrill being curator of that division as well as
head curator of the Department of Geology. Shan-
non was an accomplished analyst and a keen miner-
alogist, and an extremely diligent worker. Between
1919 and 1929 he published over 80 papers, many
of them short notes but also some lengthy ones, in-
cluding a 483-page monograph on the minerals of
Idaho. Tragically he suffered a mental breakdown in
1929 and left the Institution. He was succeeded by
Edward P. Henderson, who served for over 35 years
until his retirement in 1966, and is still active in
the department as a research associate.
In 1926 the mineral collection was enormously
enriched by two remarkable bequests, which raised
it from comparatively undistinguished stature to
one of the world's great mineral collections. These
were the Roebling and Canfield bequests.
Colonel Washington A. Roebling (the Colonel
was a Civil War veteran) was a civil engineer and
the builder of the Brooklyn Bridge. During the
latter work he suffered "cassion disease," with per-
manent impairment of his health. This caused him
to turn his attention to mineral collecting. He
sought not only beautifully developed specimens
but also those that were rare?indeed his aim was
to secure a representative of every known mineral,
however insignificant and uninteresting in appear-
ance. In this he was eminently successful, the col-
lection on his death lacking only about a dozen of
the accepted mineral species at that time. His im-
paired health made travel and field collecting im-
possible for him, but he had ample financial means,
and corresponded vigorously with mineralogists
and mineral dealers throughout the world. He died
on 28 July 1926, leaving his collection of 16,000
carefully selected specimens to the Smithsonian
Institution. His son, John A. Roebling, generously
established an endowment fund of $150,000, the
income to be used for the acquisition of additional
material. This endowment has provided for the
continued growth of the Roebling collection.
The Canfield collection really comprises three
collections. The nucleus was formed by Mahlon
Dickerson of Dover, New Jersey, who began col-
lecting in the early part of the nineteenth century.
This collection passed to his nephew, Frederick
Canfield, and to it were added many fine specimens
collected at Franklin and Sterling Hill between
1840 and 1866. This collection, comprising some
NUMBER 14
1600 specimens, was displayed in the family man-
sion near Dover, where it remained until it was
transferred to Washington in 1926. Frederick A.
Canfield, son of Frederick Canfield, who inherited
the collection, preferred to leave it intact and begin
a new one of his own. This, at the time of his
death on 3 July 1926, totaled some 7500 specimens.
Frederick A. Canfield was educated as a geologist
and mining engineer at Rutgers and Columbia
Universities. In the practice of his profession he
had the opportunity to travel and acquire many fine
specimens at first hand. He also subsidized workmen
in the railway tunnels and quarries of northern
New Jersey, and in that way obtained many choice
specimens of the trap rock minerals. His profes-
sional connection with the Franklin and Sterling
Hill mines enabled him to keep pace with the new
and rare minerals constantly found there. He also
obtained many fine suites of specimens from out-
standing mineral localities. With the bequest of
the mineral collections Mr. Canfield also provided
an endowment of $50,000 for future accessions.
Dr. Merrill died on 15 August 1929 and was suc-
ceeded as head curator of the Department of Ge-
ology by Dr. R. S. Bassler. Dr. Foshag was appointed
curator of both the Division of Mineralogy and
Petrology, and the Division of Physical and Chem-
ical Geology, in effect combining the two divisions,
although on the table of organization the Division
of Physical and Chemical Geology survived until
1942. Dr. Foshag was born on Long Island in 1894,
but his family moved to California and he gradu-
ated from the University of California in 1919. In
1917 and 1918 he served as chemist for the Riverside
Portland Cement Company at the famous min-
eral locality of Crestmore, which probably effec-
tively started his career in mineralogy. During his
service with the Smithsonian Institution he ex-
amined many famous mineral localities in the
United States and Mexico. During World War II
he spent most of his time in the latter country in
cooperative work with the Mexican authorities in
the development of mineral deposits. During that
time, in 1944, the new volcano Paricutin erupted,
and he began a long-continued study of its eruptive
processes and products, a study terminated by his
death. His Mexican work also led him to an interest
in Latin American archeology and specifically a
study of jade objects from Central America. He was
successful in pinpointing a source for this jade in
Guatemala. His contributions to mineralogy were
recognized by the ward of the Roebling Medal of
the Mineralogical Society of America in 1953.
Dr. Foshag died on 21 May 1956 and was suc-
ceeded as curator of the Division of Mineralogy
and Petrology by Dr. George S. Switzer, who had
joined the museum as associate curator in the di-
vision in April 1948. In August 1957 Paul E. De-
sautels was appointed associate curator in the
division, and later in the same year Roy S. Clarke,
Jr., was appointed chemist. These men, together
with Dr. Henderson and Dr. Switzer, comprised
the professional staff when the division was re-
formed as the Department of Mineral Sciences in
1963.
PETROLOGY
Systematic petrology in the Smithsonian Institu-
tion can be fairly dated from the appointment of
G. W. Hawes as curator of the Department of Ge-
ology on 1 January 1881. Hawes was one of the
pioneers in the use of the petrographic microscope
in the United States, having learned its use during
graduate work at Yale University with Professor
E. S. Dana, and at the University of Heidelberg
with Professor Harry Rosenbusch; he received the
Ph.D. degree from the latter university in 1880. At
the same time as his appointment to the Smithso-
nian Institution, Hawes began an investigation of
the building-stones of the United States as part of
the Tenth Census; the Annual Report for 1881
(Baird, 1882:110) states:
Dr. Hawes . . . has gathered specimens of stone from every
quarry in the United States; and a force of fifteen men, in
part detailed from the Census Office, has been occupied all
the year in preparing them for study and exhibition. . . .
Numerous analyses of building-stones by the chemical and
specific gravity methods have been made by Messrs. F. P.
Dewey and F. W. Taylor. Since the 1st of June (1881), Mr.
Geo. P. Merrill has prepared 1550 microscopic slides of
building-stones, to be used in connection with the investiga-
tion.
Unfortunately Hawes developed tuberculosis during
1881, and the disease advanced rapidly. In 1882 he
went to Colorado in the hope of checking it, but
he died in Manitou Springs on 22 June 1882. On
his death the petrographic work on the building-
stones fell to G. P. Merrill, who noted in his manu-
script (c.1925) that he was wholly inexperienced in
SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
this field. He evidently learned quickly and well,
however, since he was one o? the co-authors of
"Reports on the Building-Stones of the United
States and Statistics of the Quarry Industry for
1880" (1884), in which he contributed an article
on the microscopic structure of the building-stones
and eighteen reproductions of photomicrographs.
Merrill continued his petrographic work on
building-stones and became an authority on the
subject, writing several monographs and books
thereon. He served as a special agent in this field
for the Twelfth Census (1912). He was consulted
frequently by both private and public officials re-
garding the stone to be used for various structures.
He thus served in an advisory capacity in the selec-
tion of stone for the Natural History Building, the
Old Post Office Building, Constitution Hall, the
columns on the west face of the Treasury Building,
and numerous others. Perhaps the most momentous
question left for his decision was the choice of stone
for the Lincoln Memorial (Yule marble from Colo-
rado). Probably through his work on building-
stones, Merrill became very interested in rock
weathering, and wrote extensively on this subject,
culminating in his book, A Treatise on Rocks,
Rock Weathering, and Soils. He also authored The
Nonmetallic Minerals: Their Occurrence and Uses.
After the turn of the century, Merrill's interest
turned to meteorites, and for the remainder of his
life his research work was concentrated in this field.
The petrological collections continued to grow,
largely through accessions of material from the
U. S. Geological Survey illustrating regional mono-
graphs and bulletins. In 1916, however, Joseph P.
Iddings was appointed honorary curator of petrol-
ogy, and after his death in 1920 Whitman Cross
succeeded to this honorary position and held it until
his death in 1949. These appointments brought two
of the authors of the CIPW system of igneous rock
classification into the Smithsonian Institution; a
third, Henry S. Washington, was at the Geophysical
Laboratory of the Carnegie Institution of Washing-
ton and many of his collections were eventually
given to the Smithsonian Institution. During his
period as honorary curator, Cross developed a col-
lection of some 2000 igneous rocks comprising most
of the types established by the CIPW system. Each
rock has a thin section accompanying it, with a
meticulous description on a file card; many are
chemically analyzed. This is probably the most com-
prehensive systematic collection of igneous rocks in
existence.
After the death of Cross active curatorial and re-
search work in petrology was largely discontinued,
because of inadequate staffing. This situation was
remedied by the establishment of a Division of
Petrology within the newly organized Department
of Mineral Sciences in 1963, and the appointment
of Dr. William G. Melson in 1964 to direct it.
METEORITES
As mentioned earlier, the original Smithson be-
quest included his cabinet of minerals, which con-
tained "a valuable suite of meteoric stones, which
appear to be suites of most of the important mete-
orites which have fallen in Europe during several
centuries" (Goode, 1897:305). Some of these were
analyzed by J. Lawrence Smith in the chemical
laboratory of the Institution in 1853-54. Unfor-
tunately, these meteorites were evidently lost along
with the rest of Smithson's cabinet in the fire of
1865.
The number of falls actually represented in the collection
at the first date mentioned (1880) cannot with absolute
certainty be given owing to the imperfection of the record,
but it was small; seven localities were represented: Imilac and
Vaca Muerta, Chile; Marietta (New Concord), Ohio; Par-
nallee, India; Searsmont, Maine; and Cold Bokkeveld, South
Africa. The Tucson iron which should have formed the chief
attraction was not catalogued, though brought, through the
influence of Dr. B. J. D. Irwin, to Washington in 1863. The
Casas Grandes iron which came to Washington from the
Philadelphia Centennial should also have received attention
(Merrill, c.1925).
The collection grew rapidly, evidently largely
through the work of Dr. F. W. Clarke, who was ap-
pointed honorary curator in the Department of
Mineralogy in December 1883. In 1889 he published
a catalog of the meteorite collection as of October
1888, in which he lists 128 distinct falls and finds;
in addition, there were 217 meteorites in the Shep-
ard collection, which was deposited in the museum
by Shepard's son in 1886 (and ultimately be-
queathed to the Smithsonian Institution in 1915).
There was considerable duplication between the
two collections and many of the specimens in both
collections were fragments. Nevertheless, the build-
ing-up of the meteorite collection by Clarke, almost
entirely by exchanges, was a notable contribution at
this early time.
NUMBER 14
In 1888 Merrill published his first paper on
meteorites, and was an active researcher in the field
until his death in 1929, publishing over eighty
papers on meteorites. The collection continued to
grow. In 1902, Tassin brought the Clarke catalog
up to date. For the purposes of this catalog the
Shepard meteorites were listed along with the mu-
seum's collection. The combined collections con-
tained 348 distinct falls and finds. In 1916 Merrill
published a "Handbook and Descriptive Catalogue
of Meteorite Collections in the United States Na-
tional Museum". The combined collections at that
time contained 412 independent falls and finds. The
meteorite collection has continued to grow steadily
and may now be the world's largest and most com-
prehensive from the standpoint of the number of
meteorites represented.
The tradition of meteorite research begun by
Merrill has continued. Dr. Foshag wrote a number
of papers on meteorites during his curatorship, and
his presidential address (1940) to the Mineralogical
Society of America was entitled "Problems in the
Study of Meteorites." Meteorite research within the
Institution, however, became the major interest of
Dr. Edward P. Henderson. Beginning in 1934 and
continuing throughout his curatorship and after
his retirement, he had published over thirty papers
on meteorites. Many of these were written in collab-
oration with Stuart H. Perry (1874-1957), a Michi-
gan newspaper publisher. Perry graduated in 1894
from the University of Michigan, where he studied
chemistry, geology, and zoology. Fossils were his
first love, but around 1930 his scientific interests be-
came directed towards the study of meteorites. He
was a vigorous collector, being aided in this respect
by adequate private means. In 1944, the Smith-
sonian published Perry's "Metallography of Mete-
oritic Iron," which for many years has been the
standard reference in this field. He also compiled
and privately published nine albums of photomi-
crographs of iron meteorites (166 in all), of which
six sets were produced. Through the years that
Perry was studying and collecting meteorites he
generously shared his knowledge and specimens
with others. Although he presented specimens to
many institutions, and privately assisted others in
their studies of meteorites, he was sincerely in-
terested in the growth of the national collection and
presented what he considered to be his most impor-
tant specimens to the Smithsonian Institution. Dur-
ing his life Perry donated 192 meteorites in this
way.
After J. Lawrence Smith's death in 1883, his
widow endowed the Lawrence Smith Medal, to be
awarded by the National Academy of Sciences for
"original investigation of meteoric bodies." This
medal was awarded to George P. Merrill in 1922,
Stuart H. Perry in 1945, and Edward P. Henderson
in 1970.
GEMS AND PRECIOUS STONES
This collection owes its beginning to an exhibit
made by F. W. Clarke and paid for ($2500) by
funds appropriated for the New Orleans Exposi-
tion of 1884. The same collection was displayed at
the Cincinnati Exposition the following year, after
which it was returned to Washington and placed
on exhibition in the museum. A detailed descrip-
tion of this collection was published by G. F. Kunz
in 1889. In 1891 the collection was greatly aug-
mented by purchases from the estate of Dr. Joseph
Leidy of Philadelphia, and was exhibited at the
Columbian Exposition in Chicago in 1893. In 1894,
Mrs. Frances Lea Chamberlain gave the museum a
collection of precious stones that had been made by
her father, Dr. Isaac Lea. Later, in 1897, her hus-
band, Dr. L. T. Chamberlain, was appointed hon-
orary curator of the collection and added a large
number of desirable specimens. On his death in
1913 he bequeathed a sum of money, the income
from which is used for its further increase.
The growth of the gem collection is documented
by the 1900 "Descriptive Catalogue" by Wirt Tassin
and in the 1922 "Handbook and Descriptive Cata-
logue" by G. P. Merrill. A recent popular account is
Gems in the Smithsonian by Paul E. Desautels.
Literature Cited
Baird, S. F.
1882. Report of the Secretary. Annual Report of the
Board of Regents of the Smithsonian Institution
for the Year 1881, 839 pages.
Clarke, F. W.
1889. The Meteorite Collection in the U.S. National
Museum: A Catalogue of Meteorites Represented
November 1, 1886. Annual Report of the Smith-
sonian Institution for 1886, 2:255-265.
Desautels, P. E.
1972. Gems in the Smithsonian. Washington, D.C.: Smith-
sonian Institution Press.
10 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
Endlich, F. M.
1874. Mineralogical Collection. Annual Report of the
Board of Regents of the Smithsonian Institution
for the Year 1873, pages 51-53.
1880. List of Species and Varieties of Minerals in the
National Museum of the United States in 1879.
Proceedings of the United States National Museum,
3:333-335.
Goode, G. B.
1897. The Smithsonian Institution, 1846-1896. Washing-
ton, D.C.
Henry, J.
1854. Report of the Secretary. Ninth Annual Report of
the Board of Regents of the Smithsonian Institu-
tion, 463 pages.
1862. Report of the Secretary. Annual Report of the
Board of Regents of the Smithsonian Institution
for the Year 1861, 463 pages.
1873. Report of the Secretary. Annual Report of the
Board of Regents of the Smithsonian Institution
for the Year 1872, 456 pages.
Kunz, G. F.
1889. Gem Collection of the U.S. National Museum.
Annual Report of the Smithsonian Institution for
1886, 2:267-275.
Merrill, George P.
1879. A Treatise on Rocks, Rock Weathering, and Soils.
1st edition, 411 pages. New York: MacMillian and
Co. [2nd edition, 400 pages, published in 1906.]
1884. Reports on the Building-Stones of the United States
and Statistics of the Quarry Industry for 1880. U.S.
Tenth Census, 10:1-410.
1888. On the San Emigdio Meteorite. Proceedings of the
United States National Museum, 11:161-167.
1904. The Nonmetallic Minerals: Their Occurrence and
Uses. 414 pages. New York: John Wiley and Sons.
1916. Handbook and Descriptive Catalogue of Meteorite
Collections in the United States National Museum.
United States National Museum Bulletin, 94:1-206.
1922. Handbook and Descriptive Catalogue of the Col-
lections of Gems and Precious Stones in the United
States National Museum. United States National
Museum Bulletin, 118:1-225.
[c. 1925.] [History of the Department of Geology.] Manu-
script, files of the Department of Mineral Sciences,
National Museum of Natural History, Smithsonian
Institution.
Perry, S. H.
1944. Metallography of Meteoritic Iron. United States
National Museum Bulletin, 184:1-206.
Pogue, J. E.
1915. The Turquoise: A Study of Its History, Mineralogy,
Geology, Ethnology, Archaeology, Mythology, Folk-
lore, and Technology. Memoirs of the National
Academy of Sciences, 12 (2, 3d memoir), 162 pages,
22 plates.
Shannon, E. V.
1926. The Minerals of Idaho, United States National Mu-
seum Bulletin, 131:1-483.
Smith, J. L.
1853.?1854. Re-examination of American Minerals. Ameri-
can Journal of Science, 15:207-215; 16:41-53, 365-
373; 18:372-381.
1855. Memoir on Meteorites. American Journal of Science,
19:153-163, 322-343.
Smithson, J.
1803. A Chemical Analysis of Some Calamines. Philo-
sophical Transactions of the Royal Society (London),
93:12-28.
Tassin, W.
1900. Descriptive Catalogue of the Collection of Gems in
the United States National Museum. Annual Re-
port of the Smithsonian Institution for 1900, pages
473-670.
Tassin, W.
1906. [The Mineral Collections, 1859-1906.] Manuscript,
files of the Department of Mineral Sciences, Na-
tional Museum of Natural History, Smithsonian
Institution.
Wherry, E. T.
1917. Merrillitc, Meteoritic Calcium Phosphate. American
Mineralogist, 2:119.
New Data on Some Mercury Minerals from Terlingua, Texas
Roland C. Rouse
ABSTRACT
An X-ray study has been performed on the threemercury minerals eglestonite, mosesite, and kleinite.
Previous work is reviewed and, where necessary, cor-rected in the light of new data. Eglestonite is cubic,
Ia3d, with a =16.04 A and has the formula Hg4C1
2O. Mosesite and kleinite are mercury nitrogenchloride sulfates. Mosesite is cubic with a = 28.62
A and kleinite, hexagonal with a = 40.60 and c =
11.16 A. Although the chemical formulas of theirsupercells are still uncertain, their subcell formulas
correspond essentially to the chloride salts of Mil-Ion's base (Hg
2NOH?H2O).
Introduction
The mercury deposit at Terlingua, Brewster
County, Texas, is known mineralogically for its
suite of secondary mercury minerals. It is, in fact,
the type locality for six of them including terlin-
guaite (Hg4O2Cl2), eglestonite (Hg4Cl2O), and the
mercury nitrogen chloride sulfates mosesite and
kleinite. These minerals are known from but a few
localities and one investigator (Lipscomb, 1957) in
a review of the crystal chemistry of mercury found
it quite remarkable that kleinite and mosesite
should occur at all in nature. The latter two are of
particular interest as they are thought to be stuffed
derivatives of the high tridymite and high cristo-
balite structures, respectively. They contain NHg4
tetrahedra, which play the same structural role as
SiO4 tetrahedra do in silicates.
There exists one definitive study of the whole
group, that of Hillebrand and Schaller (1909). Sub-
sequent investigators have compiled a large body
Roland C. Rouse, Department of Geology and Mineralogy,
The University of Michigan, Ann Arbor, Michigan 48104.
of chemical, optical, and crystallographic data, but
their results have often been ambiguous or con-
tradictory. An accurate determination of the crystal
structure has been made only for terlinguaite and
since this has resolved most of the problems associ-
ated with this mineral, it was not included in the
present study. The terlinguaite structure (Aurivil-
lius and Folkmarson, 1968) is interesting from a
crystal chemical viewpoint as it contains mercury
in two valence states. Some of the mercury atoms
are divalent and form the characteristic linear
O-Hg-O groups (Grdenic, 1965). Others are ar-
ranged in equilateral triangular clusters in which
each mercury has a formal oxidation number of
+ 4/3. The Hg-Hg distances in the cluster are
shorter than those in solid mercury indicating in-
termetallic bonding, and terlinguaite is therefore a
metal cluster compound.
Experimental Methods
The study of eglestonite, mosesite, and kleinite
is rendered difficult by their tendency to occur as
multiple crystals, whether as twins, parallel growths
or random intergrowths. In the case of kleinite this
seems to be invariably the case. The specimens
used in this study were from Terlingua and Hua-
huaxtla, Mexico, and were obtained from the U. S.
National Museum collections. As reported by pre-
vious workers, eglestonite is photosensitive, its color
changing rapidly from yellow to brown to black.
The color change produced no obvious changes in
the x-ray diffraction patterns, but this matter was
not investigated systematically.
Unit cell parameters were determined from pre-
cession photographs and verified by the cone-axis
or rotating crystal methods. Space groups were
determined from precession and Weissenberg pho-
tographs taken with MoKa and CuKa radiations,
11
12 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
respectively. The cell parameters were refined from
powder data by the method of least-squares using
the program of Appleman and Evans. In the case
of mosesite and kleinite the data were obtained
from a powder diffractometer utilizing monochro-
matized CuK? radiation and quartz as an internal
standard. Since there was not enough eglestonite
available for this procedure, it was necessary to use
16 reflections between 102? and 165? 2e measured
from a Debye-Scherrer photograph. The powder
patterns in Tables 1, 2, and 3 were likewise mea-
sured from shrinkage-corrected Debye-Scherrer pho-
tographs taken with a 114.6 mm diameter camera
and filtered CuK? radiation. Intensities were esti-
mated visually.
Minerals Studied
EGLESTONITE
Eglestonite was first described by Moses (1903)
as a cubic mineral having the formula Hg6Cl3O2.
Hillebrand and Schaller (1909:145, 147) confirmed
the cubic symmetry but found the formula to be
Hg4Cl2O on the basis of three chemical analyses.
The crystal class was determined to be m3m from
the presence of the hexoctahedron as a common
form.
The crystal structure of eglestonite is of some
interest due to the presence of the binary ion
Hg2+2. Two structures have been proposed, both
by the same investigator. Hedlik (1948) found
eglestonite to be cubic, Pm3n, with a =16.07 A.
She went on to propose a structure containing the
mercurous ion and having Hillebrand and Schal-
ler's chemical formula. However, in Hedlik (1950)
there appeared a revised structure based on a cell
having a = 8.03 A and symmetry Im3m. The for-
mula was also changed to Hg6Cl4O, which is not
in accord with the existing chemical analyses. No
explanation was offered for the revisions and since
both structures were derived with powder and
rotating crystal data, neither can be considered re-
liable. To compound the confusion, Wolfe (cited
in Palache, et al., 1951:51) reported the cell pa-
rameter a =16.03 A and a probable space group
Ia3d for eglestonite.
The present study has confirmed Wolfe's results.
Eglestonite is cubic, Ia3d, with a refined parameter
a = 16.0398 ? 0.0003 A. There is also a very promi-
TABLE 1
/
10
40
100
<1
7
7
10
50
5
10
5
<1
<1
1
1
80
1
<1
3
20
1
10
<1
1
1
5
5
<1
5
1
5
10
<1
<1
5
3
<1
5
Plus ca. 65 more
.?Powder
dohs
6.549
4.006
3.271
3.142
2.925
2.832
2.599
2.529
2.471
2.313
2.265
2.179
2.135
2.034
2.003
1.887
1.858
1.811
1.788
1.707
1.687
1.633
1.617
1.586
1.554
1.526
1.499
1.475
1.461
1.384
1.373
1.334
1.308
1.290
1.266
1.237
1.215
1.182
lines to 0.777
pattern of
dcalc
6.548
4.010
3.274
3.146
2.928
2.835
2.602
2.536
2.475
2.315
2.268
2.183
2.143
2.037
2.005
1.890
1.865
1.816
1.793
1.710
1.691
1.637
1.620
1.588
1.558
1.529
1.502
1.477
1.464
1.386
1.375
1.337
1.310
1.293
1.268
1.237
1.216
1.182
A
eglestonite
hkl
211
400
422
510, 431
521
440
611, 532
620
541
444
543, 710,
550
721, 633,
552
642
651, 732
800
660, 822
831, 750,
743
752
840
664
851, 930,
754
844
941, 853,
770
10 ? 1 ? 1, 772
943
10.3? 1, 952
871
961, 10.3*3
10*4.2
11 .3?2, 10.5.3,
972
10.6.0, 866
12.C0, 884
10.7.1, 11 .5.2
12?3. 1, 983
12.4.0
10.8*2
13.2.1, H'7.2
12.6-2
nent subcell having ^4= a/2 = 8.02 A and symmetry
Im3m. That the 16.04 A cell is the correct one is
confirmed by powder data (Table 1), which can only
be indexed fully on the larger cell. The unit cell
contents calculated from the average of Hille-
brande's analyses are Hg9411 C149J6 O23.76. From
a consideration of the available equipoints in
NUMBER 14 13
Ia3d, the correct formula must be Hg4Cl2O with
Z = 24. These results indicate that Hedlik's revised
structure is necessarily wrong since it is based on
the subcell rather than the true cell and gives a
formula inconsistent with the chemical analyses.
Preliminary work on a redetermination of the
eglestonite structure is underway and a complete
study is planned.
MOSESITE
Mosesite was described by Canfield, et al. (1910)
as a new mercury-ammonium compound containing
chloride, sulfate, and water. Although cubic in
morphology, it was weakly birefringent but became
isotropic when heated to ^186?C. From this they
concluded that mosesite exists in high and low
temperature polymorphic forms with the inversion
temperature being 186?C. This, of course, is not
necessarily the case as many cubic materials show a
weak, anomalous birefringence. In a study of moses-
ite from the Fitting district of Nevada, Bird (1932)
found on the basis of powder data that mosesite
is cubic with a face-centered lattice and a = 9.57 A.
A restudy of mosesite was performed by Switzer,
et al. (1953), who obtained the first complete chem-
ical analysis of the mineral. Noting the strong
similarity of its powder pattern and chemistry to
those of Millon's base (Hg2NOH-2H2O), they de-
rived a formula Hg2N(Cl,SO4,MoO4,CO3)?H2O
(Z=8) for material from Huahuaxtla Mexico. Mo-
sesite is, therefore, the chloride salt of Millon's
base with lesser amounts of other large anions. The
structure of Hg2NOH?H2O was determined by
Lipscomb (1951) to be cubic, F43m, with ? = 9.58 A.
It is a stuffed derivative of the high cristobalite
structure with a 3-dimensional framework of linked
NHg4 tetrahedra. The OH" and H2O occupy the
large open channels in the framework. This ex-
plains the presence of large complex anions in the
mosesite formula. Switzer, et al. (1953) also pro-
posed a structure for mosesite analogous to that of
Millon's base. The structure is cubic, F43m, with
a = 9.524 A, which is consistent with the powder
pattern.
The author has reexamined the mosesite crystals
from Huahuaxtla, Mexico, analyzed by Fahey (in
Switzer, et al., 1953). Although the X-ray study of
this material is not yet complete, several important
results can be reported. Weissenberg photographs
TABLE 2.?Powder pattern of mosesite
1
75
10
90
100
40
40
25
35
25
20
30
20
20
20
10
10
15
20
15
15
Plus ca.
dobs
5.512
3.375
2.877
2.756
2.382
2.186
1.834
1.685
1.610
1.454
1.436
1.375
1.334
1.241
1.192
1.165
1.102
1.093
1.066
1.047
16 more lines
dcalc
5.508
3.373
2.876
2.754
2.385
2.188
1.836
1.832
1.686
1.683
1.612
1.607
1.455
1.453
(mostly broad bands)
hkl
333, 511
660, 822
933, 771,
755
666, 10.2.2
12-0.0
993, 13 ? 1 ? 1
999, 15.3.3
12.10.0
12-12 ? 0
17*0.0
17.5.1
14.1L0
19.5.1
18.8.0
to 0.787 A
confirm the cubic symmetry, but the space group
has not yet been determined with certainty. A
rotating crystal photograph shows that mosesite
has a supercell with a = 28.8 A, three times the value
previously reported. The refined value is a = 28.618
? 0.002 A. The subcell has A =a/3 = 9.54 A. There
is also a second subcell, defined by the strong re-
flections, which has ,4' = a/12 = 2.38A.
Powder data for mosesite are presented in Table
2. The pattern is similar to that of Switzer, et al.
(1953), except that the d-values and. intensities of
the reflections are in better agreement with those
of Hg2NOH>2H2O than are the data of Switzer,
et al. This is especially true of the line at 5.512 A,
whose intensity is much greater than reported by
those authors. The indexed reflections in Table 2
do obey the presence rule for a face-centered 9.54
A cell. Powder diffractometer scans, however, show
a small peak at 12.87 A, which can only be indexed
as 210 (dcalc= 12.80 A) of the 28.62 A supercell.
This is in accord with the single crystal results and
also seems to require a primitive lattice for mosesite.
It is possible that the 9.54 A subcell is face-centered,
14 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
while the supercell is primitive. The substructure
corresponds to the structure of Millon's base and
the superstructure is perhaps connected with order-
ing of the several kinds of large interstitial anions
in mosesite. The symmetry and lattice type of the
supercell is the subject of continuing single crystal
studies.
KLEINITE
A fourth yellow mercury mineral occurring at
Terlingua was mentioned by Moses (1903), but it
remained for Sachs (1905) to name it kleinite. Hille-
brand and Schaller (1909) found it to be another
mercury ammonium compound with chloride and
sulfate, but despite many chemical analyses they
could not derive a satisfactory formula. They also
described in detail the highly anomalous optical
properties of kleinite. Although the mineral is mor-
phologically hexagonal, basal sections are strongly
birefringent. Above about 130?C, however, kleinite
becomes uniaxial and, as in the case of mosesite,
this was ascribed to polymorphism. They also noted
that birefringent basal sections would not extin-
guish under crossed polars. This was explained by
assuming each crystal to be composed of a group of
parallel biaxial plates stacked along c. During fur-
ther heating experiments, Canfield, et al. (1910)
made the interesting observation that kleinite be-
comes isotropic above 186?C.
In an x-ray investigation Heritsch (1949) found
that all kleinite crystals are composed of several
individuals and that heating above the supposed
inversion temperature does not change this. His
crystals were all hexagonal, probably P63/mmc,
with a= 13.56 and c? 11.13 A. There is also a sub-
periodicity of a/2 = 6.78 A. A crystal structure was
deduced using rotating crystal data. In a later paper
Heritsch (1954) noted that mosesite had been
shown to be isostructural with Millon's base, which
is in turn a high cristobalite derivative. By analogy,
he proposed that kleinite is a high tridymite de-
rivative with the formula (Hg2N) (Cl,SO4)?xH2O
where x^\/4. As with mosesite the structural chan-
nels contain the large anions and water molecules.
Nijssen and Lipscomb (1954) came to the same
conclusion on the basis of the close relationship
between kleinite and Hg2NBr. They determined
the structure of the latter to be a derivative of high
tridymite. Hg2NBr is hexagonal, P63/mmc, with
0 = 6.65 and c= 11.26 A, which corresponds to the
kleinite subcell. In a later paper, however, Lip-
scomb (1957) reported evidence that the a parameter
of kleinite given by Heritsch should be tripled. He
further suggested on the basis of some intensity
calculations that the channels in the kleinite struc-
ture contain mercury atoms as well as anions.
In the present study, the interpretation of the
diffraction photographs was greatly hindered by the
fact that every crystal examined was multiple. In
some cases, however, the volume of one individual
was considerably greater than those of the others.
It was therefore possible to separate its reciprocal
lattice pattern from those of the others and study
its symmetry and cell parameters. This approach
is not infallible, but its results are supported by the
powder data, which are not subject to the uncer-
tainties introduced by the multiplicity of the single
crystals.
Rotation photographs around the c-axis are con-
sistent with Hillebrand and Schaller's model of a
stacking of parallel plates along c. Adjacent indi-
viduals, however, seem to be slightly tilted with
respect to one another by ^?1.5? in those crystals
examined.
As suggested by Lipscomb (1957), kleinite is
hexagonal with a three times the value given by
Heritsch. The refined cell parameters are a = 40.599
?0.005 and c=11.155?0.003 A. There are two
subcells. One, denned by the total pattern of me-
dium plus strong reflections, has A= a/6 = 6.77 and
C = c= 11.16 A. The other, denned by the strong
reflections alone, has /4' = a/12 = 3.39 and C' = c/2 =
5.58 A. An independent check on these parameters
is provided by the powder pattern (Table 3). It
contains four low-angle reflections (220, 222, 811,
and 702), which cannot be indexed on Heritsch's
cell but which can be indexed quite satisfactorily
on the 40.60 A cell.
The space group of the supercell could not be
determined since the superstructure reflections on
the Weissenberg photographs are too few and
scattered. However, the 6.77 A subcell does have
symmetry P63mc, P62c, or P63/mmc. This subcell
is the analog of the Hg2NBr cell.
Using the average chemical analysis of Hille-
brand and Schaller (1909), the supercell contents
are
Hg330.47^141.41*->1158.95C)24.96 W143.91H,
NUMBER 14
/
10
15
15
25
5
20
<1
1
10
5
100
70b
15
90
3
TABLE 3.?-
dth,
10.19
5.871
5.577
5.188
4.890
4.040
3.850
3.731
3.382
3.132
2.931
2.888
2.831
2.788
2.688
100
5
5V
10
10
1
10
20v
3
15
3
3
30
dcalc hkl
2.592
2.518
2.446
2.306
2.215
2.171
2.084
2.057
2.019
1.900
1.843
1.772
1.736
1.690
10.15
5.860
5.577
5.188
4.888
4.040
3.861
3.836
3.732
3.383
3.140
3.127
2.930
2.893
2.890
2.883
2.834
2.789
2.693
2,689
2.594
2.592
2.518
2.513
220
600
002
601
222
602
811
820
702
660
603
433
12?0?0
662
860
752
12 ? 0 ? 1
004
903
224, 960
12*0.2
504
604
10 .4*2
Plus ca. 40 more lines (mostly broad bands) to 0.781 A
b broad, vb very broad
On the basis of the available data it is difficult to
derive a satisfactory formula from this. A formula
for the 6.77 A subcell, however, may be derived by
analogy to that of Hg2NBr. The subcell contents
are
93C14 42 O4
or very nearly Hg9N4(Cl,SO4)5'H2O. This can be
rewritten as 4H^NCl?Hg(Cl,SO4)-H2O. Presum-
ably, the 4Hg2NCl represents that part of the sub-
structure analogous to Hg2NBr (for which Z = 4),
while the remainder represents the contents of the
structural channels.
15
The true symmetry and formula of the super-
structure can only be settled by a complete struc-
ture determination combined with a new chemical
analysis. The former objective, however, may not
be attainable due both to the lack of single crystals
and the experimental difficulties inherent in treat-
ing highly absorbing compounds with very large
supercells. The outlook for mosesite is more hope-
ful, however, and further work is anticipated on
this mineral.
Literature Cited
Aurivillius, K., and L. Folkmarson
1968. The Crystal Structure of Terlinguaite Hg4O2Cl;,.
Ada Chemica Scandinavica, 22:2529-2540.
Bird, P. H.
1932. A New Occurrence and X-Ray Study of Mosesite.
American Mineralogist, 17:541-550.
Canfield, F. A., W. F. Hillebrand, and W. T. Schaller
1910. Mosesite, a New Mercury Mineral from Terlingua,
Texas. American Journal of Science, 4th series,
30:202-208.
Grdenic, D.
1965. The Structural Chemistry of Mercury. Quarterly
Reviews (London), 19:303-328.
Hedlik, A.
1948. Uber die Formel und Struktur von Eglestonit.
Experientia, 4:66.
1950. Uber Formel und Struktur des Mercurooxychlorides
Eglestonit. Tschermaks Mineralogische und Petro-
graphische Mitteilungen, Dritte Folge, 1:378-389.
Heritsch, H.
1949. Rontgenuntersuchungen an Kleinit. Tschermaks
Mineralogische und Petrographische Mitteilungen,
Dritte Folge, 1:300-312.
1954. Bemerkungen zur Kristallchemischen Konstitution
des Kleinites. Anzeiger der Osterreichischen Aka-
demie der Wissenschaften, Mathematisch-Wissen-
schaftliche Klasse, 1:1-4.
Hillebrand, W. F., and W. T. Schaller
1909. The Mercury Minerals from Terlingua, Texas.
United States Geological Survey Bulletin, 405.
Lipscomb, W. N.
1951. The Structure of Millon's Base and Its Salts. Acta
Crystallographica, 4:156-158.
1957. Recent Studies in the Structural Inorganic Chem-
istry of Mercury. Annals of the Neiv York Academy
of Sciences, 65:427-435.
Moses, A. J.
1903. Eglestonite, Terlinguaite and Montroydite, New
Mercury Minerals from Terlingua, Texas. Ameri-
can Journal of Science, 4th series, 16:253-263.
16 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
Nijssen, L., and W. N. Lipscomb 1905. Der Kleinit, ein Hexagonales Quecksilberoxydchlorid
1954. A Hexagonal Modification of a Salt of Millon's von Terlingua in Texas. Sitzungberichte der Kon-
Base. Ada Crystallographica, 7:103-106. iglichen Preussischen Akademie der Wissenschaften
Palache, C, H. Berman, and C. Frondel (Berlin), 1091-1094.
1951. Dana's System of Mineralogy. Volume 2, 7th edition. Switzer, G., W. F. Foshag, K. J. Murata, and J. J. Fahey
New York: John Wiley and Sons. 1953. Re-examination of Mosesite. American Mineralogist,
Sachs, A. 38:1225-1234.
Dashkesanite: High-Chlorine Amphibole
from St. Paul's Rocks, Equatorial
Atlantic, and Transcaucasia, U.S.S.R.
Sara S. Jacobson
ABSTRACT
New data are given for dashkesanite, a high-chlorineamphibole, from two localities, Transcaucasia,
U.S.S.R., and St. Paul's Rocks, Equatorial Atlantic.Four new analyses (by electron microprobe) of the
mineral from Transcaucasia gave a chlorine con-
tent ranging from 4.15 to 5.34 percent. The originalanalysis by Krutov (1936) by wet chemical methods
reported Cl = 7.24 percent. Five analyses (by micro-probe) of dashkesanite from St. Paul's Rocks gave
Cl ranging from 6.03 to 6.51 percent. The mineral
from both localities is low in SiO2 and MgO andhigh in FeO, Cl, and K
2O compared to other horn-blende compositions. It is hastingsitic using the
classification of Ernst (1968).
Introduction
High-chlorine amphibole was first described by
Krutov (1936). He found the mineral in skarn rocks
surrounding sheetlike magnetite ore bodies at Dash-
kesan, Transcaucasia, U.S.S.R. These skarn rocks
occur about 1200 meters from the ore body and
extend for over 200 meters with an average thick-
ness of 0.60-0.75 meters. The surrounding country
rock is described as porphyrites and tuffs with dikes
of gabbro porphyrite occuring near the skarn. Kru-
tov (1936) named the mineral "dashkesanite" for
the locality.
The mineral was also found by Melson, et al.
(1972) in brown hornblende mylonites from St.
Sara S. Jacobson, Department of Geology, University of
California, Los Angeles, Los Angeles, California 90024.
Paul's Rocks, an ultrabasic intrusion exposed in a
series of islets in the equatorial Atlantic. Brown
hornblende mylonites are the second most abun-
dant of three rock types found on the islets; peri-
dotite mylonites are the most abundant rock type
and clinopyroxene mylonites are the least abun-
dant. All samples examined were collected in 1966
during Cruise 20 of the R. V. Atlantis II. Samples
from both of these areas were investigated in the
present study.
ACKNOWLEDGMENTS.?I wish to thank W. G. Mel-
son for suggesting the project and his help through-
out.
Petrography
The specimen from Transcaucasia available for
study (USNM 120180) consists of fine- to medium-
grained, subhedral to anhedral crystals of dashke-
sanite; minor chlorite and epidote and a fine-
grained material that may be scapolite are also
found in the specimen. This specimen of dashke-
sanite is dark green. In thin section, it shows a
strong yellow to blue-green pleochroism. Cleavage
is not well developed; rather a basal parting appears
to dominate. The mineral has inclined extinction
and positive elongation.
Dashkesanite from St. Paul's Rocks was studied
in two sections of brown hornblende mylonite,
(USNM 110393-13 and 110393-31). Both are from
the Southeast Islet. (See Melson, et al., 1972, for
map of localities.) The primary assemblage is
kaersutitic hornblende, plagioclase, and Fe-Ti ox-
ides. These occur with accessory amounts of dash-
kesanite, allanite, zircon, scapolite and titanbiotite.
17
18 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
FIGURE 1.?Portion of a hornblende augen in section USNM
110393-13. The diagonal grayish area in the central part
of the photo is dashkesanite. Width of field is 1.6 mm x 1.3
The hornblende forms large rounded augen. Dash-
kesanite occurs in veins in the augen and as over-
growths on the hornblende augen. It is also found
in veins in the fine-grained matrix. It does not oc-
cur disseminated in the matrix (Figure 1).
The optical properties of the St. Paul's Rocks
dashkesanite are similar to those described for the
mineral from the type locality. The dashkesanite
forms fine to coarse, anhedral to subhedral masses;
good crystal forms are rare. The mineral again
shows the distinctive, strong yellow-green to blue-
green pleochroism. The dashkesanite is not sheared
nor mylonitized.
Analytical Procedure
Analyses of the above samples were done on an
ARL EMX electron microprobe using a 15kV ac-
celeration potential, 0.15 juA sample current and a
TABLE 1.?Amphibole analyses
Constituent
SiO2
TiO2
A12O3
Fe2O3
FeO
MgO
CaO
Na2O
K.,O
Cl
H2O
O = C1
Total
Si
Al
Al
Ti
Fe**
Mg
Fe+2
Na
Ca
K
OH
Cl
1
36.13
0.30
10.22
7.60
19.99
3.63
10.83
1.24
2.84
7.24
1.05
101.93
1.63
100.30
5.934 |
1.978 \
0.035 /
o.94o y
0.888 I
2.747 *
0.394 1
1.906 >
0.596 \
1.154 I
2.016 S
2
33.90
1.01
11.57
-
31.91*
1.81
11.16
0.74
3.34
4.95
-
100.44
1.12
99.32
7-91 52fx | 8.00
0.26 \
0.13 1
4.67h - \ 5.50
0.47 i
4.64 )
0.25 )
2.90 2.08 I 3.07
0.74 \
317 IM
3
35.31
0.75
11.81
-
30.21 ?
2.27
11.34
0.89
2.92
4.15
-
99.75
0.94
98.81
6.06 /
1.94 \
0.40 \
0.10 1
0.58 i
4.34 *
0.33 1
2.09 >C
0.64 )
1.21
4
33.24
1.14
12.21
-
31.761
1.70
11.13
0.75
3.46
5.34
-
100.73
1.20
99.53
NUMBER
8?? S
0.27
0.15
5.42 -
0.44
4.61
0.25
3 06 2 07
0.77
1.57
5
34.92
0.54
11.99
-
30.57 *
2.27
11.34
0.80
3.47
5.06
-
100.95
1.14
99.81
OF IONS ON
!
\S
)(\
0.40 \
0.07 1
5.47 - y
0.58 4
4.38 *
0.27 1
3 09 2 08 I
0.76 )
1.47
6
36.66
0.41
11.43
-
26.99*
4.85
11.69
1.82
1.76
6.10
-
101.71
1.38
100.33
THE BASIS OF
8.00 J-JJJ8.00
0.33 \
0.05 1
5.43 - y 5.33
1.20 (
3.75 *
0.59 1
3 11 2 08 > 3 04
0.37 )
1.72
7
35.06
0.80
12.27
-
26.71
4.05
11.78
1.75
1.96
6.51
-
101.39
1.47
99.92
24 (O
5.91
2.09
0.35
0.10
-
1.02
3.76
0.57
2 13
0.42
1.86
8
35.76
0.70
12.55
-
26.32*
4.87
11.35
1.77
1.76
6.03
-
101.11
1.36
99.75
, Cl, OH)
18- ;s i
\ 0.41 X
/ 0.09 /
)5.23 - )
( 1,1 (
; 3.66;
j 0.57 1
> 3 12 2 02 i
\ 0.37 \
1.70
9
36.44
0.44
11.60
-
27.03*
4.42
11.35
1.86
1.66
6.20
-
101.00
1.40
99.60
8.00 ^[8.00
0.55 \
0.06 f
5.37 - y 5.25
0.87 (
3.77 ;
0.60 )
2 96 2 03 > 2 98
0.35 \
1.75
10
36.00
0.59
11.85
-
27.99*
3.51
11.42
1.79
1.71
6.10
-
100.96
1.38
99.58
6.05 /
1.95 \ '
0.40 \
0.07 1
- ) 5.29
0.88 I
3.94 ;
0.58 )
2 06 > 3 01
0.37 \
1.74
a All Fe calculated as FeO.
"Includes Mn 0.06.
Column 1: Krutov's (1936) analysis (includes P2O5 0.10, SO3 0.11, MnO 0.43)
Columns 2-5: Dashkesanite from Transcaucasia. Analyzed on electron microprobe.
Columns 6-10: Dashkesanite from the St. Paul's Rocks. All analyses done on section USNM-110393-31. Analyzed on electron
microprobe.
NUMBER 14 19
SjU.m beam size. The analyses were corrected
using the Bence and Albee (1968) correction pro-
cedures, and the summations corrected for chlorine
equivalents of oxygen. The results are given in
Table 1. The ionic proportions were calculated on
the basis of 24 anions with H2O not determined
and all Fe assumed to be ferrous. The choice of 24
anions rather than 23 was made because Cl~ rather
than OH" appears to nearly completely fill the
A site. Since the occupancy of this site is known
and need not be assumed, a normalization on the
basis of 24 anions is a justified approximation. The
actual number of anions is probably between 23
and 24.
Chemistry
Dashkesanite from the two localities has similar
chemistry in some respects. Both are low in SiO2
and MgO and high in FeO, Cl, and K2O compared
to other hornblende compositions (Leake, 1968;
Deer, et al., 1963a:290), while CaO, TiO2, and
A12O3 are consistent with other hornblendes. These
hornblendes are hastingsitic, using the classifica-
tion of Ernst (1968).
There are also chemical differences between the
two occurrences. The dashkesanite from the St.
Paul's Rocks is higher in SiO2, Cl, Na2O, MgO
and lower in total Fe and K2O than the dashkesan-
ite from Transcaucasia. There are also differences
between Krutov's original analyses and those done
for the present paper. The new analyses show less
Cl, SiO2, and Na2O, and higher A12O3. These
differences may be due to the differences in analyti-
cal technique.
The cation sums (Table 1) are high, generally
summing over 16. This is probably because for the
microprobe analyses all iron is assumed to be fer-
rous. Krutov's analysis shows a considerable amount
of ferric iron, and indicates that from 13 to 14
percent of the ferrous iron reported in the probe
analyses may be ferric.
Particularly, the totals for the Y site (Fe + Ti
+ Mg + octahedral Al) are high, being between
5.23 and 5.50. Leake (1968:249) postulates that
good analyses would have sums between 4.75 and
5.25 for this site. Recalculating the ionic propor-
tions assuming all iron as ferric iron reduces the
sum for this site by approximately 0.70 and reduces
the cation total by approximately 1. Depending on
the amount of ferric iron present, all the cation
proportions are probably less than those reported
for the present analyses in which all iron is as-
sumed as ferrous iron.
Discussion
Dashkesanite from the two localities is close in
optical characteristics and chemical composition in
spite of the differences in the environments in which
they are found. In St. Paul's Rocks dashkesanite is a
late stage mineral, as shown by the absence of shear-
ing or mylonitization. Melson, et al. (1972:241) in-
clude dashkesanite in the neomineral group. These
minerals occur as undeformed grains indicating
formation by recrystallization and precipitation
continuously during or after mylonitization. Melson,
et al. (1972:241) speculate the high-chlorine miner-
als found (dashkesanite, scapolite and chlorapatite)
formed either by reaction of primary minerals with
a chlorine-rich pore fluid or by direct crystallization
from such a pore fluid.
It is possible that the chlorine was derived from
sea water during weathering of the brown horn-
blende mylonites. The dashkesanite does occur with
scapolite, however, which is most characteristically
formed in upper amphibolite facies rocks (Deer,
et al., 1963b:330); this would indicate moderate to
high temperatures.
The formation of the dashkesanite may be re-
lated to the formation of the brown hornblende
mylonites. Of the three rock types present, the
brown hornblende mylonites have the largest
amount of chlorine in their bulk composition
(1.22-1.86% compared with 0.67% in the clinopy-
roxene-plagioclase mylonite and 0.05-0.20% in the
peridotite mylonites [Melson, et al., 1972]). The
higher amount of chlorine may account for the
observable presence of the dashkesanite and other
high chlorine phases in the brown hornblende
mylonites. High chlorine content appears to be
a characteristic of the strongly undersaturated
alkalic rock clan to which the brown hornblende
mylonites are chemically related (Melson, pers.
comm.).
The dashkesanite in the Trancaucasian rocks
formed by contact metamorphism. Krutov (1936)
feels that the dashkesanite was formed by reaction
of the fluids from the magma, containing large
amounts of FeO and Cl, with the surrounding
20 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
country rocks described as porphyrites and tuffs.
He projects conditions of low temperature and high
fluid pressure. Krutov (1936) thinks the source of
the fluids are the basic differentiates of the intru-
sion, which have a high chlorine content. These
basic differentiates occur as dikes of gabbro por-
phyrite near the amphibole skarn.
The differences in chemistry between the two
localities of dashkesanite as noted in the preceding
section can be accounted for by the differences in
environments. There are, however, strong similari-
ties between the two environments. Both occur-
rences are in silica-undersaturated environments
and this would account for the low SiO2 content of
dashkesanite. Although the specific conditions of
formation are not known, both occurrences of the
mineral appear to be related to the late stage
fluids derived from basic and ultrabasic intrusions.
Dashkesanite may be present but may not have
been recognized in other rocks of similar composi-
tion as the Transcaucaian rocks and St. Paul's
Rocks. Further observance of this mineral in other
silica-undersaturated rocks could add to our knowl-
edge of the geochemistry of chlorine, the concen-
tration of chlorine in such rocks, and the late stage
fluids associated with these rocks.
Dashkesanite may also prove interesting crystal-
lographically because the chlorine ion is the largest
(1.81 A) of the ions commonly reported in the A
site in amphiboles (OH~, 0.91-0.105 A and F~,
1.33 A [Bloss, 1971]).
Literature Cited
Bence, A. E., and A. L. Albee
1968. Empirical Correction Factors for the Electron
Microanalysis of Silicates and Oxides. Journal of
Geology, 76:382-403.
Bloss, F. Donald
1971. Crystallography and Crystal Chemistry. 545 pages.
New York: Holt, Rinehart and Winston, Inc.
Deer, W. A., R. A. Howie, and J. Zussman
1963a. Chain Silicates. Volume 2 of Rock-Forming Min-
erals. 379 pages. New York: John Wiley and Sons.
1963b. Framework Silicates. Volume 4 of Rock-Forming
Minerals. 435 pages. New York: John Wiley and
Sons.
Ernst, W. G.
1968. Amphiboles. 125 pages. New York: Springer-Verlag
New York, Inc.
Krutov, G. A.
1936. Dashkessanite: A New Chlorine Amphibole of the
Hastingsite Group. Bulletin de VAcademie des Sci-
ences de I'URSS, Classe des Sciences Mathematiques
et Naturelles, Serie Geologique, 341-373. [See also
Mineralogical Abstracts, 6(1937): 438.]
Leake, B. E.
1968. A Catalog of Analyzed Calciferous and Subcal-
ciferous Amphiboles Together with their Nomen-
clature and Associated Minerals. Geological Society
of America Special Paper, 98.
Melson, W. G., S. R. Hart, and G. Thompson
1972. St. Paul's Rocks, Equatorial Atlantic: Petrogenesis,
Radiometric Ages and Implications on Sea Floor
Spreading. Geological Society of America Memoir,
132:241-272.
Note on High-Alumina Basalt
Dredged near the Aleutian Trench
William G. Melson
ABSTRACT
High-alumina basalt was dredged from a smallseamount in the Adak Fracture Zone near and
south of the Aleutian Trench. The basalt, dredgedfrom magnetic anomaly 25 (about 63 million years
old) has a composition transitional between basalt
from spreading centers and basalt from seamounts.It thus may have been derived from the ancestral
East Pacific Rise or from younger off-ridge vol-canism. Extensive weathering effects have changed
the concentrations of some elements, and obscured
the primary igneous composition.
Introduction
Very little is known about the igneous rocks on
the floor of the north Pacific Ocean. Samples
dredged from this region were transferred to the
National Museum of Natural History, Smithso-
nian Institution, by Dr. Barrett H. Erickson of the
U.S. National Oceanographic and Atmospheric Ad-
ministration. These are the only samples in the
Museum's collections from this area, and their de-
scription is the subject of this report.
ACKNOWLEDGMENTS.?I am indebted to Dr. Erick-
son for providing the samples for study; to Mr.
Paul Grim, also of the U.S. National Oceanographic
and Atmospheric Administration, for helpful dis-
cussions and information on the Adak Fracture
Zone; to Mr. E. Jarosewich, National Museum of
Natural History, Smithsonian Institution, for per-
William G. Melson, Department of Mineral Sciences, Na-
tional Museum of Natural History, Smithsonian Institution,
Washington, D. C. 20560.
forming the chemical analyses; and to Dr. S. R.
Cann and his associates, University of East Anglia,
England, for performing trace element analyses on
the samples.
Specimen Features
Two basalt samples were received, both identical
in hand specimen features and weighing 195 gm
(USNM 112471) and 400 gm (USNM 112472). Both
are angular, and have very thin, black, hydrous
iron-manganese oxide coatings on at least one sur-
face. Other surfaces are free of the coatings, indi-
cating that the fragments were broken off larger
boulders or outcrops by the dredge. A thin, 5 to 10
mm, reddish altered zone extends inward into the
rock from the oxide-encrusted surfaces. The unen-
crusted surfaces are mottled by small black circular
encrustations, probably composed of hydrous iron-
manganese oxides, about 1 to 2 mm across, on a
yellowish background of altered basalt. The less-
altered appearing interior is dark brown, contains
scattered large (up to 1 cm across) rounded, multi-
crystal aggregates of plagioclase (An80), which com-
pose about 3 volume percent of each sample. The
groundmass contains abundant plagioclase micro-
lites (average An60-An70), and small granules and
sheaflike quench crystals of clinopyroxene. The
large plagioclase crystals are altered to white clay
minerals on the outer, oxide-encrusted surface. The
very fine grain-size of the samples shows that they
cooled near quenched surfaces, although neither
fresh glass nor palagonitized glass surfaces occur.
There are no obvious grain-size gradients across
the samples. These features indicate that the
samples are from the interior, but probably less
than 20 centimeters from a cooling surface, of a
21
22 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
pillow lava or from the contact of a shallow intru-
sion.
Bottom Exposure Age
The thickness of the hydrous iron-manganese
oxide coating on deep-sea rocks is proportional to
the length of time of exposure on the sea-floor.
The rate of deposition ranges between 1 and 10
mm per million years (Bender et al., 1966; Ku and
Broecker, 1967; Somayujulu, 1967). The maximum
thickness of 1 mm on these samples thus indicates
a bottom exposure age of from 0.1 to 1 million
years.
Sample Locality and Oceanic Basement Age
The sample was dredged from a depth of 4714
m on the south side of a small seamount which rises
to 3790 m. The locality (49?30.43' N, 177?32.88'
W) is on magnetic anomaly 25 (Grim and Erick-
son, 1969; Erickson and Grim, 1969), which has an
age estimated at 63 million years according to the
magnetic reversals time scale of Heirtzler, et al.
(1968). The thin hydrous iron-manganese encrusta-
tion shows that the sample has been exposed at the
sea floor for a much shorter period of time than
this. The inference, then, is that it has only re-
cently been exposed at the sea-floor by faulting or
slumping off of overburden, and/or it is a geo-
logically recent eruptive rock.
The unnamed seamount from which the samples
were dredged is elongated in an east-northeast direc-
tion, and rises 1210 m over a distance of 15 km in
this direction. Steepest slopes are perpendicular to
this direction, and average about 10?. The bathy-
metry itself would indicate formation of the feature
as an off-ridge submarine volcano. However, the
direction of elongation is nearly paralled to the
trend of the regional magnetic anomalies. This
feature raises the possibility that the seamount was
a bathymetrically high ridge formed originally at
a spreading center. The composition of the basalt
from the seamount, as will be discussed, does not
indicate unambiguously which of these possibilities
is true.
Composition
The Adak Fracture Zone basalt samples are high-
alumina basalt with evidence of significant weather-
TABLE 1.?Chemical composition of high-alumina
basalt from the Adak Fracture Zone
Constituent
SiO2
A1,O3
FesO,
FeO
MgO
CaO
Na2O
K2O
H2O+
H2O-
TiO2
MnO
Total
Total FeO*
Rb
Sr .
Y
Zr
Nb
46.80
17.99
7.52
3.52
4.48
12.14
2.84
0.37
1.66
0.77
1.67
0.17
0.19
100.12
10.05
49.61
16.01
2.21
7.19
7.84
11.32
2.76
0.22
0.63
0.45
1.43
0.14
0.18
0.72
0.85
0.74
1.25
0.90
0.64
0.25
0.12
0.22
0.30
0.29
0.07
0.04
49.84 48.16
14.09 18.31
3.06
8.61
8.52
10.41
2.15
0.38
4.24
5.89
4.87
8.79
4.05
1.69
2.52 2.91
0.26 0.93
0.16 0.16
11.49 1.27 11.36 9.71
TRACE ELEMENTS
7.2
148
57
111
0.8
123
43
100
46
12
42
* All Fe recalculated as FeO.
Column 1: Using classical methods, E. Jarosewich, analyst;
trace elements by X-ray fluorescence by J. R. Cann and co-
workers.
Column 2: Average ocean-floor basalt (Cann, 1971), 94 analy-
ses; Fe2O3, FeO, H2O + , H2O ?, and trace element averages
(Melson and Thompson, 1971).
Column 3: Standard deviations for column 2.
Column 4: Average of 181 olivine tholeiites and tholeiites from
the Hawaiian Islands (MacDonald and Katsura, 1964).
Recalculated on an anhydrous basis.
Column 5: Average of 10 samples of alkali basalt from sub-
marine volcanoes and islands of the eastern Pacific Ocean
(Engel, et al., 1965). Recalculated on an anhydrous basis.
ing (Table 1). The ferric to ferrous iron ratio is
considerably greater than fresh oceanic basalts, and
reflects submarine oxidation. The water content, on
the other hand, is higher (2.43 versus 1.08) than
most midocean ridge basalts (Table 1). Weathering
of deep-sea basalts commonly produces potassium-
rich smectite (Melson and Thompson, 1973) which
significantly increases the bulk K2O content of even
slightly weathered submarine basalts. Although
clay minerals are abundant in the mesostasis, the
K2O content of the sample, 0.37 percent, is barely
higher than the average for fresh ocean-floor basalt
(0.22, Table 1).
NUMBER 14 23
TABLE 2.?Trace element composition (ppm) of the
Adak Fracture Zone high-alumina basalt compared
to other regions (all data and averages by Dr. J. R.
Cann and colleagues; averages from Pearce and
Cann, 1971)
Sample
Adak Basalt
USXM 112471 . . .
USNM 112472 . . .
Average Ocean
floor basalt
Average Island
arc andesitc
Average Hawaii
basalt
Ti
10012
-
7760
4050
15360
Rb
7.2
9.7
-
-
-
Sr
148
144
-
-
-
Y
57
49
27
21
24
Zr
111
108
80
106
162
Nb
0.8
4.2
-
-
?
Origin of High-Alumina Content
High-alumina content can be a feature of erupted
liquid, or a reflection of an abundance of already
present calcic plagioclase phenocrysts. In the Adak
high-alumina basalt, the high-alumina content is a
reflection of a high-alumina liquid composition.
Although large rounded plagioclase crystals occur,
they compose on the order of 3 volume percent of
the samples, an amount much too small to account
for the high-alumina content of the sample. The
large, rounded calcic plagioclase crystals are prob-
ably cognate, and may reflect convective movements
or settling of early forming plagioclase crystals
from cooler to hotter portions of the pre-eruption
magma chamber.
In this region of the Adak Fracture Zone old
oceanic crust formed by sea-floor spreading at the
northern extension of the east Pacific rise about
63 million years ago? Or, is the seamount from
which it was dredged the product of much younger,
off-ridge volcanism? These two modes of origin
correlate with distinct differences in chemical com-
position. A number of elements whose concentra-
tions commonly differ in these two settings were
determined in the Adak sample. These include Ti,
K, and Zr, which are characteristically higher in
volcanic islands and seamounts compared to deep-
sea basalts from spreading ridges.
The most striking compositional differences be-
tween the Adak Fracture Zone basalt and most
ocean-floor basalts, are its high alumina content,
low magnesia content, and high total iron to mag-
nesia ratio. These features are probably not a
result of weathering. The Zr and Ti contents are
transitional between seamount basalt and basalt
derived from spreading centers (Table 2). Chem-
ically, it is thus not clear to which category the
Adak Fracture Zone basalt belongs.
If actually derived from a spreading center about
63 million years old, its chemically and perhaps
even bathymetric corresponding half would be
identifiable on the opposite side of the spreading
center at magnetic anomaly 25. This test of a sea-
floor spreading origin for the basalt cannot be
made because the eastern counterpart of the re-
gion from which the Adak basalt may have come
has long since been subducted beneath North
America.
Literature Cited
Bender, M. L., T. Ku, and W. Broecker
1966. Manganese Nodules: Their Evolution. Science, 151:
325-328.
Cann, J. R.
1971. Major Element Variations in Ocean-Floor basalts.
Philosophical Transactions of the Royal Society of
London, A.268:495-505.
Engel, A. E. J., C. G. Engel, and R. G. Havens
1965. Chemical Characteristics of Oceanic Basalts and the
Upper Mantle. Geological Society of America
Bulletin, 76:719-734.
Ericksson, B. H., and P. J. Grim
1969. Profiles of Magnetic Anomalies South of the Aleu-
tian Island Arc. Geological Society of America
Bulletin, 80:1387-1389.
Grim, P. J., and B. H. Ericksson.
1969. Fracture Zones and Magnetic Anomalies South of
the Aleutian Trench. Journal of Geophysical Re-
search, 74:1488-1494.
Heirtzler, J. R., G. O. Dickson, E. M. Herron, W. C. Pitman
III, and X. Le Pichon
1968. Marine Magnetic Anomalies, Geomagnetic Field
Reversals, and Motions of the Ocean Floor and
Continents. Journal of Geophysical Research, 73:
2119-2136.
Ku, T., and W. S. Broecker
1967. Uranium, Thorium and Protactinium in a Manga-
nese Nodule. Earth and Planetary Science Letters,
2:217-320.
MacDonald, G. A., and T. Katsura
1964. Chemical Composition of Hawaiian Lavas. Journal
of Petrology, 5:82-133.
Melson, W. G., and G. Thompson
1971. Petrology of a Transform Fault Zone and Adjacent
Ridge Segments. Philosophical Transactions of the
Royal Society of London, A.268:423-441.
1973. Glassy Abyssal Basalts, Atlantic Sea Floor near St.
24 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
Paul's Rocks: Petrography and Composition of Analysis Using Ti, Zr, and Y. Earth and Planetary
Secondary Clay Minerals. Geological Society of Science Letters, 12:339-349.
America Bulletin, 84:703-716. Somayujulu, B. L. K.
Pearce, J. A., and J. R. Cann 1967. Beryllium-10 in a Manganese Nodule. Science, 156:
1971. Ophiolite Origin Investigated by Discriminant 1219-1220.
Composition of Three Glass Phases
in an Apollo 15 Basalt Fragment
George S. Switzer
ABSTRACT
The fragment studied contains glass of three dis-
tinct types: A: colorless, high silica glass; B: brown,
high-iron interstitial glass. Microprobe analyses show
high-iron interstitial glass. Microprobe analyses show
there has been strong partitioning of certain ele-
ments, especially of K into the high-silica glass,
and P and Ti into the high-iron glasses. The high-
iron glasses are thought to have been formed as im-
miscible liquids.
Introduction
Some silicate liquids on cooling separate into two
or more liquids of markedly different compositions.
This process, generation of immiscible liquids, may
be an important process in the generation of cer-
tain classes of igneous rocks, such as those in the
syenitic and granitic families (Roedder, 1951; Phil-
potts and Philpotts, 1969; Roedder and Weiblen,
1970). Immiscibility has long been known in cer-
tain experimentally studied systems (Greig, 1927;
Roedder, 1951), but was first documented in natural
materials in Apollo 11 basaltic rocks (Roedder and
Weiblen, 1970). Here I describe one of the best
examples of liquid immiscibility yet reported in
lunar samples.
During a study of Apollo 15 soil sample 15272,11
an unusual basalt fragment was encountered that
contains glass of three distinct types, which for
convenience are designated types A, B, and C: A
George S. Switzer, Department of Mineral Sciences, National
Museum of Natural History, Smithsonian Institution, Wash-
ington, D.C. 20560.
is a colorless, high-silica glass; B is brown, high-
iron blebs immiscible in A; and C is a brown, high-
iron interstitial glass.
The sample is an angular, 2.0 mm fragment com-
posed primarily of plagioclase, orthopyroxene, glass,
cristobalite, ilmenite, and clinopyroxene, in that
order of abundance. The density of the fragment is
> 2.90 < 3.00. A photomicrograph is shown in
Figure 1.
Microprobe analyses and microscopic examina-
tion yielded the following data on the principal
minerals. Plagioclase crystals up to 0.2 X 0.8 mm in
size have a composition near An86, with slightly
more sodic (An83) rims. The orthopyroxene, in
crystals up to 0.5 mm, has composition En75Fs25Wo2.
A single crystal of clinopyroxene is En34Fs26Wo40.
FIGURE 1.?Thin section of basalt fragment from Apollo 15
soil sample 15272,11. Maximum dimension of fragment 2.0
mm. Transmitted light.
25
26 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
Vo ????? A
FIGURE 2.?Umenite crystal surrounded by interstitial Type C glass except where in contact
with plagioclase. Area on left with mosaic structure is cristobalite. Note darkening of glass
due to reaction with ilmenite. a, Transmitted light, b, Reflected light. Length of bar = 0.1
mm.
The cristobalite shows typical mosaic structure.
Droplets of iron and possibly some sulfide are
scattered throughout the specimen. The ilmenite
contains 2.8 percent MgO.
Ilmenite-Glass Reaction
Ilmenite is a prominent phase (about 5 volume
percent). Where ilmenite is in contact with Type
C glass, the glass is frequently dark brown to
opaque, with the depth of color decreasing away
from the contact, as though ilmenite has reacted
with the glass (Figure 2a, b). A microprobe step
scan for Fe and Ti at 2/j.m intervals across an
ilmenite-glass contact (along A-Ar of Figure 3a)
yielded the data plotted in Figure 4. There has
been some reaction between the two phases, but an
apparent more extensive reaction seen in the thin
FIGURE 3.?Ilmenite crystal in contact with (to left) Type C interstitial glass, and (to right) with
high-silica glass (Type A) with high-iron (Type B) immiscible blebs. A-A' is line along which
step scan analysis was made (see Figure 4). a, Transmitted light, b, Reflected light. Length
of bar = 0.1 mm.
NUMBER 14 27
u
UI
CL
z
Ul
uac
uia.
30 -
20 -
10 -
10 -
-
-
HIGH IRON
GLASS /
jTYPE C
1 1 1 1
Tl ^? '
r1
in
i?
z
S
_?
i ___-?
??
1
\" <
11 -J **i
r1
1\
\ ^
i
HIGH SILICA
GLASS
TYPE A
HIGH IRON
GLASS
TYPE B
r
1
1 i i
HIGH SILICA
GLASS
TYPE A
Tl
20 40 60
MICRONS
80
FIGURE 4.?Microprobe step scan for Fe and Ti along A-A' of Figure 3.
100
T '
FIGURE 5.?Large area of immiscible glass (left) consisting of high-silica colorless glass (Type
A) with blebs of high-iron (Type B) glass. Right half of each photograph is same area as
shown in Figure 3. a, Transmitted light, b, Reflected light. Length of bar = 0.1 mm. Circular,
highly-reflective area in upper left is iron.
28 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
TABLE 1.?Composition of glass phases
Constituent
SiO,
A12O,
FeO
CaO
MgO
K2O
Na2O
TiO2
P,O5
MnO
BaO
Total
Qtz
Or
Ab
An
Di
Hy
11
Ap
C
Apollo 15 Apollo 11 and 12
Type A Type B Type C
12 3 4 5 6 7
ANALYSES
75.7 53.1 43.4 76.3 75.8 42.2 44.1
11.9 8.3 8.0 11.5 11.4 5.4 3.0
2.7 24.8 24.6 3.0 2.5 32.9 31.2
1.5 7.1 10.3 1.6 1.8 10.7 10.9
0.3 0.5 4.7 0.07 0.25 0.79 1.9
6.6 1.9 0.8 6.7 6.4 0.41 0.25
1.1 0.4 0.4 0.14 0.35 0.16 0.11
0.5 2.7 6.3 0.68 0.53 4.2 3.8
0.0 1.1 0.8 0.15 - 0.57 1.2
0.0 0.2 0.2 0:05 - 0.32
0.3 0.6 0.2 0.62 - 0.0
100.6 100.7 99.7 100.81 99.03 97.65 96.46
NORMS
38.5 14.5 1.3 44.7 43.4 0.5 5.0
39.0 11.2 4.7 39.6 37.8 2.4 1.5
9.3 3.4 3.4 1.1 3.0 1.4 0.9
7.4 15.2 17.7 7.0 8.9 12.8 7.0
11.4 24.2 - - 32.4 34.6
4.9 36.7 34.4 4.7 4.3 38.9 37.6
1.0 5.1 12.0 1.3 1.0 8.0 7.2
2.6 1.9 0.4 - 1.3 2.8
0.2 - - 1.5 0.6
Column 1: High-silica immiscible colorless glass.
Column 2: High-iron immiscible brown glass bleb.
Column 3: High-iron brown interstitial glass (average of 2 analyses).
Column 4: High-silica colorless glass from Apollo 12.
Average of 15 analyses (Roedder and Weiblen, 1971).
Column 5: High-silica colorless glass from Apollo 11.
Average of 35 analyses (Roedder and Weiblen, 1971) .
Column 6: High-iron brown glass from Apollo 12.
Average of 5 analyses (Roedder and Weiblen, 1971) .
Column 7: High-iron brown glass from Apollo 11.
Average of 7 analyses (Roedder and Weiblen, 1971).
section is due to inclined grain boundaries in a
section that is approximately 40 ^m thick.
The Glass Phases
Microprobe analyses of the three glasses are given
in Table 1. The analyses were made using close
standards and corrected by use of the ABFAN pro-
gram of the Geophysical Laboratory (Boyd, et al.,
1969). For comparison are given compositions of
immiscible glass in Apollo 11 and 12 rocks de-
scribed by Roedder and Weiblen (1971).
The areas of high-silica glass with immiscible
blebs of high-iron glass (Figure 5a, b) are similar
to those described in Apollo 11 and 12 rocks by
Roedder and Weiblen (1970, 1971). Comparison of
the high-silica glasses (Table 1, columns 1, 4, and
5) shows them to be similar, and characterized (in
addition to high-silica), by low iron, calcium, and
magnesium, and high potassium.
The correspondence between analyses of the
brown immiscible blebs (columns 2, 6, and 7) is not
NUMBER 14 29
\ 1
..<
FIGURE 6.?Area of Type C high-iron glass interstitial to plagioclase and surrounding crystal
of ilmenite. Triangular area above is high-silica glass (Type A) with high-iron glass blebs
(Type B). a, Transmitted light, b, Reflected light.
as close. However, it should be pointed out that
Roedder and Weiblen's analyses of high-iron glasses
in Apollo 11 and 12 rocks are averages of seven and
five analyses, repectively, with a range in SiO2 of
36.5-50.2 percent (for Apollo 11), and 40.0-45.0 (for
Apollo 12). These authors point out that the high-
iron glass in their samples varied somewhat in com-
position between samples, between blebs, and even
within blebs. The new analysis presented here for
Apollo 15 is of a single 20jU,m diameter bleb.
An analysis of the high-iron Type C glass is
given in Table 1, column 3. This glass is clear and
vitreous and typically interstitial (Figure 6), and
quite distinct in this latter feature from the Type
B immiscible blebs. It also differs chemically from
Type B, as can be seen by comparison of columns
2 and 3 of Table 1.
Discussion
Apollo 15 glass analyses in Table 1 show that
there has been strong partitioning of certain ele-
ments into one or the other liquid, the same as
found by Roedder and Weiblen (1970, 1971) in
their studies of immiscible glasses in Apollo 11 and
12 rocks. Of particular interest is the high con-
centration of K in the high-silica glass and the high
concentration of P in the high-iron glasses. The
depletion of phosphorus in the high-silica (gra-
nitic) fraction is a reverse trend to that in normal
fractional crystallization and is thought to be the
result of the separation of the liquids by immisci-
bility (Anderson and Greenland, 1969, and Roed-
der and Weiblen, 1970, 1971).
Because phosphorus is enriched in both types (B
and C) of high-iron glasses described here, one can
conclude that both were formed as immiscible
liquids. The question then is why one of the two
high-iron glasses occurs as blebs in high-silica glass,
and the other interstitial to the plagioclase and
pyroxene. One possible explanation is that, due to
a complex cooling history, initially two immiscible
liquids formed, of compositions (A + B) and C.
Then on further cooling (A + B) further separated
into Type B blebs immiscible in Type A.
Literature Cited
Anderson, A. T., and L. P. Greenland
1969. Phosphorus Fractionation Diagram as a Quantita-
tive Indicator of Crystallization Differentiation of
Basaltic Liquids. Geochimica and Cosmochimica
Ada, 33:493-505.
Boyd, F. R., L. W. Finger, and F. Chayes
1969. Computer Reduction of Electron-probe Data. Car-
negie Institution Year Book 67, Annual Report of
the Director, Geophysical Laboratory, pages 210-215.
Greig, J. W.
1927. Immiscibility in Silicate Melts. American Journal of
Science, series 5, 13:1-44.
30 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
Philpotts, A. R., and J. A. Philpotts
1969. Liquid Immiscibility Between Syenitic and Gabbroic
Magmas (Abstract). Geological Society of America
Abstracts 1969, pages 176-177.
Roedder, E.
1951. Low Temperature Liquid Immiscibility in the
System K2O-FeO-Al2O3-SiO2. American Mineralo-
gist, 36:282-286.
Roedder, E., and P. W. Weiblen
1970. Lunar Petrology of Silicate Melt Inclusions, Apollo
11 Rocks. Pages 801-837 of volume 1 in A. A. Levin-
son, editor, Proceedings of the Apollo 11 Lunar
Science Conference, Houston, Texas, January 5-8,
1970. Geochimica and Cosmochimica Ada, supple-
ment 1, volume 34.
1971. Petrology of Silicate Melt Inclusions, Apollo 11
and Apollo 12 and Terrestrial Equivalents. Pages
507-528 of volume 1 in A. A. Levinson, editor,
Proceedings of the Second Lunar Science Con-
ference, Houston, Texas, January 11-14, 1971.
Geochimica and Cosmochimica Acta, supplement
2.
Petrographic Analysis of Apollo 16 Samples
66083,1 and 67943,1
Brian Mason
ABSTRACT
Two samples of 2-4 mm fines collected on the
Apollo 16 mission consist largely of fragments of
anorthositic rock and breccias derived therefrom.
Sample 66083,1 was a 0.5 g sample of 2-4 mm
fines from a clod of white impact ejecta collected at
Station 6 on the southwest wall of a subdued 10
m crater on the lowest "bench" of Stone Mountain,
near its base; the astrogeology report (Apollo 16,
1973) says "possibly from South Ray crater, but
location generally shadowed from South Ray crater
ejecta."
The sample as received contained 12 fragments
> 2 mm, with a total weight of 0.433 g, and an
average weight of 0.036 g; there was 0.066 g of
material < 2 mm. The > 2 mm fragments were
density-fractionated in methylene iodide-acetone
liquids, with the results as shown in Table 1.
Polished thin sections were made of fragments
B, C, D, E, F, and of one from 60083,1 A (designated
as 66083,1AA). Half of fragments AA and F were
ground and fused with lithium tetraborate flux,
and the resulting glasses analyzed with the micro-
probe, with the results given in Table 2. A polished
thin section (66083,1H) was also prepared from the
0.2-0.4 mm material obtained by sieving from the
coarser fragments; this material contained 334 in-
dividual particles, which were classified as shown
in Table 3.
Brian Mason, Department of Mineral Sciences, National
Museum of Natural History, Smithsonian Institution, Wash-
ington, D.C. 20560.
There appears to be a continuous sequence be-
tween breccia and crystalline rock fragments, and
for some fragments the assignment to one class or
the other is to some degree arbitrary.
Microprobe analyses of mineral grains gave the
following results (number of grains analyzed in
parentheses): plagioclase (10): An88_97, mean An93;
olivine (4): Fa2o-25> mean Fa23; orthopyroxene (6):
Wo2-4, Fs21_26, mean Wo3Fs25; augite (1): Wo41Fsn
En48. Six glass fragments were also analyzed, with
the results as shown in Table 4. Glasses 1, 2, and
3 could be melted regolith from this site; 2 and 3
are very similar to AA in Table 2, while 1 is some-
what more feldspathic. Glass 6 corresponds to a
completely melted mare basalt. The interpretation
of the compositions of glasses 4 and 5 is not so
clear-cut, and they probably represent mixtures of
different components.
The petrography of the fragments larger than 2
mm is as follows:
AA. A breccia with a glassy matrix containing
about 20-30 percent of crystal fragments. The frag-
ments are predominantly plagioclase (maximum
TABLE 1.?Density-fractionation results
for 66083,1
Fragment
A
B
C
D
E
F
Total
Density
<2.70
2.72
2.85
2.90
3.03
3.07
Number of
fragments
7
1
1
1
1
1
12
Grams
0.208
0.030
0.042
0.029
0.025
0.099
0.433
31
32 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
TABLE 2.?Analyses and norms of two fragments
from 66083,1 (analyst: J. Nelen)
TABLE 4.?Analysis of six glass fragments from
66083,1 H
Constituent
SiO2 ....
TiO2
A12O3 . .
Cr2O3 . .
FeO ...
MnO
MgO .
CaO .
Na2O . .
K2O . .
P2O5 ...
Total
An
Ab
Or
Di
Hy
01
II
Ap
Cr
AA
ANALYSES
44.8
0.62
25.5
0.14
5.76
0.12
6.11
14.7
0.52
0.18
0.1
47.3
0.92
19.2
0.18
9.20
0.16
9.86
11.4
0.59
0.32
0.3
98.6 99.4
NORMS
66.4
4.4
1.1
4.7
13.5
6.6
1.2
0.2
0.2
48.8
5.0
1.9
4.9
31.6
3.9
1.8
0.7
0.3
AA = fragment with D<2.70; F = fragment with D = 3.O7.
size 0.3 mm), with minor olivine and orthopy-
roxene, and some lithic fragments, also mainly
plagioclase. Some of the plagioclase fragments show
bending of the twin lamellae. This fragment ap-
pears to be typical for the material with D <2.70;
its analysis (Table 2) is very similar to that of fines
TABLE 3.?Classification of 334 particles from
66083 JH
Particles
Crystalline rock fragments . .
Breccia fragments
Glass (clear and devitrified)
Plagioclase fragments
Plagioclase, partly melted . . .
Spherules, glass or devitrified
Metal fragment
Total
Number Percent
126
125
36
31
8
7
1
334
37.7
37.4
10.8
9.3
2.4
2.1
0.3
100.0
Constituent
SiO2
TiO2
A12O3
FeO
MnO
MgO
CaO
Na2O
K2O
Total
1
44.0
0.54
31.2
3.58
0.03
3 39
180
0.26
0.03
101.0
2
46.6
0.60
27.8
5.32
0.03
5.87
15.8
0.94
0.06
103.0
3
44.5
0.88
26.8
5.99
0.03
6.49
14.3
0.69
0.11
99.8
4
47.7
0.53
20.9
7.08
0.03
12.0
12 0
0.62
0.08
100.9
5
50.9
0.69
17.5
10.9
0.10
9.45
10.2
0.75
0.11
100.6
6
45.7
3.88
9.89
20.4
0.15
10.0
7.94
0.86
0.24
99.1
from the same location (66081, H. Rose pers.
comm.), indicating that it represents partially
melted soil.
B. A fragment of partly devitrified glass, with
some plagioclase fragments (maximum size 0.1 mm)
in one area; one spherule of metal, 0.1 mm di-
ameter, was noted.
C. A fine-grained feldspar-rich rock, made up
of a felted aggregate of small plagioclase laths
(up to 0.01 mm long) with about 10 percent of
interstitial pyroxene and olivine; and minor
amounts of opaques (probably ilmenite); occasional
larger equidimensional grains of plagioclase (up to
0.1 mm) are also present. The texture suggests a
recrystallized breccia, with the larger plagioclase
grains being relict fragments from a coarse-grained
rock.
D. A breccia consisting of clasts of plagioclase
(up to 0.25 mm) and plagioclase-rich rock (up to
2 mm) in a fine-grained microcrystalline feldspathic
groundmass; four metal spherules, the largest 0.35
mm in diameter, are present. Microprobe analyses
give an average plagioclase composition of An90
(range An88-An94); four pyroxene grains show a
range of Wo43.116, Fs17.6_22.o; one grain of olivine
gave Fa29.
E. A fragment consisting of plagioclase clasts
(up to 0.3 mm) in a fine-grained groundmass of
plagioclase and ferromagnesians (olivine and pi-
geonite) in subequal amounts, with some opaques
(ilmenite and a few metal grains); the texture of
the groundmass suggests that it may represent a
recrystallized glass. Microprobe analyses of the
individual minerals (number of grains analyzed in
parentheses) gave plagioclase (9), An84_An95, mean
NUMBER 14
TABLE 5.?Density-fractionation results for 67943,1
Density
Number of
fragments
TABLE 6.?Analyses and norms of three fragments
from 67943 (analyst: J. Nelen)
Grams
<2.70
2.70-2.80
2.80-2.90
2.90-3.00
2
6
5
3
0.061
0.051
0.199
0.049
An90; pigeonite (7), Wo8_12, Fs22_24, mean Wo9,Fs23;
olivine (14), Fa25_27, mean Fa26.
F. A crystalline rock made largely of subequal
amounts of plagioclase and ferromagnesians (py-
roxenes and olivine), with a minor amount of
opaque material (mainly ilmenite, but with some
prominent metallic spherules up to 0.4 mm in
diameter). There are distinct areas of different
grain size, suggesting the rock is a recrystallized
breccia of lithic and mineral fragments; this is
supported by the presence of occasional large
(up to 0.4 mm) plagioclase grains, probably por-
phyroclasts.
Microprobe analyses of the individual minerals
gave the following results (number of grains ana-
lysed in parentheses): plagioclase (12): An90_98,
mean An95; olivine (18): Fa38_43, mean Fa41; ortho-
pyroxene (17): Wo3_7, Fs25_32, mean Wo4Fs28; pi-
geonite (6): Wo]0_12, Fs30_31, mean WonFs31; augite
(18): Wo34_40, Fs16_20, mean Wo37Fs17; metal (7):
Ni 4.7-6.5 percent, mean 5.1 percent, and Co
0.58-0.77 percent, mean 0.67 percent.
The analysis (Table 2) shows that the fragment
is of noritic composition, but the texture is horn-
felsic rather than igneous, suggesting recrystalliza-
tion of a norite breccia at high temperatures
(perhaps high enough to produce a metal-sulfide
melt, represented by the metallic spherules, but not
high enough to melt the silicates). The fragment is
very similar in composition and texture to 65015, a
1.8 kg rock collected at the adjacent Station 5 and
ascribed by the Geology Investigation Team to
South Ray crater ejecta.
Sample 67943,1 was a 0.5 g sample of 2-4 mm
fines, one of the sieve fractions from 67940, a soil
sample collected from the House Rock area on the
southeast of North Ray crater (Station 11 of the
Apollo 16 traverses).
The sample as received contained 16 fragments
>2 mm, with a total weight of 0.453 g, and an
Constituent
SiO2 ....
TiO2
A12O3 . . .
FeO
MnO . . .
MgO
CaO . .
Na2O . .
K2O ...
Total
An
Ab
Or
Di
Hy
Ol
II
An
Fa
Pyx
D = 2.70-2.80 D = 2.70-2.80 D=2.80-2.90
44.9
0.42
29.6
3.52
0.05
3.22
16.8
0.68
0.10
99.3
77.6
5.8
0.6
4.9
4.8
4.9
0.8
93
35.8
ANALYSES
43.4
0.35
29.2
3.58
0.08
3.24
16.4
0.60
0.09
96.9
NORMS
76.6
5.1
0.6
3.9
3.5
6.4
0.7
93
36.8
44.6
0.44
27.9
4.29
0.07
5.43
15.4
0.67
0.15
99.0
72.4
5.7
0.9
3.2
6.1
9.5
0.8
91
29.8
Ca25Mg48Fe27 Ca2eMg47Fe27
average weight of 0.029 g; there was 0.047 g of
material <2 mm. The >2 mm fragments were
density-fractionated in methylene iodide-acetone
liquids, with the results as shown in Table 5. Com-
pared to 66083, another sample of 2-4 mm fines
analyzed, this sample shows a smaller density range,
and a concentration of material where D = 2.70-
2.90.
Polished thin sections were made of 11 fragments
covering the full density range; for three of them
sufficient material remained after sectioning for
microprobe analysis after fusion with lithium tetra-
borate flux, and the results are given in Table 6.
Except for two fragments in the D = 2.90-3.00
fraction, the individual fragments could all be
classed as brecciated anorthosites; they differed one
from another mainly in the amount of glass in the
matrix, the lower density fragments having larger
amounts of glass. This glass was usually pale brown
and isotropic, sometimes showing incipient devitri-
fication. The analyses in Table 6 confirm the
anorthositic composition of these fragments; they
all show over 80 percent normative plagioclase of
34 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
anorthite composition, with minor amounts of
olivine and pyroxene.
Of the two non-anorthositic fragments, both
with density between 2.90 and 3.00, one was an
equigranular plagioclase-rich microgabbro, with in-
dividual grains averaging 0.1 mm across. The esti-
mated mineral composition is 65 percent plagioclase
(An94), 15 percent olivine (Fa19), 15 percent pigeon-
ite (Wo12Fs23En65), and 5 percent opaques (mostly
ilmenite). The compositions given for the minerals
are averages of microprobe analyses; plagioclase and
olivine are very uniform in composition, whereas
pigeonite shows minor variability from grain to
grain.
The other non-anorthositic fragment appeared to
be a composite particle, consisting of a relatively
coarse (plagioclase grains up to 0.6 mm) troctolite
welded to a fine-grained (maximum grain size 0.1
mm) gabbroic anorthosite. However, the miner-
alogy of the two parts of the fragment are identical
?plagioclase (An94), olivine (Fa21), minor hyper-
persthene (Fa19), and a little opaques?and the
fine-grained material may be comminuted and re-
crystallized troctolite. The boundary between the
two parts is quite sharp.
These two non-anorthositic fragments may repre-
sent primary igneous rocks from the Apollo 16 site
that have escaped the ubiquitous brecciation and
recrystallization so characteristic of most of the
rocks collected on this mission. Alternatively, they
may be exotic fragments transported to the Apollo
16 site from some other region on the Moon.
ACKNOWLEDGMENT.?This research was supported
by a grant from the National Aeronautics and
Space Administration (NGR 09-015-146).
Literature Cited
Apollo 16 Preliminary Examination Team
1973. The Apollo 16 Lunar Samples. Science, 179:23-34.
The Allende Meteorite: Chondrule Composition
and the Early Histroy of the Solar System
Andrew L. Graham
ABSTRACT
Fifty chondrules and aggregates from the Allende
meteorite (Type III carbonaceous chondrite), havebeen analyzed for the major elements using a fu-
sion technique and subsequent microprobe analysis
of the resulting glass bead. A portion of eachsample was retained and made into a polished thin
section for mineralogical studies. The majorelement chemistry suggests that the Ca/Al-rich
chondrules and aggregates exhibit compositional
changes, whereas the Mg/Fe-rich chondrules donot. There appears to be no simple relationship
between these two chondrule groups. Concomitantwith a decrease in the (CaO + Al
2O3) content is adecrease in TiO
2 and an increase in Na2O. This
is paralleled by changes in the mineralogy of thechondrules. Although the (CaO-f A1
2O3) to TiO2relationship is that which has been predicted in
terms of a condensing solar nebular model, that of(CaO + Al
2O3) with Na2O does not conform to this
model. Other evidence against the primary natureof the Ca/Al-rich Allende chondrules is cited.
Introduction
The Allende meteorite is a Type III carbona-
ceous chondrite of the Vigarano subgroup (Clarke,
et al., 1970). It is composed of chondrules and
microcrystalline aggregates distributed heteroge-
neously throughout a black, olivine-rich matrix.
Some of the chondrules and aggregates contain min-
erals rarely found in meteorites. These include
melilite, magnesian spinel, fassaite, wollastonite,
cordierite, hibonite, and grossular (Clarke, et al.,
Andrew L. Graham, Department of Mineralogy, British
Museum (Natural History), London S.W. 7, England.
1970; Fuchs, 1969, 1971; Marvin, et al., 1970). From
a possible composition of the primitive (that is, un-
differentiated), solar nebular gas (e.g., table in
Cameron, 1968:127), and making assumptions as to
the initial temperature of this gas, the condensation
sequence with falling temperature of stable mole-
cules can be calculated. Such calculations (Lord,
1965; Grossman, 1972) indicate that corundum,
spinel, and melilite form early in this sequence. It
is, therefore, tempting to suggest that the occurrence
of these minerals in Type III carbonaceous chon-
drites is evidence supporting the thermodynamically
derived condensation sequence. It is not at present
known whether the Ca/Al-rich aggregates and the
melilite-bearing chondrules of the Allende and simi-
lar meteorites are "bulk primary" condensates or
not. Does their bulk composition represent a quasi-
equilibrium state, similar to that shown by ter-
restrial igneous processes, in which the crystalline
solids represent one side of an equilibrium solidifi-
cation sequence, or are these chondrules and aggre-
gates the result of the association of material after
the condensation sequence had finished?
The structure and chemistry of the carbonaceous
chondrites provide the basis for a strong argument
against the metamorphism of these meteorites sub-
sequent to their accretion. This is particularly
shown by the wide variation in the Fe/Fe + Mg of
the olivines they contain, and by the presence of
glass within the chondrules. The bulk chemistry
of the chondrules has not, therefore, been changed
by metamorphism since the formation of the mete-
orite. In this work chondrules and aggregates
separated from the Allende meteorite have been
analyzed for their major elements to see whether
there is a sequential relationship in their chemistry.
Do they represent particular stages in the condensa-
35
36 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
TABLE 1.?The composition of some chondrules and aggregates from the
Allende meteorite
Constituent
SiO2
TiOa
A12O3 . .
FeO
MgO
CaO
Na2O
ICO
Cr2O3
Total
1
43.1
0.2
5.7
8.9
38.6
3.8
0.7
0.03
0.5
101.53
2
50.9
0.4
1.2
7.6
37.9
0.9
0.2
0.03
0.8
99.93
3
40.6
0.2
7.0
6.6
38.7
4.5
0.5
0.03
0.7
98.83
4
32.0
1.4
27.3
4.8
12.4
21.5
0.8
0.03
0.2
100.43
5
41.5
1.1
25.8
2.2
14.8
12.7
1.3
0.03
0.4
99.83
6
26.9
1.3
30.0
1.0
9.8
30.9
0.2
0.03
0.05
100.18
7
35.7
1.2
24.8
0.9
7.1
29.8
0.4
0.03
0.02
99.95
8
30.9
1.0
28.1
0.3
9.8
30.3
0.2
0.03
0.01
100.64
Columns 1-3: Mg/Fe-rich chondrules
Columns 4-5: Fine-grained aggregates
Columns 6-8: Coarse-grained Ca/Al-rich chondrules
tion of the nebular gas or are they random associa-
tions which have been formed for material produced
by this primitive condensation? If the latter is the
case then no chemical trends would be expected,
while if the former is the case then a sequential
chemistry may be shown.
ACKNOWLEDGMENTS.?This work was performed
during the tenure of a Smithsonian Research Fel-
lowship. Dr. B. Mason and Dr. R. F. Fudali are
thanked for their assistance and also Dr. L. Walter
who gave advice on the analytical procedure used.
The constructive criticism and suggestions of Dr.
A. A. Moss and Dr. R. F. Fudali much improved
the manuscript.
Method
Chondrules and aggregates were picked from
fractured faces of specimens of the Allende mete-
orite. The aggregates, which are friable, were re-
moved from the bulk material with a stainless steel
needle. Each sample was then crushed in an agate
mortar. An aliquot of between 1 and 2 mg was
mixed with 2.5 times its weight of a lithium tetra-
borate/lithium carbonate flux prepared according
to the method of Norrish and Hutton (1969). The
mixture was fused to a glass bead in a small gold
crucible by maintaining it at 960? C for five
minutes in a vertical quench furnace. After quench-
ing, the bead was mounted in a lucite disc, polished
and analyzed for nine elements in an A.R.L. EMX
electron microprobe. United States Geological
Survey standards Wl, AGV1, DTS1 and the Allende
bulk standard were prepared in the same way and
used as standards. The homogeneity of the stan-
dards and unknowns was checked by performing
multiple analyses of the standards and three sepa-
rate analyses of each unknown. Reproducibility
of the whole procedure was checked by making six
separate beads of a standard and analyzing each as
an unknown. The relative mean deviation of the
analyses in this case was ?3 percent for elements
present in the undiluted material at concentrations
greater than 0.5 percent by weight. Summation to
100 percent was also used as a test; any analysis that
did not total 100 ? 2 percent was rejected. The
values used for the composition of the USGS stan-
dards were those given by Flanagan (1969), and
those of the Allende bulk standard were given in
Clarke, et al. (1970).
Results
The analyses of 50 chondrules and aggregates
have been used to plot Figures 1-3 and in Table
1 representative analyses are quoted for eight indi-
vidual samples. The mineralogy of most of the
samples was examined by preparing a thin section
from a fragment of each. Analyses 1-3 of Table
1 are of Mg/Fe-rich chondrules. Analyses 4 and 5
are of fine-grained aggregates; these are too fine
for microscopic identification of their constituent
minerals, but x-ray powder diffractrograms show
the presence of clinopyroxene, melilite, and spinel.
NUMBER 14 37
gehlenite)
spinel \ ?>
minor clinopyroxene \
melilite)
clinopyroxene I ?
anorthite \
enstatite
> anorthite
olivine
FIGURE 1.?Ternary plot of SiO2 against (CaO + ALO3) and
(FeO+MgO), all in weight percent, for the chondrules and
aggregates from the Allende meteorite. The line shown is a
possible fractionation trend.
? well crystallized Ca/Al-rich chondrules
? finely crystalline aggregates
^ olivine-rich chondrules
Analyses 6-8 are of coarse-grained chondrules com-
posed of melilite and clinopyroxene with minor
spinel. The composition of a melilite and some
clinopyroxenes from similar chondrules are given
in Clarke, et al. (1970) and in Marvin, et al. (1970).
Figure 1 is a ternary plot of (CaO + AI2O3)
against SiO2 and (MgO + FeO) for the data ob-
tained in this work. FeO was combined with MgO
as most of the analyses showed less than 5 percent
FeO, except for the olivine chondrules in some of
which the FeO content rises to 14 percent. From
the distribution of points it can be seen that there
are two chondrule/aggregate groups. One shows
evidence for a development trend, that is to say
its composition could be regarded as changing
regularly in some manner, but the second shows
no such vector properties in the diagram. The
former group is made up of the Ca/Al-rich bodies
and the latter of chondrules containing olivine
and orthopyroxene and, occasionally, glass. The
group with a trend in bulk chemistry also shows
a corresponding mineralogical development. At
the Ca/Al-rich end, the chondrules are composed
predominantly of gehlenite with a little spinel and
clinopyroxene. Progressing from this extreme to-
wards the (MgO + FeO)-SiO2 join the following
trends are observed:
This is accompanied by a decrease in the TiO2
content and an increase in the Na2O content of
the chondrules. The concentration of K2O is too
low to be meaningful with the analytical method
used. Figure 2 shows the distribution of TiO2 con-
tents. The average concentration of TiO2 in the
melilite plus clinopyroxene chondrules is 1.2 per-
cent, and this decreases on moving from the most
Ca/Al-rich compositions plotted in Figure 1 to-
wards the SiO2-(FeO + MgO) join. The olivine
chondrules show a uniform TiO2 content of around
0.2 percent by weight. The changes in the concen-
tration of Na2O in the same sequence is more com-
plex, as shown in Figure 3. The maximum Na2O
content is observed in those chondrules whose com-
positions plot in the middle of Figure 1; the MgO +
FeO-rich chondrules are almost as low in Na2O as
are the gehlenite-rich bodies. Petrographic exami-
nation of the thin sections confirms the distinction
between the two groups shown in Figure 1. The
group richest in CaO and A12O3 contains soda-free
melilite and clinopyroxene with spinel and occa-
10 15 20
TiO
FIGURE 2.?A histogram of TiO2 concentrations in Allende
chondrules and aggregates. The peak at 1.2 percent TiO2
occurs at the Ca/Al-rich end of the line drawn on Figure 1.
38
FIGURE 3.?A histogram of Na2O concentrations in Allende
chondrules and aggregates. The peak at about 0.2 percent
Na2O is due to the melilite chondrules and that at 0.6 per-
cent Na^O to the olivine chondrules.
sionally anorthite while the other group contains
olivine and orthopyroxene with occasional glass.
The coexistence of olivine and melilite was not
seen nor was that of olivine and clinopyroxene. A
further interesting observation, not expected from
experience with terrestrial samples, is the coexis-
tence of anorthite and melilite. Schairer and Yoder
(1969) reported the coexistence of these minerals
in experimental runs and commented that this had
not, so far, been observed in nature.
Discussion
The distinctive petrographic and chemical differ-
ences between the Ca/Al-rich chondrules and ag-
gregates and the Mg/Si-rich chondrules strongly
suggest that there was very little brecciation and
mixing of chondrule fragments prior to their in-
corporation in this meteorite. It is, therefore, possi-
ble to regard the bulk composition of the Ca/Al-
rich chondrules and aggregates as being that of the
solid phases of a fractionation sequence and so any
trends in their major element chemistry can be
interpreted as reflecting changes in the composi-
tion of their parent material. Changes in the pres-
sure and temperature of the parent material, the
nebular gas of the condensation hypothesis, will
have very little effect upon this relationship, since
dP/dT for the vapor to solid transition for melilite,
enstatite, forsterite, and spinel is almost constant
over a pressure range of 10~7 to 10~2 atmospheres,
and has very nearly the same value for all these
SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
phases (Grossman, 1972). A further point is that
fine-grained aggregates are similar in composition
to the more coarsely crystalline Ca/Al-rich chon-
drules. It is not possible to distinguish between
them in terms of those elements used to construct
Figure 1. This may, in part, be due to the small
number of analyses, data for only six fine-grained
aggregates being currently available, but these all
plot within the field defined by the analyses of the
coarse-grained Ca/Al-rich chondrules. The implica-
tion is that either the conditions favoring the crys-
tallization of material from a similar parent were
not uniform during the formation of the Ca/Al-
rich bodies, or that subsequent recrystallization of
the first-formed microcrystalline aggregates pro-
duced the coarsely crystalline chondrules. Evidence
suggesting that some recrystallization has occurred
comes from the Rb/Sr systematics of the Ca/Al-rich
bodies. They do not define a single isochron (Gray,
et al., 1973), implying that there has been some
movement of Rb and/or Sr since their formation.
It has been shown, however, that the chondrule
olivines show radiation damage whereas the matrix
olivines do not (Green, et al., 1971). Since this
damage is removed by heating to about 500? K
(Green, et al., 1971), the chondrule olivines must
have remained below this temperature during and
after the accretion of the meteorite. Further, the
presence of grossular in some of these aggregates
(Clarke, et al., 1970) and of andradite (Fuchs, 1971)
implies temperatures of the order of 700? K. Gros-
sular has been synthesized at 800? C from a glass
(Yoder, 1954) and at 400? C from a mixture
of 3SiO2-Al2O3-3CaCO3 (Christophe-Michel-Levy,
1956). Thus an upper temperature limit of about
700? K can be given for the postformation history
of the Ca/Al-rich bodies of this meteorite, and of
about 500? K for the Mg/Fe-rich chondrules. These
temperatures are too low to cause the formation
of the coarse-grained chondrules from the fine-
grained aggregates by recrystallization during or
after the accretion of the meteorite.
The changes in the bulk chemistry of the two
chondrule types, Ca/Al-rich and Mg/Fe-rich, are
such that it is not possible at present to connect
them unequivocally into a single fractionation se-
quence with a common parent. It may be that any
sequential change in the Mg/Si ratio of the olivine-
rich chondrules that developed as a result of early
condensation processes has been modified by sub-
NUMBER 14 39
sequent exchange between these chondrules and
the residual material prior to the accretion of the
Allende meteorite. In particular, the magnesium
silicates, enstatite, and forsterite will become more
iron-rich by this process, which moves their com-
position toward the MgO + FeO apex of Figure
1. Plotting this figure on an iron-free basis does
not, however, cause the magnesium-rich chondrules
to plot on an extension of the line drawn; the
distinction between the two groups of chondrules
remains.
The alignment of chondrule compositions in the
Ca/Al-rich portion of Figure 1 suggests that their
bulk compositions were fixed by a fractionation
sequence similar to that derived by Grossman
(1972) from theoretical considerations of the primi-
tive nebular gas. Fuchs and Blander (1973), how-
ever, give a summary of the phenomena that they
have observed in Ca/Al-rich chondrules from the
Allende meteorite that suggests a liquid precursor
for these chondrules, rather than their formation
from a vapor by direct condensation. In particular
they note the spherical nature of the chondrules
and the occurrence of eutectic intergrowths of py-
roxene and anorthite and also gehlenite and anor-
thite. Grossman and Clark (1973) maintain that the
formation of any liquid took place after the main
condensation sequence, from solids produced in this
sequence. This is required by their theory because
the calculated condensation sequence predicts the
formation of some phases, for example clinopy-
roxene and anorthite, at temperatures that are
below their solidi. The direct condensation hy-
pothesis has difficulty in explaining the observed
Mg/Si ratios of the Ca/Al-rich chondrules, which
are almost constant during the decrease in (CaO +
A12O3) content (Figure 1). The direct condensation
hypothesis predicts spinel as an early solid phase,
subsequent condensates being more siliceous. There
should then be a decrease in the Mg/Si ratio of
the succeeding chondrules as condensation proceeds.
This trend is not observed for the chondrule com-
positions plotted in Figure 1; actually there is a
slight increase in this ratio.
The evidence from the Allende Ca/Al-rich chon-
drules suggests that, if they are "primary conden-
sates,'.' they have had a more complicated history
than simply that of condensation and incorpora-
tion into Type III carbonaceous chondrite matrix.
How far this has erased evidence of the primitive
state of these condensates is not clear. The great
heterogeneity of the meteorite is remarkable and it
may be that further samples will provide a better
means for distinguishing between phenomena that
could be produced by the early condensation se-
quence in the solar nebula, and those that are the
result of processes subsequent to this event.
Literature Cited
Cameron, A. G. W.
1968. A New Table of Abundances of the Elements in the
Solar System. Pages 125-143 in L. H. Ahrens, editor,
Origin and Distribution of the Elements. New York:
Pergamon.
Christophe-Michel-Levy, M.
1956. Reproduction artificielle des granats caliques: gros-
sulaire et andradite. Socie'te Francaise de Mineralogie
et de Cristallographie Bulletin, 79:124-128.
Clarke, R. S., Jr., E. Jarosewich, B. Mason, J. Nelen, M.
Gomez, and J. R. Hyde
1970. The Allende, Mexico, Meteorite Shower. Smith-
sonian Contributions to the Earth Sciences, 5:1-52.
Flanagan, F. J.
1969. U. S. Geological Survey Standards-II: First Compila-
tion of Data for the New USGS Rocks. Geochimica
et Cosmochimica Ada, 33:81?120.
Fuchs, L. H.
1969. Occurrence of Cordierite and Aluminous Orthoen-
statite in the Allende Meteorite. American Mineral-
ogist, 54:1645-1653.
1971. Occurrence of Wollastonite, Rhonite and Andradite
in the Allende Meteorite. American Mineralogist,
526:2053-2067.
Fuchs, L. H., and Blander, M.
1973. Calcium-Aluminium Rich Inclusions in the Allende
Meteorite: Textural and Mineralogical Evidence for
a Liquid Origin (abstract) . Transactions of the
American Geophysical Union, 54:345.
Gray, C. A., D. A. Papanastassiou, and G. J. Wasserburg
1973. The Identification of Early Condensates from the
Solar Nebula. Icarus, 20:213-239.
Green, H. W. II, S. V. Radcliffe, and A. H. Heuer
1971. Allende Meteorite: A High Voltage Electron Petro-
graphic Study. Science, 172:936-939.
Grossman, L.
1972. Condensation in the Primitive Solar Nebula. Geo-
chimica et Cosmochimica Ada, 36:597-619.
Grossman, L., and S. P. Clark, Jr.,
1973. High Temperature Condensates in Chondrites and
the Environment in Which They Formed. Geo-
chimica et Cosmochimica Ada, 37:635-650.
Lord, H. C, III
1965. Molecular Equilibria and Condensation in a Solar
Nebula and Cool Stellar Atmospheres. Icarus, 4:
279-288.
40 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
Marvin, U. B., J. A. Wood, and J. S. Dickey, Jr.
1970. Ca-Al-rich Phases in the Allende Meteorite. Earth
and Planetary Science Letters, 7:346-350.
Norrish, K., and J. T. Hutton
1969. An Accurate X-ray Spectrographic Method for the
Analysis of a Wide Range of Geological Samples.
Geochimica et Cosmochimica Ada, 33:431-454.
Schairer, J. F., and H. S. Yoder, Jr.
1969. Critical Planes and Flow Sheet for a Portion of
the System CaO-MgO-ALO3-SiO2 Having Petrological
Significance. Carnegie Institution Year Book, 68:
202-214.
Yoder, H. S., Jr.
1954. Garnets and Staurolite. Carnegie Institution Year
Book, 53:120-121.
The Pulsora Anomaly: A Case against
Metamorphic Equilibration in Chondrites
Kurt Fredriksson, Ananda Dube, Eugene Jarosewich, Joseph A. Nelen,
and Albert F. Noonan
ABSTRACT
Four Indian H-group chondrites, Andura, Butsura,Pulsora, and Sitathali, with almost identical chem-
ical composition but differing in texture and struc-
ture, are described. All have essentially the sameand constant ("equilibrated") olivine and pyroxene
composition. Pulsora, however, contains lithic frag-ments, some of which are coarsely crystalline with
almost no metal, while others contain skeletal oliv-
ine in silica-rich glass. Thus Pulsora could not havebeen "equilibrated" after the final agglomeration,
nor is it necessary to postulate any such "metamor-phic" equilibration to explain the texture and min-
eralogy of other chondrites.
Introduction
Circumstances of fall, morphology, and general
classification of the four Indian H-group chon-
drites, Andura, Butsura, Pulsora, and Sitathali,
have been reported by Murthy, et al. (1969), who
also referred to some previous work by Coulson
(1940). A more detailed description of Sitathali was
published by Viswanathan, et al. (1971). Andura
and Butsura were grouped together by Murthy, et
al., while according to Coulson (using the Rose,
Tschermak, Brezina classification) they were respec-
tively Cck, crystalline spherical, and Ci, intermedi-
ate chondrites. Pulsora and Sitathali were described
Kurt Fredriksson, Eugene Jarosewich, Joseph A. Nelen, and
Albert F. Noonan, Department of Mineral Sciences, National
Museum of Natural History, Smithsonian Institution, Wash-
ington, D. C. 20560. Ananda Dube, Geological Survey of
India, Calcutta 13, India.
by Murthy, et al (1969) as brecciated chondrites
with "fine" and "coarse" matrix respectively and as
Cib, brecciated intermediate, and Chob, brecciated
howarditic chondrites by Coulson (1940). All four
belong to the H-group (high iron) as defined by
Urey and Craig (1953) and Keil and Fredriksson
(1964). According to the subdivision, mainly on de-
gree of "crystallinity," proposed by Van Schmus
and Wood (1967), Butsura is H6 and Sitathali H5.
Andura and Pulsora have not been classified pre-
viously according to this system. However, by
applying the criteria of Van Schmus and Wood,
Andura is definitely within the H5-H6 range. Pul-
sora on the other hand shows a variety of features
allowing it to be placed in any group from H3 to
H6 similarly as the Hedjaz L-group chondrite
(Kraut and Fredriksson, 1971).
These four meteorites were selected for detailed
comparative studies from a number of meteorites
currently being investigated under a joint program
of the Geological Survey of India and the Smith-
sonian Institution, because of their essentially
identical chemical composition in spite of gross
morphological and textural differences as indi-
cated by the studies referred to above as well as our
data.
ACKNOWLEDGMENTS.?We are grateful to the Di-
rector General of the Geological Survey of India,
Dr. M. K. Roy Chowdhury, as well as his prede-
cessors for encouraging this cooperative work over
many years. Mr. R. K. Sundaram and Drs. S. V. P.
Iyengar and M. V. N. Murthy, all of the Geological
Survey of India, also gave much assistance. For
various data reductions we are indebted to the
Smithsonian Information Systems Division and B.
41
42 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
FIGURE 1.?Photomicrograph of thin section of the Andura bronzite-chrondrite (H6). Only
one chondrule, right center, is discernible; its boundaries are indistinct. Part of the matrix
is relatively coarse and equigranular; other parts, however, are still fine-grained. White to gray
are silicates; black, metal and troilite. Length of section 5 mm. Compare with frontispiece (a).
J. Fredriksson. This work has been supported in
part by grants from the Smithsonian Research
Foundation and NASA (NGR 09-015-207).
Mineralogy and Textures
Figures 1-4 and the frontispiece show low magni-
fication photomicrographs of thin sections of the
four chondrites. The distinction between Sitathali
(H5), Andura (H5-6), and Butsura (H6) with re-
gard to grain size, degree of crystallinity and "inte-
gration" between chondrules and matrix (Van
Schmus and Wood, 1967) is clearly open to rather
subjective judgment. Pulsora has parts similar in
texture to the other three, but Figure 5 shows a
relatively coarsely crystalline fragment within Pul-
sora with clear crystals of plagioclase (typical of
Type 6), diopside and minor whitlockite. This
fragment is also nearly devoid of metal and troilite,
like so-called amphoterites, or LL-group chondrites,
but its olivine is typical of the H-group (also nearly
the same as in the main mass of Pulsora, see Tables
1, 5, and 6) and its texture is achondritic (perhaps
a "Grade 7" in the Van Schmus and Wood (1967) 1
NUMBER 14 43
FIGURE 2.?Photomicrograph of thin section of the Butsura bronzite chondrite, classification
H5 (towards H6). Chondrules are somewhat more discernible than in Figure 1, while the
matrix shows some relatively fine-grained areas. Length of section 5 mm.
to 6 petrographic scale). The Pulsora section shown
in Figures 6 and 7, on the other hand, includes
a fragment (chondrule?) with skeletal to euhedral
olivine in clear glass of the kind usually found only
in Type 3 chondrites. The detailed composition
and significance of these fragments are discussed
further below.
Table 1 summarizes electron probe analyses of
the major silicate phases, olivine, low and high
calcium pyroxenes and plagioclase in the four
meteorites. In addition, the modal composition was
determined by point counting by probe analysis of
(Fe, Ca, Mg) on a grid 0.3 X 0.3 mm. The calcu-
lated weight percentages of these four silicates are
also indicated in Table 1. The complete modal
analysis is reported in Table 3 and compared to
the norm. The technique and the computer pro-
gram used is described in the Appendix.
The phase compositions in Table 1 are averages
from analyses performed at various times with
different instruments and we estimate the accuracy
to be at the best ?2 percent relative but usually
better than ?6 percent.
Olivine: The iron value, which is the most re-
liable, is no more accurate than ?2 percent of the
values given (this is about the best which can be
44 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
FIGURE 3.?Photomicrograph of thin section of the Sitathali bronzite chondrite (H5). The
structure is similar to that shown in Figure 2. Chondrules, e.g., center and top left, are
somewhat more abundant and distinct. Length of section 5 mm.
achieved by using a close standard and the Bence
and Albee (1968) correction factors); this makes the
olivines in all four chondrites essentially identical.
The precision, however, while measuring olivine
in a single sample is higher (?0.1% FeO), and
relative differences in olivine composition in Pul-
sora are discussed below. Manganese, magnesium,
and silicon were also measured but with less ac-
curacy.
Pyroxenes: A calcium-poor orthopyroxene is the
most abundant in all four meteorites and has es-
sentially constant composition within each stone
but small and probably real differences between
the meteorites. Like the olivines they are relatively
coarse, well defined and distinguishable in the
modal analysis. The less abundant calcium-rich
pyroxenes show greater chemical variations (es-
pecially in Pulsora), are fine-grained, and occur as
inclusions in the other minerals (lamellae in Ca-
poor pyroxenes) or intergrown with plagioclase in
devitrified areas. Like the plagioclase they are,
therefore, hard to distinguish during modal analy-
sis (Appendix) and the average composition given
in Table 1 is rather uncertain.
Plagioclase: This phase is almost impossible to
identify microscopically (except in a fragment in
NUMBER 14 45
FIGURE 4.?Photomicrograph of thin section of the Pulsora bronzite chondrite. The classification
for this section is H4 to H5, but compare also Figures 5 to 8, and frontispiece (b). Chondrules
and lithic fragments, e.g., right center, are clearly visible in the mostly fine-grained, gray to
opaque matrix. Large, coherent black areas are metal, troilite, and other opaque minerals.
Length of section 5 mm.
Pulsora) and almost invariably contains pyroxene
crystallites. The reported composition is based on
the calcium content obtained in areas where iron
and magnesium values approach minimum (^
background). In the modal analyses allowance was
made for substantial apparent contents of mag-
nesium and, especially, iron.
Metal, both taenite and kamacite, troilite and
whitlockite were identified in our modal analyses
and although numerous grains were analysed and
counted detailed descriptions are not warranted be-
cause they are of average composition and abun-
dance. Also, the absolute abundance of these phases
can be estimated more accurately from the bulk
chemical analyses. A number of common accessory
minerals, e.g., chromite and ilmenite, as well as
more exotic ones, like copper and spinels, were also
identified (cf. Ramdohr, 1973) but these phases
seem to contribute little to the understanding of
current models for the origin of chondrites, es-
pecially ours involving impact-melting and breccia-
tion.
46 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
TABLE I.?Abundance and average composition of major silicate phases
Constituent
AN DURA
Olivine
Pyroxene 1
Pyroxene 2
Plagioclase
BUTSURA
Olivine
Pyroxene 1
Pyroxene 2
Plagioclase
PULSORA
Olivine
Pyroxene 1
Pyroxene 2
Plagioclase
SlTATHALI
Olivine
Pyroxene 1
Pyroxene 2
Plagioclase
Mode
wt.% FeO CaO MgO SiOt AltO3 NatO K2O MnO
29.0 18.4 42.4 39.6 -
27.2 11.3 0.8 31.4 56 M).l 0.2
6.0 5 21 18.0 55 0.8 0.5
6.1 - 2.4 - 65.6 21.5 10.4 (0.7)
25.5 18.3 - 42.4 39.4 -
27.8 11.4 0.8 31.1 56.5 -
7.8 5 21 17 56 -
4.9 - 2.6 - 66 21.7 10.3 (0.7)
22.6 18.3 - 42 39.7 -
26.5 11.0 1.3 30 55 0.3 0.1
11.3 5 19.6 17.5 54.5 0.7 0.6
3.0 - 2.5 - 65.4 21.8 10.4 1.0
28.7 18.0 - 42.3 39.6 -
24.6 11.2 0.6 31.2 56.5 -
6.7 4.1 21.4 17.6 56
5.3 - 2.5 - 65.8 21.6 10.3 (0.7)
^0.2
^0.2
Chemical Composition
The analyses reported in Table 2 are almost
identical. Only Andura shows less metallic Fe and
more FeO. Apparently some of its metal (and/or
troilite) has been more susceptible to atmospheric
oxidation. This assumption is strongly supported by
the fact that the calculated normative fayalite con-
tent of the olivine, Fa23 (Table 3), is considerably
higher than the measured value, Fa19 (Table 1),
which is the same as in the three other stones.
Andura also has a more "rusty" appearance both
macroscopically and in thin section. Pulsora and
Butsura seem to have higher carbon contents than
the other two but the possibility of contamination
cannot be entirely disregarded. (The analysis of
Sitathali reported by Viswanathan, et al. (1971) is
apparently erroneous, especially with regard to the
iron content as the authors suspected.)
In Table 3 the normative and modal composi-
tions are compared. The mode was obtained by
electron probe analysis as described above and in
the Appendix. The composition of the normative
minerals agrees closely with the values obtained
TABLE 2.?Chemical composition (analysts: J. Nelen,
Andura; and E. Jarosewich, Butsura, Pulsora,
Sitathali)
Constituent
Fe
Ni
Co .
FeS
SiO2
TiO2 ....
A12O3
Cr2O3
FeO
MnO
MgO
CaO
Na2O
K2O
P2O3
H2O +
H2O-
C . .
Total
Total Fe . ...
Andura
13.72
1.62
0.08
5.57
36.85
0.16
1.85
0.53
12.23
0.39
23.42
1.97
0.91
0.12
0.29
0.19
0.10
0.05
100.05
26.74
Butsura
16.61
1.75
0.07
5.89
36.27
0.12
1.98
0.56
9.65
0.31
23.21
2.00
0.85
0.08
0.31
0.53
0.13
0.15
100.47
27.85
Pulsora
17.23
1.78
0.10
4.99
35.97
0.12
2.26
0.52
10.58
0.29
22.77
1.59
0.84
0.10
0.26
0.00
0.10
0.14
99.64
28.62
Sitathali
16.08
1.79
0.09
5.76
36.65
0.14
2.37
0.43
9.96
0.31
23.45
1.71
0.87
0.10
0.30
0.10
0.06
0.05
100.22
27.48
NUMBER 14 47
TABLE 3.?Normative and modal (weight percent) compositions
Phase
Ni/Fe
FeS . .
Ap
Cr . .
11
Or
Ab
An
Wo
En
Fs
Fo
Fa
C
Residue . . .
Total
Andura
Norm
15.42
5.57
0.97
0.85
0.30
0.72 )
7.70 I 9.0
0.58 \J
2.70 )
17.34 I 26.7
6.71 \}
12.18 \ 40"7
0.05
Mode
17.1
5.8
0.5
_
*
6.1
6.0*
27.2 f
]
29.0
8.4
100.1
Norm
18.43
5.89
0.74
0.90
0.23
0.44 -
7.18
1.36 '
2.73 j
k 33.2 20.56 I
6.30 \
25.90
8.75
0.15
But sura
Mode
16.3
3.6
_
> 9.0 4.9
27.8 t
J
? 34.7 25.5
14.1
100.0
Pulsora
Norm
19.11
4.99
0.60
0.83
0.24
0.56 )
7.07 I 9.7
2.11 \
1.71 )
35.6 20.09 I 28.7
6.86 \
25.47 )
9.62 }
0.14
Mode
17.1
4.0
0.2
_
3.0
26.5 f f
22.6
15.4
100.1
Norn
17.96
5.76
0.71
0.68
0.26
0.25
7.34
2.25
1.79
20.28
6.38
26.52
9.20
0.05
Sitathali
i Mode
16.1
3.9
0.2
_
? 9.8 5.3
28.5 6A'7*, { 31.324.6 f (
J
? 35.7 28.7
14.4
99.9
?Pyroxene 2, see Table 1.
f Pyroxene 1, see Table 1.
by probe analysis (except for the high Fa value in
Andura, explained above). Although in the modal
analysis only two pyroxenes were distinguished, i.e.,
those with <1% CaO (Pyroxene 1) and >1% CaO
(Pyroxene 2), the average composition calculated
according to Table 1 is Wo10En75Fs15 while the
normative pyroxenes (Table 3) are in the range
Wo6_9En72_76Fs18_19. The amount of pyroxene in
the norm, however, is consistently lower than in
the mode while the reverse is true for olivine and
plagioclase. Because the "Residue" in the modal
analyses is substantial, we have attempted to estab-
lish its chemical composition by deducting the ele-
ments in the modal minerals as well as normative
metal, troilite, chromite, ilmenite, and apatite. In
Table 4 these residues are compared with the chem-
ical composition of the bulk silicates obtained by
deducting FeO for ilmenite and chromite and CaO
for apatite from the chemical analyses in Table 2.
It seems clear that the residue is severely deficient
in silica but enriched in alumina and alkalies.
Since a large part of the residue in these fine-
grained rocks must be due to "overlap" between
two or more phases, we attempted to remove first
olivine, requiring least silica, and secondly py-
roxene, having the same composition as in the
"bulk" (but different proportions), from the resi-
due. In all the four cases the end result is substan-
tial excess of alumina and FeO, ^-'l percent and
0.5 to 1.5 percent, respectively, of the total weight.
Alkalies show a similar excess in all cases while
calcium appears to be deficient in all but Andura.
This calcium deficiency is probably not real but is
due to uncertainty in the analysis of the Ca-rich
pyroxenes.
Iron and magnesium, as might be expected, are
about the same in bulk silicates and residue, again
excepting Andura for which the results are dis-
torted because of the oxidation discussed previously.
This supports the probability that most of the fine-
grained "matrix" is indeed "rock flour" similar in
composition to the bulk meteorites. The excess
iron may possibly be explained as admixed ultra-
fine metal and/or troilite. The excess of alumina
and alkalies, however, is more puzzling. Three
possible interpretations are offered: (1) Our modal
and phase analyses are too inaccurate to allow these
kinds of calculations. The consistency of the results,
however, speaks against this, as well as the fact that
if olivine is underestimated and pyroxene overesti-
mated in the mode, which seems most likely, the
discrepancies become worse. (2) A small amount of
an undiscovered discreet phase(s) is present, e.g.,
corundum or spinel for alumina, or some silicate
48 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
TABLE 4.?Chemical composition of bulk silicates from chemical analyses and of unidentified fractions
residue in the modal analyses
Constituent
SiOo .
A12O3
FeO
MgO
CaO .
Na2O
K2O
Andura
Silicate
fraction
(76% of
total)
47.8
2.4
15.4
30.4
2.6
1.2
0.2
Residue
(8.4% of
total)
33 7
5.7
38.1
18.0
0.2
3.3
1.0
Butsura
Silicate
fraction
(73% of
total)
49.2
2.7
12.7
31.5
2.7
1.2
0.1
Residue
(14.1% of
total)
38.0
11.3
14.0
32.0
0.0
4.5
0.7
Pulsora
Silicate
fraction
(73% of
total)
48.7
3.1
13.9
30.9
2.2
1.1
0.1
Residue
(15.4% of
total)
38.0
11.9
21.8
27.0
0.0
3.5
0.6
Sitathali
Silicate
fraction
(74% of
total)
49.0
3.2
12.9
31.3
2.3
1.2
0.1
Residue
(14.4% of
total)
42.0
13.2
15.5
25.4
0.0
3.3
0.6
strongly enriched in alumina and alkalies and pos-
sibly iron. Our studies have not so far revealed
any such exotic minerals though we have searched
for all the ones recently described by Ramdohr
(1973) and Mason (1972). (3) The fine-grained
ground mass does contain a "real matrix" entirely
different from the bulk meteorite. This is in ac-
cordance with suggestions by Fredriksson and Keil
(1964) that some carbonaceous chondrites are mix-
tures of low and high temperature fractions. Similar
mixture models have been pursued in great detail
by the Chicago group led by E. Anders in a number
of papers, e.g., Anders (1964) and Laul, et al. (1973);
(this paper has numerous references), for especially
gas-rich and/or unequilibrated chondrites. A study
of carbon distribution in some chondrites (Fred-
riksson and Nelen, 1969) led to the conclusion that
also ordinary equilibrated chondrites may have
small amounts of a low temperature real matrix
and current work (Fredriksson, et al., in prep.) indi-
cates that this matrix is also enriched in some heavy
volatiles, e.g., Pb, Zn, and possibly also N and F
(Kerridge, pers. comm.). From our present study,
however, we can only conclude that if such a matrix
is present, its bulk chemistry is still uncertain al-
though Al/Si, Na/Si, and probably Fe/Si are higher
than in the bulk silicates of H-group chondrites (as
in some carbonaceous chondrites).
The Pulsora Anomaly
It has been demonstrated that Pulsora is chem-
ically and mineralogically almost identical to three
other ordinary equilibrated chondrites although
its petrographic grade (Van Schmus and Wood,
1967), even on cursory inspection, seems somewhat
lower, i.e., 4-5, than the others of 5-6. Detailed
studies of a number of thin sections, however, have
revealed several unusual lithic fragments that are
not conspicuous in hand specimens. Two of these,
extremely different from each other and from the
bulk, are described in some detail because they have
profound implications as to the thermal history of
Pulsora and by analogy of other H-group chon-
drites. The achondritic fragment (Figure 5) referred
to previously consists of relatively coarse, equi-
FIGURE 5.?Crystalline "achondritic" (H7?) fragment, lower
right quarter, in Pulsora. Diopside (D), plagioclase (P) and
whitlockite occur as large clear crystals together with normal
olivine and orthopyroxene. Note the scarcity of metal and
troilite (black) as compared to the main mass. Length of
section 4 mm.
NUMBER 14 49
TABLE 5.?Phase compositions in Pulsora
achondritic fragments
Constituent
SiO2
TiO2
A12O3
FeO
MgO
CaO
Na2O . ...
K2O
Olivine Pyroxene 1
-
18.0 10.8
40.5 28.8
2.2
Pyroxene
-
1.6
5.2
17.9
19.0
2 Plagioclase
1.6
granular (in strong contrast with the bulk) olivine,
ortho-pyroxene, diopside, plagioclase and apatite,
in order of decreasing abundance. The composition
of the major silicate phases are given in Tables 5
and 6. Although the accuracy of our iron determi-
nations is no better than ?2 percent relative, we
believe that we obtained a precision better than ?
0.1 percent FeO in alternate measurements between
fragment and matrix and thus established a small
difference in olivine composition, i.e., 18.0 (Table
5) in the fragment and 18.3 (Table 1) in the bulk.
The ortho-pyroxene in the fragment (Pyroxene 1)
has decidedly higher (2.2%, Table 5) CaO content
than in the bulk (1.3%, Table 1) indicating a higher
temperature of "equilibration," whereas the plagio-
clase is somewhat less calcic, An5_7 versus An10, in
the bulk. In the fragment, troilite and metal are
scarce (^0.5%), the latter consisting of about equal
amounts of taenite (r~45% Ni) and kamacite (^
5% Ni). In contrast the main mass contains 17
percent metal with only minor amounts of taenite
with ^45 percent Ni; its kamacite has ^7 percent
Ni. The overall Ni content of the Pulsora metal
(Table 2) is 9 percent corresponding to ^95%
kamacite and ^5% taenite.
The glassy fragment (Figures 6, 7) consists of
euhedral and skeletal crystals of olivine (^65%)
displaying normal zoning, in clear glass ('?'35%)
enriched in silica; the composition is given in Table
6. The olivine has somewhat lower average iron
content than olivine in the bulk and in other lithic
fragments. The variation indicated in Table 6 is
mostly due to zoning in the individual crystals, i.e.,
the cores have almost constant composition with
10-12 percent FeO. The glass is enriched in Si, Al,
Ca, and Na and depleted in Mg and Fe. In its over-
all composition this fragment resembles some
glassy chrondrules described from unequilibrated
chondrites (e.g., Fredriksson and Reid, 1965) al-
though the iron content of the olivine is unusually
high and the glass has > 5 percent normative
quartz, also unusual in glassy chondrules. In addi-
tion it contains a small, 0.5 mm, metal-sulfide
spherule (Figure 8) with ^65 percent kamacite
(7.3% Ni), ^35 percent troilite and a eutectic tex-
ture. The troilite contains 0.9 percent Ni in sharp
contrast to troilite in the main mass of Pulsora, as
well as in other ordinary chondrules, which have
< 0.05 percent Ni. The texture indicates rapid
crystallization and the 0.9 percent Ni content of
the troilite would indicate quenching from ^1000?
C (Kullerud, 1963) without any appreciable reheat-
ing.
This fragment might be a deformed chondrule
even if applying the restricted definition proposed
by Fredriksson, et al. (1973), but the irregular shape
and the fact that other similar but angular frag-
ments were found lead us to believe that it is a
TABLE 6.?Phase compositions in Pulsora glassy
fragments
Constituent
SiO2
TiO2
A12O3
FeO
MgO
CaO
Na2O
K2O
Qtz .
C ..
Or. .
Ab. .
An. .
Wo. .
En. .
Fs. . .
Fo...
Fa ..
II. ...
Olivine Glass Calculated bulk*
41-39
-
-
10-15
49-44
-
-
-
ANALYSES
65
0.4
7.8
11.0
6.5
5.7
2.5
0.3
NORMS
5.5
0.3
2.3
30.6
40.8
-
8.7
11.2
48.5
0.1
2.6
12.2
33.7
1.9
0.8
0.1
_
-
0.8
11.7
6.3
1.9
21.2
5.6
0.5
40.5
11.8
0.2
* Modal analysis gave ^V% olivine and ^14 glass of the
composition indicated in the two preceding columns. Com-
pare bulk silicates in Table 4.
50 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
FIGURE 6.?Thin section of Pulsora including a lithic fragment consisting of olivine in a
silica-rich glass (outlined by black, dotted line). Compare Figure 7. The black sphere consists
of metal and troilite; see Figure 8. Length of section 11 mm.
fragment of a pre-existing rock. In Table 6 we also
present the calculated bulk composition of the sili-
cate in this glassy fragment. It is remarkably similar
to the composition of the Pulsora bulk silicates as
given in Table 4, again indicating a common parent
material.
Clearly, the glassy fragments have not been equil-
ibrated within themselves or with the rest of Pulsora
either before or after incorporation in the mete-
orite. The achondritic fragment might have been
equilibrated or recrystallized and drained of metal
before incorporation in the bulk which then, if
metamorphic equilibration is invoked, must have
been metamorphosed somewhere else, then broken
up and reassembled together with its fragments,
most of which have identical olivine and pyroxene
composition. Pulsora, or at least its main mass,
must have been "metamorphosed" to exactly the
same degree as the higher grade Andura, Butsura,
and Sitathali, since their chemistry and mineralogy
are essentially identical, implying recrystallization
at various temperatures and/or length of time with-
out any differentiation. This seems incomprehensi-
ble and to explain the chemical similarities between
our four chondrites as well as their textural differ-
ences it seems necessary to have (1) a common
parent material, (2) heating and cooling rapid
enough to retain olivine in silica-rich glass and to
NUMBER 14 51
FIGURE 7.?Detail of Figure 6. Skeletal to euhedral zoned
olivine (Fai3, center to Fa10, margin) in clear, quartz-normative
glass. The bulk composition equals that of the silicates in
Pulsora. Length of section 0.9 mm.
prevent chemical differentiation (except for drain-
ing of metal), and (3) repetitious brecciation and/
or recrystallization processes which again do not
cause gross chemical changes and consequently no
equilibration.
FIGURE 8.?Metal (kamacite, 7% Ni) and troilite spherule in
the glassy fragment shown in Figures 6 and 7. The eutectic
texture and high (0.9%) Ni content in the troilite indicate
quenching from about 1000? C. Diameter 0.5 mm.
Conclusion
We have demonstrated that the Pulsora chon-
drite, in spite of its constant olivine-pyroxene com-
position, has never been exposed to thermal
metamorphism after agglomeration as has been sug-
gested for most chondrites (e.g., Merrill, 1921; Van
Schmus and Wood, 1967; Dodd, 1969). Still, Pul-
sora, including its greatly variable (H3 to H7) lithic
fragments, is practically identical, both chemically
and with regard to major phase (and possible ma-
trix) composition, to Andura, Butsura, and Sitath-
ali. These latter three stones may seem to be of a
higher average petrographic grade but they do
contain, as might be perceived from Figures 1-4,
more fine-grained parts similar to the main mass of
Pulsora (while Pulsora has high grade fragments).
Thus, by analogy, we contend that there is no com-
pelling evidence for the assumption that the three
other chondrites here described, or any other ordi-
nary chondrite, ever suffered severe, long term ther-
mal metamorphism, or that the constant olivine-
pyroxene compositions were established by such
processes. Rather, it seems that rapid, multiple
processes acting on a common (at least chemically)
parent material may create conditions under which
olivine-pyroxene may retain their constant initial
composition once established by subliquidus, or
even subsolidus, crystallization (Fredriksson, 1963;
Kurat, 1967; Blander, et al., 1970). Impacts on mod-
erate sized bodies, > 102 to ^ 103 km, seem to be
the most plausible and currently most accepted ex-
planation for the contradictions presented by or-
dinary chondrites. Recent investigations of lunar
impact breccias, glasses, and chondrules and the
interpretation thereof also seem to corroborate the
impact origin of chondrites (Fredriksson, et al.,
1973).
Appendix
MODAL ANALYSIS WITH THE ELECTRON PROBE
Like most chondrites the four stones investigated
in this study are fine grained and the minerals are
intricately intergrown. Thus modal analysis by
point counting under the microscope is practically
impossible as illustrated in Figures 1-4 and espe-
ciallly in the frontispiece. The proportions of metal,
troilite (some oxides) and silicates can be readily
estimated (e.g., Keil, 1962) but even estimates of
52
TABLE 7.?Identification of phases by number of
total counts
Phase
1 Kamacite . . .
2 Taenite ....
3 Troilite ....
4 Olivine ....
5 Pyroxene-1 .
6 Pyroxene-2 .
7 Plagioclase .
8 Apatite ....
9 "Others"
or no fit
FeKa
>
21000
15000
15800
2500
1500
500
20
20
-
<
99000
21000
17500
5000
3000
3000
500
400
-
CaKa
>
50
200
100
50
150
600
500
10000
-
<
500
500
200
150
600
10000
1500
25000
-
MgKa
>
20
20
20
4500
1500
1300
20
20
-
<
500
500
500
10000
5000
5000
300
1000
-
Note: Taenite is distinguished from troilite by the
count, i.e., the background, the difference (somewhat uncer-
tain) being caused by the different average atomic number
(this adds another variable) . In some cases "Taenite 2" was
substituted for apatite by changing the criteria for Phase 8
to Fe, >10000and <15000; Ca. (i.e., background) >200 and
<500; and Mg >20 and <500.
the ratio of olivine to pyroxene in Andura (rela-
tively coarse) differed by more than 50 percent
between two of us and the amount of residue (un-
identified) varied from 18 to 30 percent. In a study
of Forest Vale (Noonan, et al., 1972) more rigorous
optical criteria were applied which resulted in bet-
ter olivine/pyroxene ratios but left /?' 45 percent
of the points (i.e., > 50% of the silicates) unidenti-
fied. We, therefore, decided to resort to point count-
ing by electron probe, simultaneously registering
the relative abundance of three elements. Similar
techniques have been used before (e.g., Keil, 1962,
1967), but we believe that the procedure used for
this study warrants a brief description for three
reasons: (1) to illustrate the accuracy (or lack) of
some of our calculations, (2) some of the modifica-
tions appear to be new and easily adapted to other
systems that include computerized data reduction,
and (3) the inherent possibility to identify ultrafine
interstitial or intergrown minerals, especially if
relative abundances of six or more elements are
determined simultaneously, as is possible with vari-
ous modern instruments.
In our study we used an ARL EMX instrument
run at 20 kV, and ,- 0.25/xA beam (-? 0.025^A
sample) current. The three spectrometers were
SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
tuned to Fek?, Caka and Mgk? and intensities re-
corded in silicate standards were approximately 125
cps (counts per second) per 1 percent Fe, 220 cps/1
percent Ca and 115 cps/1 percent Mg. Counting
time was 2 seconds per point and the sample
was moved mechanically (manually) under the
beam on a 0.3 X 0.3 mm grid during the print out
(typewriter and punch cards) time of <~? 3 seconds;
thus each point required an average of r~> 6 sec-
onds and a typical thin section ^ 1 cm2 was
analyzed within I1/2 hours. No real concentrations
were estimated. Rather, phases were identified
by number of total counts as established for each
run as illustrated in Table 7. Similarly any
other phases can be introduced and of course any
suitable elements can be substituted for Fe, Ca,
Mg to characterize these phases. The developed
computer program readily accommodates such
changes, i.e., it accommodates eight "fixed" phases,
one "unknown" (frequently caused by overlap of
different phases) and one "reject," i.e., if all three
channels register background or less, the point is
not counted but the number of such points is reg-
istered and might in special cases be used to identi-
fy a special "phase," e.g., holes or pores in a sample,
etc. If the counts for each of the three channels are
registered on punch card only, the limits as exem-
plified in the table have to be added, i.e., 48 digit
(up to 107 counts per channel!) numbers which can
be accommodated on 5 IBM cards. (A print out of
the current program in Fortran IV is available on
request.) The program at present is designed to
accomodate up to 750 points, which we consider
more than adequate for most thin sections. An
average run on a Honeywell 1250 takes a maximum
of 2 minutes.
In the near future an electron probe capable of
registering simultaneously nine elements will be
available. This will make it possible to define more
sharply a greater number of phases, but this may
not lead to a much greater resolution of phases in
fine-grained rocks, especially those with a glassy or
crypto-crystalline matrix. It should, however, be
possible to obtain the average composition of such
a fine-grained matrix and/or apportion (i.e., calcu-
late the norm) unidentified points among the major
phases, leaving only small amounts of excess ele-
ments or compounds.
NUMBER 14
Literature Cited
Anders, E.
1964. Origin, Age and Composition of Meteorites. Space
Science Reviews, 3:583-714.
Bence, A. E., and A. L. Albee
1968. Emperical Correction Factors for the Electron Mi-
croanalysis of Silicates and Oxides. Journal of
Geology, 76:382-403.
Blander, M., K. Keil, L. S. Nelson, and S. R. Skaggs
1970. Heating of Basalts with Carbon Dioxide Laser.
Science, 170:435-438.
Coulson, A. L.
1940. Catalogue of Indian Meteorites. Memoirs of the
Geological Survey of India, 75:1-346.
Dodd, R. T.
1969. Metamorphism of the Ordinary Chondrites: A Re-
view. Geochimica et Cosmochimica Ada, 33:161-203.
Fredriksson, K.
1963. Chondrules and the Meteorite Parent Bodies. Trans-
actions of the New York Academy of Sciences, series
2, 25:756-769.
Fredriksson, K., and K. Keil
1964. The Iron, Magnesium, Calcium and Nickel Distri-
bution in the Murray Carbonaceous Chondrite.
Meteoritics, 2:201-217.
Fredriksson, K., and J. Nelen
1969. Carbon Distribution in Chondrites. Meteoritics,
4:271-272.
Fredriksson, K., A. Noonan, and J. Nelen
1973. Meteoritic, Lunar and Lonar Impact Chondrules.
The Moon, 7:475-482.
Fredriksson, K., and A. M. Reid
1965. A Chondrule in the Chainpur Meteorite. Science,
149:856-860.
Keil, K.
1962. Quantitative-erzmikroskopische Integrationsanalyse
der Chondrite. Chemie der Erde, 22:281-348.
1967. The Electron Microprobe X-ray Analyzer and Its
Application in Mineralogy. Fortschrittee der Miner-
alogie, 44:4-66.
Keil, K., and K. Fredrikkson
1964. The Iron, Magnesium and Calcium Distribution in
Coexisting Olivines and Rhombic Pyroxenes in
Chondrites. Journal of Geophysical Research, 69:
3487-3515.
Kraut, F., and K. Fredriksson
1971. Hedjaz an L-3, L-4, L-5 and L-6 Chondrite. Mete-
oritics, 6:284.
Kullerud, G.
1963. The Fe-Ni-S System. Pages 175-189 in Annual
Report of the Director of the Geophysical Labora-
tory, 1962-1963. Carnegie Institution of Washington.
[Also as number 1412 in Papers from the Geophysi-
cal Laboratory.]
Kurat, Gero
1967. Einige Chondren aus dem Meteoriten von Mezo-
Madaras. Geochimica et Cosmochimica Ada, 31:
1843-1857.
Laul J., C. R. Ganapathy, Edward Anders, and John W.
Morgan
1973. Chemical Fractionation in Meteorites-VI: Accretion
Temperature of H-, LL-, and E-Chondrites, from
Abundance of Volatile Trace Elements, Geochimica
et Cosmochimica Ada, 37:329-357.
Mason, B.
1972. The Mineralogy of Meteorites. Meteoritics, 7:309-
326.
Merrill, G. P.
1921. On Metamorphism in Meteorites. Bulletin of the
Geological Society of America, 32:395-414.
Murthy, M. V. N., S. N. P. Srivastara, and A. Dube
1969. Indian Meteorites. Memoirs of the Geological Survey
of India, 99:1-192.
Noonan, A. F., K. Fredriksson, and J. Nelen
1972. Forest Vale Meteorite. Smithsonian Contributions to
the Earth Sciences, 9:57-64.
Ramdohr, P.
1973. The Opaque Minerals in Stony Meteorites. 245
pages. Amsterdam: Elsevier Publishing Company.
Urey, H. C, and H. Craig
1953. The Classification of the Stone Meteorites and the
Origin of the Meteorites. Geochimica et Cosmochi-
mica Ada, 4:36-82.
Van Schmus, W. R., and J. A. Wood
1967. A Chemical and Petrologic Classification for the
Chondritic Meteorites. Geochimica et Cosmochimica
Ada, 31:747-765.
Viswanathan, T. V., N. R. Sen Gupta, D. R. Das Gupta, and
S. Banerjee
1971. The Sitathali Meteorite. Mineralogical Magazine,
38 (September):335-343.
Impact Survival Conditions for Very Large Meteorites,
with Special Reference to the
Legendary Chinguetti Meteorite
Robert F. Fudali and Dean R. Chapman
ABSTRACT
Contrary to popular belief, very large meteorites(>> 1000 tons) can be sufficiently slowed by aero-
dynamic drag to survive impact with the earth'ssurface provided that they enter the atmosphere
at very low angles. This is a stringent requirement
and survival probabilities for large, unguided ob-jects are low; but they are not zero. Based on high-
velocity impact experiments and published tabula-tions of the parameters of shallow angle entry
trajectories, we estimate the probability of survival
for an iron meteorite approximately the size andshape of the legendary Chinguetti meteorite (100
meters x 40 meters x 20-40 meters) to be between0.1 and 1 percent. Together with a limiting esti-
mate of the flux of such bodies encountering the
earth, this leads to an expected survival rate of oneper 10
8?10? years on the earth's land surface.
Introduction
In 1916, a small (4.5 kg) meteorite was found in
the Adrar region of Mauritania (then French West
Africa) by a captain in the French army, Gaston
Ripert. It was designated the Chinguetti Meteorite,
after the nearby oasis of Chinguetti (20? 28'N, 12?
20'W). The meteorite presently resides in the
Museum National d'Histoire Naturelle (Paris).
The fascinating part of the story is that Ripert
Robert F. Fudali, Department of Mineral Sciences, National
Museum of Natural History, Smithsonian Institution, Wash-
ington, D. C. 20560. Dean R. Chapman, Thermo- and Gas-
Dynamics Division, NASA Ames Research Center, Moffett
Field, California 94035.
claimed the meteorite was but the tiniest part of a
gigantic mass that he was led to, with the greatest
reluctance, by the then chieftain of Chinguetti, Sidi
Ahmed Ould Zein. Following is Ripert's description
of the find (as originally transcribed by Lacroix,
1924, and cited in Monrod, 1952):
This specimen was collected about 45 km. to the southwest
of Chinguetti and to the west of Aouinet N'Cher. It was
lying on top of an enormous metallic mass measuring about
100 m. on one side and about 40 m. in height, which stands
up in the middle of dunes covered by a desert plant called
'sba' [sic]. (The mass) is in the form of a compact, unfrac-
tured parallelepiped. The visible surface is vertical, standing
like a cliff above the sand, which, driven by the wind, has
grooved its base to such an extent that its upper edge over-
hangs; the part worn by eolian erosion is polished like a
mirror. The sand has piled up against the opposite face and
hides it entirely, a fact that made it impossible to estimate
its third dimension. The upper surface of the mass bristles
with little needles, which the Arabs have tried to remove;
but, because of the malleability of the metal, these were only
bent. Some less important blocks (of the same material) are
scattered about the neighborhood.
Understandably, this report caused considerable ex-
citement in 1924. If the meteorite existed it would
be of enormous mass. Based on dynamical consid-
erations to fix the unreported third dimension we
estimate a minimum mass of 500,000 tons and a
more likely mass in excess of 1,000,000 tons. Since
it would have had to tumble and skid to rest from
a high velocity, the most stable attitude was prob-
ably achieved during the final stages of deceleration
and this would be with the smallest dimension
vertical. In the event that the vertical dimension is
instead the intermediate dimension it is highly un-
likely that the unknown horizontal dimension could
55
56 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
be significantly smaller than the vertical dimension.
We therefore assume that the unknown dimension
would be at least 20 meters and more likely in ex-
cess of 40 meters.
In addition to the scientific interest, the meteorite
would be large enough to generate thoughts of
commercial gain. Over the next 14 years there were
several, more or less random, attempts made to re-
locate the original site, without success. Curiously,
it was not until 1932 that Ripert was contacted for
help in these endeavors and, by then, he apparently
retained only a vague notion of the general area of
the site. Further, his guide had long since died. In
reviewing the problem, Monod (1952) reproduced
parts of two letters written by Ripert at about this
time. In 1932 Ripert (translated by Brady) wrote,
in part, to Dr. Jean Bosler, Director of Le Verrier
Observatory in Marseille:
Unfortunately, 16 years have passed since I had occasion
to describe the existence of this meteorite. The times were
such that everything else was blotted out by the dramatic
events that were occurring in France, and I could not realize
at the time all the interest that this meteorite might arouse.
I spoke of it incidentally, when I was leaving Mauritania, to
one of my friends, M. H. Hubert, a Doctor of Science at
Dakar, who did not seem to attach any great importance to
its existence.
Moreover, in order to satisfy the urgent demands of my
guide I had taken with me neither compass nor any material
that would permit me to make notes or any measurements
whatsoever; also, I could remain only a very short time, be-
cause of the guide's haste to leave the spot; so my observa-
tions were extremely cursory. I was able to state that the
meteorite formed a sort of cliff, with numerous projecting
ridges of sand, lying in an approximately southwest direction,
while the northeast side was completely covered by sand. I
found there a small fragment, well rounded at the edges,
which I turned over to M. Hubert. On one corner (of the
large mass), toward the west, I think, sand had carved out
fairly thick metallic needles, which I could not loosen or re-
move. Of the metallic nature of the large mass there could
be no doubt at all. Examination of the small specimen given
to M. Hubert proves this fact. I tried to detach one of the
aforesaid needles by pounding it with the small heavy lump
(= the small meteorite). The specimen showed at every
point of contact traces of hammer marks, the same condition
being true of the needles; and, finally, the surface appearance
of the large mass was in no way comparable to that of the
blackish polished surface of the rocks that are found on the
'reg' and on the sandstone plateaus of the Adrar. On my
return, I wrote down my observations (based on my memory,
which was very good at the time), but, since then, I have
moved so much from place to place that I do not know
where I put those notes. In any case, I told M. Hubert every-
thing that I had seen which was then quite clear in my mind.
As you may imagine, this meteorite, half-covered by sand in
1916, is very likely completely covered now, but, even so, I
am astonished that it has not been found by the civilian and
military expeditions that have been sent out since 1929 to
look for it. It seems likely that they have been blocked by a
conspiracy on the part of the natives, who seem anything but
anxious that its existence should be known to Europeans. I
realized this (fact) when I overheard a conversation among
the camel drivers when I was with the Adrar camel corps. I
had spoken of (the meteorite) to the head man of Chin-
guetti, Sidi Ahmed Zein, who at first vigorously denied its
existence, and it was only after long conversations (negotia-
tions?) that he consented, in spite of his fears, to lead me
secretly to the spot, on condition that I would take nothing
with me that might permit me to describe the exact position
of the meteorite, or to make notes. Sidi Ahmed died, ap-
parently of poison; and, without wishing to establish any
connection between the facts, I cannot now point out anyone
who could serve as a guide.
And in 1934 he wrote, in part, to Monod (1952;
translated by Brady):
I know that the general opinion is that the stone (rock)
does not exist: that, to some, I am purely and simply an im-
poster who picked up a metallic specimen on the 'reg' of
Mauritania; that, to others, I am a simpleton who mistook
an exposure of sandstone, blackened and polished as is so
often the case in that country, for an enormous meteorite. I
shall do nothing to undeceive them. I care very little for
what any of them may think. I know only what I saw, be-
cause I saw it, and I could probably have found (the mass)
again if I had been directed to do so, when my visual mem-
ory and my age still permitted.
Monod in 1934, and one A. Pourquie, in 1938,
conducted search expeditions into the Chinguetti
region?both with negative results (Monod, 1952).
At that point World War II intervened and the
matter was not seriously broached again until the
early 1950s. In his 1952 paper Monod was still con-
vinced the huge meteorite could very well exist.
However, based on his extensive experience in the
region, he cautioned:
Let me add that if the meteorite is actually located among
the dunes of Ouarane, 10 hours to the southeast' of Chin-
guetti, the very nature of this terrain is unsuited to easy
exploration. It is foolish, of course, to pursue a search (for
the meteorite) in a random fashion merely in the hope of
making a lucky discovery.
Thus far, this point has been academic as there
have been no further efforts to relocate the site.
The reason for this neglect is that it is widely be-
lieved to be physically impossible for such an enor-
1 The original report of Lacroix describing the find site as
southwest of Chinguetti was subsequently corrected by Ripert
to southeast of Chinguetti (Monod, 1952).
NUMBER 14 57
mous mass to survive an impact with the earth's
surface. This belief is based on simple aerodynamic
calculations, which have been generally interpreted
as precluding any significant atmospheric braking
if the incoming meteoritic mass exceeds a maximum
of 100 to 1000 tons. This has been uniformly as-
serted in several meteoritic texts that have enjoyed
wide circulation (e.g., Mason 1962; Hawkins, 1964;
Heidi, 1964; Wood, 1968).
With little or no atmospheric braking a meteorite
impacts the earth's surface with enough kinetic
energy to destroy both itself and a significant por-
tion of the target rock. And there surely is ample
evidence?in the form of a number of meteorite
impact craters of modest dimensions?that mete-
orites of mass far less than that estimated for Chin-
guetti do normally explode upon encountering the
earth's surface. But what is not generally appreci-
ated is that this is a normal, rather than an inevi-
table, result. In this paper we demonstrate that,
under special conditions, a Chinguetti-size mete-
orite could survive a terrestrial impact essentially
intact. In the following sections we delineate the
conditions necessary for the survival of such a
meteorite, and we compute the approximate prob-
ability of encountering such conditions. As one
might suspect, the conditions are stringent and the
concomittant probabilities are small?but they are
certainly not zero.
TABLE 1.?Fate of aluminum projectiles impacting
sand targets at high velocities and low incident
angles (NASA Ames Light-Gas Gun Range experi-
ments)
Round
DLG
209
211
221
215
233
VGP
788
791
792
796
(km/sec)
6.1
6.5
5.7
6.3
5.9
1.9
1.8
1.9
1.8
Ti Yr Fate of projectile
(degrees)
2.1 0.9 Essentially intact
4.75 1.1 Two main fragments
7.33 1.15 Three main fragments
15 ? Fragmented into small pieces
30 - Fragmented into small pieces
2.1 1.5 Intact
4.75 3.6 Intact
7.33 2.85 Intact
15 1.32 Essentially intact
1/16
1/4
1/2
0.8
1.1
1.2
TABLE 2.?Variation of the effective strength of
aluminum projectiles as a function of their diameter
(NASA Ames Light-Gas Gun Range experiments)
Projectile Impact velocity for Corresponding energy
diameter same degree of deformation for same degree of
(inch) (km jsec) deformation (ergs)
2xlO7
2xlO?
2xlO10
The foregoing should in no sense be considered
an endorsement for the existence of the legendary
Chinguetti meteorite. The primary purposes of this
paper is not to demonstrate that it does exist, but
that such a large meteorite can exist. Clearly this is
the absolute minimal requirement necessary if a
systematic search for the meteorite is ever seriously
contemplated.
ACKNOWLEDGMENTS.?Donald Gault and John
Wedekind of the NASA Ames Research Center
generously furnished us with the unpublished, ex-
perimental high-velocity impact data vital to our
study. For this, and for a number of stimulating
discussions on all aspects of the study, we express
our deep appreciation. We also thank Lionel Levy
of the NASA Ames Research Center for the com-
puter-generated trajectories of a large iron mete-
orite subsequent to ricochet at the earth's surface.
Experimental Data on Impact Survival
There is, of course, no possibility of slowing a
Chinguetti-size meteorite to terminal velocity by
atmospheric braking alone. In the following section
we show that the minimum possible velocity such a
body could have upon impact would be 3-4 km/sec,
and higher velocities are much more probable. We
must therefore consider the possibility of survival
only for impact velocities of 3 km/sec and higher.
Some of the hypervelocity impact experiments
done at the light-gas gun range of NASA Ames Re-
search Center have a direct bearing on this prob-
lem. Most of these impact experiments have been
characterized by high angles of impact (> 30?).
But the few experiments carried out at low incident
angles show that, for aluminum projectiles impact-
ing sand targets at velocities between approximately
2 and 6 km/sec, the fate of the projectile is a func-
58 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
0 246
Impact Angle I? degrees
FIGURE 1.?Relationship between impact angle r\, impact
velocity Vlmp, and rebound angle IY Aluminum projectiles
into sand targets.
tion of both velocity and incident angle. For a
given velocity, as the incident angle is lowered, a
value is reached below which the projectile will
ricochet intact. The pertinent experimental data
are shown in Table 1. Figure 1 plots the angle of
ricochet against the angle of incidence for three
impact velocities. Note that in all cases the ricochet
angle is significantly less than the incident angle.
We have used the data in Table 1 to approxi-
mately fix the boundary separating the fragmenta-
tion domain from the domain of intact ricochet on
plots of impact velocities versus incident angles in
Figures 2 and 3. Admittedly the data are scanty and
14
12 -
10
6 -
- 4
2 -
Intact
- Ricochet
Domain
Atm. Brakinc
Point
.003 \
Fragmentation
Domain (Al into
\
\
~ .006 <y?-^
.006
, 1.1.1,1
sand)
11.2 13
i
/ /'?5
/.03
/.02 /
;L
/
i
16
/.05
/ i
/.04 /
03 /
, 1 ,
20 Km/Sec./
/
1,1,
2 4 6 8 10 12 14 16
Earth Impact Velocity Km/sec
18 20
there is a rather large uncertainty associated with
this boundary placement, especially at higher veloci-
ties. However, it seems likely that the boundary, as
drawn, is conservative. For a Chinguetti-size,
nickel-iron meteorite the true boundary should lie
somewhere to the right of the boundary shown in
Figures 2 and 3. There are two reasons for this:
(1) Nickel-iron has a much greater strength than
aluminum, and (2) scaling up from the tiny ex-
perimental projectiles to a large meteorite should be
favorable for impact survival. Although the shock
wave strengths are independent of scale, the decel-
eration or body forces acting on each element of the
projectile vary inversely with the scale of the pro-
jectile. We therefore expect a large projectile to
? 10
?? 4
20 Km/Sec
FIGURE 2.?Possible earth impact conditions for shallow
angle entry trajectories of an iron meteorite with 7T=2Om.
0 2 4 6 8 10 12 14 16 18
Earth Impact Velocity Km/sec
FIGURE 3.?Possible earth impact conditions for shallow
angle entry trajectories of an iron meteorite with h = 40m.
effectively behave as if it were stronger than a small
projectile. Some experimental confirmation of this
expectation is shown in Table 2.
The conditions a projectile experiences subse-
quent to ricochet are specifically discussed in the
following section. Here we merely point out that
nothing subsequent to the initial impact can begin
to approach this initial event in potential destruc-
tiveness. This is because the projectile loses energy
during ricochet, and is subject to further atmo-
spheric braking after ricochet.
Possible Impact Conditions
Some years ago, Chapman (1959, 1960) published
an analytical method for studying the motion of
NUMBER 14 59
extraterrestrial bodies on shallow entry trajectories
into the earth's atmosphere. So that these trajectory
calculations could be generally applied to any plan-
etary atmosphere and to incoming bodies with any
physical properties, the trajectories were described
in terms of two dimensionless functions. The conic
perigee parameter F p delineates the conditions at
the perigee point of the hypothetical conic trajec-
tory that the incoming body would have had if
there were no planetary atmosphere. It characterizes
the trajectory prior to atmospheric encounter and is
defined as
F = 2(MjCDA) ? (1)
The trajectory through the atmosphere is character-
ized by a dependent variable termed the Z function:
pU
)X2(M/CDA)
where, for both equations p = air density, r=radius
from the planet's center, /3 = the atmospheric density
decay parameter, M = mass (grams) of the incoming
body, A? cross-sectional area normal to the flight
path (cm2), CD = the atmospheric drag coefficient,
w = the ratio of the horizontal velocity component
at any point along the flight path to the circular
orbital velocity (8 km/sec), and subscript v refers to
conditions at the conic perigee.
For an incoming body of given Fp, initial velocity,
and lift force to drag force ratio (LjD), Z can be
determined at all points on its atmospheric tra-
jectory. If Z is known, a number of other pertinent
entry motion parameters can be calculated, includ-
ing the horizontal velocity component and the
trajectory angle with respect to the local horizontal.
This has been done for a large number of trajec-
tories and the results are assembled in computer-
generated tables (Chapman and Kapphahn, 1961).
One such table is reproduced herein for illustrative
purposes (Table 3).
In order to use these tables for the problem under
consideration, we begin by determining Zju at sea
level (i.e., at impact) for Chinguetti-size meteorites,
using equation 2. This quantity is conveniently
given if we use the density of air at set level, p0, for
the calculation. For purposes of these calculations
the following assumptions and approximations are
made: (1) The meteorite is oriented by aerody-
namic forces such that the minimum dimension is
TABLE 3.?Atmospheric behavior of an object enter-
ing the earth's atmosphere under the following
initial conditions: Initial velocity?IS km/sec, Lift/
drag ? 0, Conic perigee parameter 7^ = 0.466, and
Initial atm. encounter angle? ?5.00? (modified
from Chapman and Kapphahn, 1961:155)
Z/fi
0.54
0.59
1.12
il
1.593
1.590
1.585
1.580
1.575
1.570
1.560
1.550
1.500
1.450
1.400
1.350
1.300
1.250
1.200
1.150
1.100
1.050
1.000
.950
.900
.850
.800
.750
.700
.650
.600
.550
.500
.450
.400
.350
.300
.250
.200
.150
.100
.050
Z
.00261
.00986
.02070
.03074
.04026
.04938
.06667
.08297
.15425
.21384
.26505
.30958
.34848
.38253
.41234
.43842
.46123
.48116
.49864
.51405
.52781
.54031
.55202
.56340
.57496
.58727
.60091
.61654
.63485
.65657
.68242
.71309
.74914
.79078
.83749
.88673
.92946
.92114
G
.13
.47
.99
1.46
1.91
2.33
3.13
3.87
6.96
9.32
11.15
12.55
13.61
14.36
14.86
15.14
15.23
15.17
14.97
14.66
14.26
13.79
13.26
12.69
12.09
11.47
10.84
10.20
9.55
8.90
8.24
7.55
6.82
6.04
5.18
4.21
3.13
2.00
r
-5.00
-4.38
-3.99
-3.77
-3.60
-3.48
-3.28
-3.13
-2.67
-2.38
-2.18
-2.03
-1.90
-1.80
-1.72
-1.66
-1.61
-1.59
-1.58
-1.59
-1.62
-1.67
-1.76
-1.87
-2.02
-2.22
-2.47
-2.79
-3.19
-3.69
-4.34
-5.17
-6.27
-7.76
-9.91
-13.26
-19.33
-33.68
S
.000
.018
.029
.036
.041
.044
.050
.055
.069
.078
.085
.091
.096
.100
.105
.108
.112
.116
.119
.122
.126
.129
.132
.135
.138
.141
.143
.146
.149
.151
.154
.156
.158
.161
.163
.165
.166
.168
T
0
9.2
14.9
18.3
20.7
22.6
25.6
27.9
35.5
40.5
44.5
47.9
51.0
53.9
56.7
59.4
62.0
64.7
67.4
70.1
72.9
75.8
78.8
81.9
85.2
88.6
92.3
96.1
100.2
104.6
109.4
114.5
120.2
126.5
133.8
142.6
154.2
172.1
u = Horizontal velocity component normalized to circular
satellite velocity.
Z= Dimensionless function proportional to density times
velocity.
G = Deceleration in earth g's.
r = Flight path angle relative to local horizontal in degrees.
S = Circumferential distance traveled.
T = Time elapsed in seconds.
60 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
parallel to the flight path. The calculations are
carried out for two minimum dimensions, 20 meters
and 40 meters. (2) Density of the nickel-iron mete-
orite is 8 gm/cc. (3) The meteorite has no aerody-
namic lift, either positive or negative (i.e., ballistic
entry). The lift-to-drag ratio (L/D) is therefore
zero. (4) The drag coefficient CD^\.b over the en-
tire flight path. (5) The variation of atmospheric
density with height is exponential and \A/? ?^ 30
everywhere in the earth's atmosphere. (6) r?? radius
of the earth = 6 X 108 cm. (7) For shallow angle
entries, the (normalized) horizontal velocity com-
ponent, u, approximately equals the meteorite's
entire velocity.
Now MIA ? Sh gm/cm2, where h is the average
depth of the meteorites (in cm) in the direction
parallel to the flight path. Therefore M/CVA =
5M gm/cm2. Also Po = 0.0012 gm/cm3. Substitut-
ing in equation (2) we have:
_ 0.0012 gm/cm3 X (6 X 10 8 cm)
~ 10.6 X 30 xTgm/cm2
_ 2264 dimensionless,
for Ji = 40 m = 4 X 10 3 cm, (Z/u)0 = 0.57,
for h = 20 m = 2 X 10 3 cm, (Z/u)0 = 1.14.
(3)
(4)
Table 3 displays the trajectory parameters of a
body with zero lift, entering the earth's atmosphere
at an initial velocity of 13 km/sec and an initial
angle of ?5.00? (the negative sign merely denoting
that the body is descending). From this table we see
that a meteorite characterized by (Z/u)o = 0.57
would impact the earth's surface at a velocity of
7.4 km/sec and an angle of ?1.65?, 75 seconds after
first encountering the atmosphere. If (Z/u)0= 1.14,
the meteorite would impact at 4.4 km/sec and
? 2.8?, 96 seconds after first encountering the atmo-
sphere. It might at first seem paradoxical that the
impact angle can be less than the initial entry angle
since the trajectory must always steepen (in an ab-
solute sense) due to the combined effects of gravity
and aerodynamic deceleration. However, this may
indeed be the case for an impact angle measured
zvith respect to the local horizontal since the earth's
surface curves away from the trajectory path. Both
sets of parameters define points within the intact
ricochet domain as delineated in Figures 2 and 3.
Using the complete tables of Chapman and Kap-
TABLE 4.?Computer-generated trajectory param-
eters and conditions at second impact for an iron
meteorite of h = 40 meters ricocheting with veloci-
ties Vj and angle Tl
(km I sec)
5.3
4.5
4.5
5.0
6.0
6.0
7.0
7.0
8.0
8.0
(degrees)
1.3
1.9
1.5
1.0
0.5
0.8
0.5
0.7
0.4
0.7
Range
(km)
161
153
125
114
106
152
172
211
233
306
Max. height
(km)
1.1
1.4
0.9
0.6
0.3
0.6
0.5
0.9
0.7
1.5
(km /sec)
3.9
3.4
3.5
4.0
4.8
4.4
4.9
4.6
5.0
4.5
(degrees)
-1.8
-2.4
-1.9
-1.3
-0.7
-1.1
-0.8
-1.2
-0.9
-1.6
phahn (1961), a number of such points can be
plotted, generating the family of lines shown in
Figures 2 and 3. Each line contains all the impact
velocities and angles (of interest here) for a single
initial encounter velocity. Lines corresponding to
encounter velocities of 11.2, 13, 16, and 20 km/sec
are plotted. An encounter velocity of 11.2 km/sec
(escape velocity) is, of course, the lower limit for
an incoming meteorite and we consider velocities
between 13 and 16 km/sec to be the most likely
range for large iron meteorites. Clearly, at least for
all encounter velocities less than 20 km/sec, a sig-
nificant portion of the generated lines lie in the
intact ricochet domain.
For all impact velocities below about 8 km/sec,
the associated impact angle is less than ^5? for a
meteorite of ft = 20 m and less than,?-4? for a meteo-
rite of h = 40 m. Combining these data, with those
shown in Figure 1, we see that the only possible
rebound angles are less than 3? and most are less
than 2?. We have used these rebound angles to
compute the after-ricochet trajectories of a mete-
orite of ft = 40 m, and the conditions at second
impact. The computer-generated results are shown
in Table 4. In all cases there is a significant reduc-
tion in velocity at second impact and, in every case
but one, the second impact angle is less than 2?. The
velocities at second impact would actually be some-
what less than those shown since we did not con-
sider any energy (and velocity) loss due to the
initial impact and ricochet. Atmospheric decelera-
NUMBER 14 61
tion after rebound is most effective for the higher
velocity projectiles, so there is not a great deal of
difference between the velocities and angles of the
various cases at second impact?velocities varying
from 3.4 to 5.0 km/sec and impact angles from
? 0.6? to ?2.4?. At these lower velocities and very
low incident angles we feel that such a meteorite
would most likely tumble and skid to rest without
ricocheting a second time. Since aluminum pro-
jectiles vertically impacting sand targets at 2.5 km/
sec remain essentially intact (although severely de-
formed), we expect a nickel-iron meteorite to sur-
vive such a skid to rest essentially intact.
Probabilities of Impact Survival
If certain pre-atmospheric trajectories will permit
the impact survival of a Chinguetti-size meteorite,
then there must be an envelope or entry corridor
containing these trajectories. An unguided, extra-
terrestrial body on a collison course with the earth
may or may not encounter this corridor. To a first
approximation, the probability of a "favorable"
encounter is merely the ratio of the cross-sectional
area of the annular ring defining the entry corridor
to the cross-sectional area of the earth itself:
= log1 z) X 16 km (7)
2Ar,
irr (6)
where /\r is the width of the annular ring and r is
the earth's radius. To be sure, the equation is not
precise, nor is it strictly the equation of applica-
bility since the earth's gravity field creates an effec-
tive capture Cross-section and an annular entry
corridor both of which are larger than the compar-
able terms in equation (6). A precise derivation,
however, is complex and, considering the approxi-
mations and assumptions already made, the in-
crease in precision would be meaningless. The first
approximation is at worst within a factor of 2 of a
more precise derivation?for a meteorite with the
lowest possible geocentric velocity (11.2 km/sec).
The approximation improves as geocentric velo-
cities increase.
For an exponential atmosphere, /\r between any
two pre-atmospheric trajectories can be determined
from the ratio of their conic perigee parameters
since these parameters vary directly with air density
(equation 1). The atmospheric density changes by
a factor of 10 for every 16 km change in altitude
and the corridor width is simply:
The upper limit to any entry corridor will be a
trajectory resulting in just enough atmospheric
braking to transfer the incoming body to an un-
stable elliptical orbit, which will decay and lead to
atmospheric entry on a subsequent pass. For all
practical purposes this atmospheric entry will be
identical to that for a body with circular satellite
velocity, regardless of the initial encounter velocity.
(Thus all the curves in Figures 2 and 3 degenerate
to the same point). Beyond this limiting trajectory
the incoming body will swing around the earth on
a nonreturn hyperbolic trajectory. The nonreturn
boundaries for zero-lift bodies with initial veloci-
ties of 13 and 16 km/sec are characterized by Fp's
of 0.036 and 0.10, respectively (Chapman, 1960,
fig. 14). The lower corridor boundary can be any
arbitrarily chosen trajectory whatsoever. For ex-
ample, Table 3 describes the atmospheric be-
havior of incoming bodies with initial velocities of
13 km/sec and pre-atmospheric trajectories char-
acterized by Fp = 0.466. The ratio 0.466/0.036 = 12.9
= 101-11, so
Ar=l.ll X 16 = 17.8 km (8)
The concomitant, approximate probability of an
unguided, incoming body (with initial velocity of
13 km/sec) encountering this corridor is
2 X 17.8
6 X 10 3: 0.006. (9)
This probability can be related to survival as fol-
lows: In Figure 3, the 13 km/sec curve encompasses
possible impact velocities and angles for a mete-
orite of ft = 40 m. Table 3 was used to generate one
point on that line?the point labeled .006. For that
class of meteorites encountering the earth, char-
acterized by h=40 m and ^=13 km/sec, 0.6 per-
cent will impact at velocities and angles given by
the segment of the 13 km/sec curve lying between
the .006 point and the atmospheric braking point.
All the other numbered points in Figures 2 and 3
were derived in the same way and have the same
meaning.
Discussion and Conclusions
From the foregoing sections it should be clear
that meteoritic masses very much greater than 1000
62 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
tons can survive a terrestrial impact under certain
conditions. In fact, mass is not strictly the govern-
ing parameter and a meteorite even more massive
than the specific Chinguetti-size bodies considered
herein could also survive a terrestrial impact, pro-
vided the critical parameter?its linear dimension
parallel to the atmospheric flight path?remains
modest. The probability of survival, however, will
decrease with increasing mass because of the need
for a meteorite to assume a favorable flight attitude
in the atmosphere. If a meteorite has a stable flight
attitude, it will be with the minimum dimension
oriented parallel to the flight path. For the present
study we have assumed that if all dimensions are
small enough to permit significant deceleration by
the atmosphere, then the meteorite will become
oriented in its stable attitude prior to maximum
deceleration. This holds for the maximum dimen-
sion considered (100 m) on most shallow-angle entry
trajectories. As the other dimensions increase be-
yond this, the pre-atmospheric orientations that will
lead to stable atmospheric orientation are reduced
and thus the survival probability decreases, al-
though it never goes to zero as long as the min-
imum dimension is fixed.
With respect to the specific question of the possi-
ble existence of the Chinguetti meteorite, we reit-
erate the fact that the curves, boundaries and
probabilities displayed in Figures 2 and 3 are all
approximate. Nevertheless, the data are surely pre-
cise enough for order of magnitude conclusions.
Bearing this in mind we estimate the survival prob-
ability of a Chinguetti-size meteorite (whose geo-
centric velocity lies somewhere between 11.2 and
20 km/sec) to be between 0.1 and 1 percent. Put
another way, such a body has between one chance
in 100 and one chance in 1000 of surviving a terres-
trial impact. Now, based very crudely on the num-
ber of more recent, known and probable terrestrial
impact craters due to meteorites with kinetic en-
ergies similar to the hypothetical Chinguetti body,
the upper limit to the encounter frequency of such
bodies with the earth's land surface cannot exceed
one per 10 6 years (over the past few tens of million
years). This means that the expected survival (and
potentially recoverable) rate can be no higher than
one every 10 8 to 10 9 years. Because the complete
oxidation (and thus destruction) of large iron
meteorites on the earth's surface is accomplished in
a time that is short compared to 100 million years,
we are led to the following conclusions. First, the
Chinguetti meteorite can exist. Second, from a
purely statistical aspect, the odds are heavily against
its existence. Of course, the specific existence or
nonexistence of the Chinguetti meteorite is not
purely a statistical problem, since it involves a re-
ported observation. Thus there are only two possi-
ble probabilities: Either the observer, Gaston Rip
ert, was mistaken (or lied), or his report was correct.
In the former case, the probability that the Chin-
guetti meteorite exists is 0. In the latter case, the
probability is 1. No other probability calculations
are meaningful, nor is it possible to discriminate
between 0 and 1 at this point. The combination of
survival and weathering rates, however, does tell
us, clearly and irrevocably, that if the Chinguetti
meteorite does indeed exist, it must be, now and
forever in the history of mankind, unique.
Literature Cited
Chapman, D. R.
1959. An Approximate Analytical Method for Studying
Entry into Planetary Atmospheres. NASA Technical
Report, R-ll: 44 pages.
1960. An Analysis of the Corridor and Guidance Re-
quirements for Supercircular Entry into Planetary
Atmospheres. NASA Technical Report, R-55: 47
pages.
Chapman, D. R., and A. K. Kapphahn
1961. Tables of Z Functions for Atmospheric Entry
Analysis. NASA Technical Report, R-106: 271 pages.
Hawkins, G. S.
1964. Meteors, Comets and Meteorites. 134 pages. New
York: McGraw Hill.
Heidi, F.
1964. Meteorites. 144 pages. Chicago: The University of
Chicago Press.
Lacroix, A.
1924. On a New Type of Meteoric Iron Found in the
Desert of Adrar in Mauritania. Comptes Rendus,
179:303-313.
Mason, B.
1962. Meteorites. 274 pages. New York: John Wiley and
Sons.
Monod, Th.
1952. Le probleme de la meteorite de Chinguetti. Bulletin
de la Direction des Mines, Gouvernement General
de I'Afrique Occidentale Francaise, 15 (2):405-412.
[Translated by L. F. Brady, "The Problem of the
[French West Africa] Chinguetti Meteorite [CN =
0127,202:]", and published, in 1955, in Meteoritics,
1 (3): 308-314.]
Wood, J. A.
1968. Meteorites and the Origin of Planets. 117 pages.
New York: McGraw Hill.
Preliminary Data on Eight Observed-Fall Chondritic Meteorites
Roy S. Clarke, Jr., Eugene Jarosewich,
and Albert F. Noonan
ABSTRACT
Eight observed-fall meteorites, five olivine-bronzitechondrites and three olivine- hypersthene chondrites,
have been chemically analyzed and assigned a VanSchmus-Wood petrographic classification. The
names, dates of fall, and classification of the mete-
orites are as follows; Ankober, 7 July 1942, H4;Schenectady, 12 April 1968, H4; Dwaleni, 12 Octo-
ber 1970, H4 dark fraction and H6 light fraction;Kiffa, 23 October 1970, H5; Kabo (Gwarzo), 25
April 1971, H4 dark fraction and H5 light fraction;
Tathlith, 5 October 1967, L6; Malakal, 9-15 Au-gust 1970, L5; Wethersfield, 8 April 1971, L6.
Introduction
During the past several years we have obtained a
number of newly fallen chondritic meteorites.
These specimens, or samples from them, have been
made available as promptly as possible to investiga-
tors engaged in time-dependent studies. Distribu-
tion of samples for other purposes has also been
made, depending upon availability of specimen
material. In an effort to support these outside stud-
ies, as well as to fulfill our curatorial responsibili-
ties, we have undertaken preliminary examinations
of each meteorite. Typically, this includes obtain-
ing as much information as possible about the
circumstances of fall and recovery of the meteorite,
performing a bulk chemical analysis, a preliminary
petrographic examination, and an electron micro-
probe examination of the major minerals. Our
Roy S. Clarke, Jr., Eugene Jarosewich, and Albert F. Noonan,
Department of Mineral Sciences, National Museum of Natural
History, Smithsonian Institution, Washington, D.C. 20560.
objective is sound classification, and the data is
circulated promptly to those who have need for it.
This paper summarizes data on eight observed-fall
chondrites.
The meteorites studied are listed by date of fall
in Table 1, together with their geographic loca-
tion, date and time of fall, and the weight of
material recovered. Bulk chemical analyses and cal-
culated normative mineral compositions are given
in Table 2. In Table 2 the meteorites are listed in
order of date of fall within the two major classifica-
tion groups. The first five meteorites, Ankober,
Schenectady, Dwaleni, Kiffa, and Kabo are all
olivine-bronzite chondrites (H-group meteorites).
The last three meteorites, Tathlith, Malakal, and
Wethersfield are olivine-hypersthene chondrites
(L-group meteorites).
RELATED LITERATURE.?A number of papers have
been published giving cosmic-ray exposure ages and
rare gas measurements on meteorites from our list
of eight. References are given here, with the names
of the meteorites studied, for papers that are not
mentioned in our discussion: Bogard, et al. (1973),
Tathlith, Malakal, Dwaleni, Kiffa, Wethersfield,
and Kabo; Cressy (1970), Tathlith; Fireman and
Goebel (1970), Ankober; Fireman and Spannagel
(1971), Dwaleni; Reynolds, et al. (1971), Dwaleni;
Smith and Firemen (1973), Dwaleni, Kabo, Kiffa,
Malakal, and Wethersfield; Tobailem and Lalou
(1972), Kiffa; Heimann, Parekh, and Herr (1974),
Malakal,
ACKNOWLEDGMENTS.?The Smithsonian Institu-
tion's Center for Short-Lived Phenomena has been
responsible for obtaining our specimens of the Ma-
lakal, Dwaleni, Kiffa, and Kabo meteorites. We are
indebted to the Center's director, Robert A. Citron,
and to David Squires for their assistance. Joseph A.
63
64 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
TABLE I.?Recovery information on eight observed-fall chondrites (listed by date of fall)
Name
Ankober, Ethiopia
Tathlith, Saudi Arabia
Schenectady, New York
Malakal, Sudan
Dwaleni, Swaziland
Kiffa, Mauritania . . .
Wethersfield, Connecticut ....
Kabo, Nigeria . . ....
Location
9?32'N, 39?43'E
19?21'N, 43?44'E
(recovery site, both ?3')
42?51'39"N, 73?57'1"W
9?30'N, 31?45'E
27?12'S, 31?19'E
16?35'N, 11?21'W
41?42'N, 72?39'E
11?51'N, 8?13'E
Date
7 Jul 42
5 Oct 67
12 Apr 68
9-15 Aug 70
12 Oct 70
23 Oct 70
8 Apr 71
25 Apr 71
Time
10:00 a.m.
local time
5:15-5:30 a.m.
local time
8:30 p.m.
local time
Not known
10:30 hours
South African time
About 2:55 p.m.
GMT
2-6 a.m.
local time
(0700-1100 hrs. GMT)
About 4:30 p.m.
local time
(15.30 GMT)
Recovered Weight
7 kg
^V2kg
283 g
>2kg
3.2 kg
'Vl 5 kg
350 g
>10kg
Nelen did preliminary electron microprobe exam-
inations on several of these meteorites. A grant from
the National Aeronautics and Space Administra-
tion provided funds for the purchase of the
Ankober and Wethersfield meteorites, and to pay re-
covery expenses for the Kabo meteorite. Dr. Robert
Hutchison, British Museum (Natural History) read
the manuscript and provided us with several help-
ful suggestions.
Meteorite Descriptions
The procedures and techniques used in this study
are standard for this type of investigation. Chemi-
cal analyses were perfomed following the general
procedures described earlier (Jarosewich, 1966).
Normative mineral compositions were calculated
from the chemical data. Olivine and pyroxene com-
positions were determined using the electron mi-
croprobe. Individual mineral grains were selected,
and Fe, Ca, and Mg were determined simultane-
ously. The number of grains analyzed depended
both on the homogeneity of the sample and on
their availability in the sections at hand. Olivine
compositions are reported as mole percent fayalite
(Fa) and pyroxene compositions are mole percent
ferrosilite (Fs). A petrographic classification is given
for each meteorite following the procedure of Van
Schmus and Wood (1967).
In the following sections, brief descriptions of
each of the meteorites are given in the order in
which they are listed in Table 2.
ANKOBER, BOLEDE, ETHIOPIA
The Ankober meteorite fall of 7 July 1942 was
observed and attracted local attention at the time.
We first heard of it more than 10 years later, when
Professor Guglielmo Sensi (1953) of the Pasteur In-
stitue, Addis Ababa, wrote in behalf of the owner,
Ato Assafa Abye. Professor Sensi reported that the
meteorite had been incandescent in the air and
that it had fallen into a field of whitish clay caus-
ing no damage. A more dramatic version of the fall
attributed to Ato Assafa Abye reflects the excited
state of the local inhabitants: "After having
touched the earth, it sank to a depth of about one
meter, giving forth steam. The natives of the region
dug at the ground to remove it, but it was still
burning so much they could not take it immedi-
ately. They therefore left it in place, saying that it
must be a shell fired by the English." In early 1954
we received a specimen of the meteorite (USNM
2609) and its identification was confirmed by E. P.
Henderson. In 1967 Robert Citron, at that time
manager of the Smithsonian Institution Astrophysi-
cal Observing Station in Ethiopia, obtained the
main mass of the meteorite (USNM 3399). The
essentially complete individual stone of 6955 g was
subsequently deposited in the meteorite collection
in Washington.
NUMBER 14 65
TABLE 2.?Analytical data on observed-fall chondrites (values in weight percent unless otherwise indicated;
chemical analyses, Eugene Jarosewich; microprobe Fa-Fs determinations, Albert F. Noonan; classifications
according to Van Schmus and Wood, 1967)
Constituent
Ankober Schenectady Dwaleni Kiffa Kabo Tathlith Malakal Wethersfield
USNM 3399 USNM 5722 USNM 5447 USNM 5490 USNM5611 USNM 5448 USNM 5446 USNM 5596
H4 H4 H4fi6 H5 H4JI5 L6 L5 L6
Fe
Ni
Co
FeS
SiO3
TiO2
Al-A
Cr2O3
FeO
MnO
MgO
CaO
Na2O
K2O
P2O5
H2O+
H2O-
C
Total
Total Fe ....
Olivine
Orthopyroxene
Diopside ....
Albite
Anorthite ....
Orthoclase . . .
Chromite
Ilmenite
Apatite
Nickel-Iron ..
Troilite
Fa
Fs
ANALYSES
13.96*
1.71
0.10
5.29
35.25
0.12
2.32
0.55
13.05*
0.32
22.86
1.72
0.84
0.12
0.33
0.45
0.36
0.10
99.55
27.47
43.6
18.8
3.4
7.1
2.2
0.7
0.8
0.2
0.8
15.8
5.3
19
17
15.01*
1.72
0.09
5.29
36.20
0.12
2.25
0.55
11.77*
0.32
23.00
1.77
0.83
0.10
0.29
0.69
0.23
0.12
100.35
27.52
38.7
23.2
3.8
7.0
2.1
0.6
0.8
0.2
0.7
16.8
5.3
19
17
17.23
1.75
0.09
5.21
36.49
0.12
2.30
0.51
9.52
0.27
23.16
1.72
0.88
0.09
0.22
0.18
0.03
0.04
99.81
28.41
33.9
25.8
4.0
7.4
2.1
0.5
0.8
0.2
0.5
19.1
5.2
17.65
1.77
0.09
5.62
36.14
0.12
2.14
0.52
9.15
0.31
23.26
1.77
0.82
0.10
0.22
0.00
0.10
0.01
99.79
28.33
17.56
1.80
0.09
4.82
36.54
0.12
2.15
0.53
9.60
0.33
23.24
1.72
0.82
0.08
0.28
7.61
1.14
0.06
6.58
39.44
0.14
2.40
0.52
13.84
0.33
24.35
1.88
0.99
0.11
0.21
0.03
0.01
99.72
28.09
NORMS
33.6
25.6
4.4
6.9
1.9
0.6
0.8
0.2
0.5
19.5
5.6
33.1
27.5
3.8
6.9
2.0
0.5
0.8
0.2
0.7
19.4
4.8
0.11
0.03
99.74
22.55
43.3
23.6
5.0
8.4
1.8
0.6
0.8
0.3
0.5
8.8
6.6
OLIVINE-PYROXENE COMPOSITIONS (mole percent)
18 19 L 26
16-19D
17 L
15-18 D
16
19 L
19D
17L
14-20D
21
7.56
1.18
0.05
4.30
40.38
0.14
2.14
0.50
15.32
0.28
24.88
1.85
0.95
0.10
0.17
<0.1
0.03
0.03
99.86
22.30
45.4
24.6
5.5
8.0
1.3
0.6
0.7
0.3
0.4
8.8
4.3
25
21
7.00
1.25
0.06
5.46
39.42
0.12
2.26
0.50
15.48
0.34
24.64
1.82
1.05
0.11
0.18
0.00
0.10
0.01
99.80
22.50
49.0
19.5
5.4
8.9
1.1
0.6
0.7
0.2
0.4
8.3
5.5
25
21
?Metallic Fe values are low and calculated FeO values are high in the Ankober and Schenectady analyses. This is primarily
due to oxidation of the sample prior to analysis. The brown color of the analysis powder and the high total water values are
consistent with this interpretation.
L = light fraction, D = dark fraction.
The stone is blocky in shape and about 90 per-
cent of its surface is covered with fusion crust.
Late-stage spalling of fusion crust was restricted to
edges and corners where the essentially flat sur-
faces of the meteorite intersect. The stone appears
to have tumbled during flight and there is no ob-
66 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
vious anterior surface. One of the main surfaces is
markedly less smooth than the other surfaces. This
suggests a break-up at an early stage during atmo-
spheric passage and the probability that the Anko-
ber fall produced at least two individuals. The
fusion crust is brown in color, probably due to
oxidation by atmospheric moisture. It is assumed
that the stone has been protected from the weather
since shortly after its recovery, but we have no cer-
tain information on this prior to our obtaining the
specimen.
Freshly cut surfaces of Ankober quickly become
brown in color due to limonitic staining. Slices are
compact and hard, due largely to the metal forming
a continuous structural framework. Under low mag-
nification (X 10) cut and broken surfaces are seen
to have a surprisingly open structure. Large num-
bers of submillimeter size cavities are present, into
which chondrules, euhedral crystals, and crystal
and chondrule fragments protrude. The meteorite
has a very uniform texture with no obvious breccia-
tion or veining.
In thin section Ankober is seen to consist of dis-
tinct complete and broken chondrules in a matrix
of fine-grained silicates and metallic phases. Chon-
drules do not normally exceed one millimeter in
diameter, but a few unusually large ones up to 6
mm across were observed. Minerals present include
olivine, bronzite, clinobronzite, taenite, kamacite,
and troilite, with minor amounts of chromite,
whitlockite, and copper. Forty microprobe analyses
indicated a constant silicate composition of Fa 19
for olivine and Fs 17 for the low-Ca pyroxene. Abun-
dant clinobronzite occurs both in chondrules and
matrix, often with devitrified glass. Ankober is clas-
sified as an H4 chondrite.
SCHENECTADY, SCHENECTADY COUNTY, NEW YORK
The Schenectady meteorite fall of 12 April 1968
produced an individual stone that was recovered
following a widely observed fireball. Details of the
fall and recovery have been reported by Fleischer,
et al. (1970). These authors also classified the mete-
orite, presented rare gas data, radioactive-ages and
cosmic-ray track information. The meteorite is the
property of The Schenectady Museum, Schenectady,
New York, and it was loaned to us for study in
August of 1969.
The physical characteristics of Schenectady are
similar to those described for Ankober. Exposure to
moisture from the air and during sawing has re-
sulted in limonitic staining. The fusion crust how-
ever, is still black. The stone is compact and
hard, but under low magnification (X 10) many
small cavities can be seen. The openness of the
structure, however, is less than that observed in
Ankober. The stone has a uniform texture, free of
brecciation or veining.
Chondrules are abundant, distinct, and as large
as two millimeters in diameter. They are seen in
thin section to be in a relatively coarse-grained
matrix. The chondrules are composed of olivine,
orthobronzite, minor clinobronzite, and devitrified
glass. Forty-five microprobe analyses show the prin-
cipal silicate minerals to be essentially homoge-
neous, with an olivine composition of Faa9 and a
pyroxene composition of Fs]7. Other minerals iden-
tified were taenite, kamacite, troilite, chromite,
whitlockite, and copper. The concentration of cop-
per was unusually high with as many as 20 grains
found in one polished section. Schenectady is classi-
fied as an H4 chondrite.
DWALENI, SWAZILAND
Loud explosions accompanied the fall of the
Dawleni meteorite near the Swaziland-Transvaal
border on 12 October 1970. Three essentially com-
pletely fusion-crusted individuals were recovered
promptly from a 12 km 2 area centering on the loca-
tion given in Table 1. The individuals weighing
2.37, 0.51, and 0.35 kg were recovered from 15 to
18 cm below ground level in moist soil. We received
two pieces of the 0.35 kg stone (156 g and 10 g) on
3 November 1970. This sample material and infor-
mation on the fall and recovery were sent to us by
J. G. Urie, Acting Director, Geological Survey and
Mines Department, Mbabane.
The Dwaleni chondrite is a compact rock with a
pronounced light-dark structure (Figure 1). The one
surface available to us shows approximately equal
areas of light and dark material, with thin dark
veins penetrating both types. The light-colored
areas are of uniform composition, have sharp out-
line and are tightly bound to the darker material.
Bogard, et al. (1973:2429) have noted that the dark
fraction of Dwaleni is unusually rich in rare gases.
Both light and dark materials undoubtedly have
similar bulk compositions, but minor differences in
NUMBER 14 67
FIGURE 1.?The Dwaleni, Swaziland, chondrite. A broken
surface showing light-dark structure and veining. One and a
half times actual size.
individual mineral compositions between these two
materials were noted. The lighter material contains
relatively large crystals of homogeneous olivine,
bronzite, and plagioclase. Chondrules are absent or
indistinct. Twenty-five grain analyses gave olivine,
pyroxene, and plagioclase compositions of Fa2o, Fs17
and An12. The petrographic classification of this
material is H6. The dark material, however, con-
tains olivine and pyroxene of slightly higher
magnesium content and of somewhat variable com-
position. Eighteen grains analyses gave olivine and
bronzite compositions of Fa1G_19 and Fs15_18. Chon-
drules up to one millimeter in diameter are well
defined and contain turbid glass. The petrographic
classification of the dark material is H4. Kamacite,
taenite, troilite and minor chromite, and copper
occur in Dwaleni, mostly in the light-colored frac-
tion.
KIFFA, MAURITANIA
At 2:55 p.m. GMT on 23 October 1970, detona-
tions were heard and a smoke cloud observed in the
neighborhood of the town of Kiffa. About 10 a.m.
the next morning a meteorite specimen was found
by a child 8 km southwest of the town. It had made
a shallow hole about 20 cm deep in fine sand. The
meteorite shattered into many pieces on landing
and less than 1.5 kg was recovered. We received a
36 g piece on 17 February 1971 from M. Henri
Gruenwald, Chef de Service Geologique, Direction
Des Mines et de la Geologie, Nouakchott. M.
Gruenwald was also our source of information on
the fall and recovery of the meteorite.
The Kiffa meteorite is a friable gray chondrite
with distinct chondrules reaching a maximum di-
ameter of approximately 2 mm. The chondrules
can be separated with ease from the fine-grained
white matrix, which appears to cement the chon-
drules and metal phases together. The texture is
quite uniform and there are no obvious foreign in-
clusions. Olivine and bronzite are the principal
silicate minerals in both chondrules and matrix.
Clinobronzite is rare. Twenty microprobe analyses
show chondrule olivine and pyroxene to be con-
stant in composition, the fayalite and ferrosilite
value being Fa18 and Fs1(i, respectively. The white
matrix is composed, at least in part, of what appears
to be comminuted chondrules and devitrified glass.
In addition to olivine and bronzite, clinobronzite,
kamacite, taenite, troilite, chromite, and whitlock-
ite were identified. Although chondrules are di-
stinct, Kiffa is classified an H5 chondrite principally
on the basis of silicate homogeneity and the pres-
ence of only minor clinopyroxene.
KABO, GWARZO DISTRICT, KANO STATE, NIGERIA
The name Gwarzo has been used as a synonym
for Kabo.
The Kabo meteorite shower of 25 April 1971
attracted considerable local attention, and at least
four stones with a total weight of over 10 kg were
recovered. We received a 134 g sample on 9 June
1971. Both the sample and a description of the fall
and recovery were provided by Professor M. O.
Oyawoye, Geology Department, University of Iba-
dan. A detailed description of this meteorite has
been prepared by Hutchison and coworkers (Hutch-
ison, et al., 1973).
Kabo is an ordinary chondrite with a subtle
light-dark structure. Bogard, et al. (1973) have
shown that its dark fraction contains an enrichment
of several trapped rare gases. Veining and distorted
metal grains in Kabo seem to be preferentially as-
sociated with the dark fraction. Chondrules, which
68 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
range up to two millimeters in diameter in both
fractions, are well defined in the dark fraction and
less well denned in the light fraction. Clinobronzite
occurs in both the light and dark fractions.
The minerals identified in Kabo are olivine,
bronzite, clinobronzite, kamacite, taenite, troilite,
chromite, ilmenite, whitlockite, chlorapatite, and
copper. Thirty grain analyses in the dark fraction
show pyroxene variability over the composition
range Fs14_20. The olivines, however, are essentially
homogeneous with a composition of Fa19. Glass oc-
curs as fragments and in chondrules in minor
amounts, indicating that the dark fraction is less
equilibrated than the light fraction. Thirty olivine
and pyroxene grain analyses in the light fraction
indicate homogeneous compositions of Fa19 and
Fs17. The textural and compositional differences re-
sult in an H4 classification for the dark fraction and
an H5 classification for the light fraction.
TATHLITH, SAUDI ARABIA
The Tathlith meteorite fell early in the morning
on 5 October 1967 about 15 km from the village of
Tathlith. A single blocky-shaped, fusion-crusted
individual was recovered immediately. The exact
weight of the stone is not known to us, but estimates
based on photographs indicate that it weighed
about 2 kg (Figure 2). We received a 205 g slice of
Tathlith from Glen F. Brown, Chief, U.S. Geologi-
cal Survey Field Party, Jiddah, Saudi Arabia on 22
November 1967. It had been identified as a meteo-
rite by Gerhard W. Leo in the U. S. Geological
Survey Laboratories in Jiddah. The main mass of
the meteorite is believed to be in the possession of
the Emir of Abha.
Tathlith is a highly recrystallized gray chondrite
of uniform texture. Chondrules range in size up to
3 mm in diameter and are easily discernible on a
cut or broken surface. In thin section, however, the
chondrules appear relict and intergrown with gran-
ular plagioclase and metal. Minerals identified are
olivine, hypersthene, clinohypersthene, plagioclase,
kamacite, taenite, troilite, chromite, and copper.
Twenty grain analyses show the major silicates to
be homogeneous with olivine composition of Fa26
and hypersthene of Fs2i- The plagioclase is also
homogeneous with a composition of An10. Small
amounts of clinohypersthene, which are probably
a it^yy^H
'?i-i
FIGURE 2.?The Tathlith, Saudi Arabia, meteorite, a, Cut
surface from which our 205 g specimen was removed, b,
Meteorite sitting on cut surface. Scale in inches. Photographs
by Gerhard W. Leo, U.S. Geological Survey.
remnant, occur in some of the larger chondrules.
Tathlith is an L6 chondrite.
MALAKAL, UPPER NILE PROVINCE, SUDAN
The fall of the Malakal meteorite during the
first half of August 1970 into an active military
area has been described by Dawoud and Vail
(1971). The meteorite fragment that reached Khar-
NUMBER 14 69
toum on 15 August 1970, weighed about 2 kg, and
photographs of that piece indicate that it is a part
of a considerably larger stone. It is not known if
other fragments were recovered. We received a 194 g
piece and 4.4 g of fragments from Professor J. R.
Vail of the University of Khartoum on 3 November
1970. A comprehensive description of the Malakal
meteorite has recently been published by El Rabaa,
etal. (1974).
Malakal is a unique meteorite from the point of
view of its cosmic-ray history. Cressy and Rancitelli
(1974) note that cosmic-ray produced 2(iAl and 58Mn
are anomalously high in Malakal. They conclude
that the meteorite must have experienced a two-
stage irradiation, and that during the earlier stage
Malakal was exposed to a cosmic-ray flux several
times larger than that incident on other chondrites
or that measured by space probes.
Malakal is a moderately veined and brecciated
hypersthene chondrite, with the dark and ap-
parently shock-produced vein material also occur-
ring in patches. Kamacite and troilite also appear
to have been partially shock melted. Their inter-
faces with each other are frequently lacy and
diffuse, as are their interfaces with surrounding
silicates. Abundant spherules of metal or sulfide are
observed within the silicates near the metallic
phases. Chondrules are present, ranging up to 5 mm
in diameter, and the silicate matrix is moderately
recrystallized. The minerals identified are olivine,
hypersthene, kamacite, taenite, troilite, ilmenite,
chromite, and whitlockite. Twenty grain analyses
show the major silicates to be homogeneous with
olivine composition of Fa25 and pyroxene of Fs2i.
Malakal is classified as an L5 chondrite.
WETHERSFIELD,, HARTFORD COUNTY,, CONNECTICUT
The Wethersfield meteorite fell into the home of
Mr. and Mrs. Paul J. Cassarino during the early
morning hours of 8 April 1971. It penetrated a roof
of asbestos shingles and three-quarter-inch plywood,
then passed through four inches of insulating ma-
terial and nearly went through a half-inch sheetrock
ceiling in a second floor room. Debris on the floor
of the room called attention to the presence of a
350 g meteorite precariously suspended in the
broken ceiling. Richard E. McCrosky of the Smith-
sonian Astrophysical Observatory, Cambridge,
visited the Cassarino's and obtained the specimen
FIGURE 3.?The Wethersfield, Connecticut, chondrite. The
bottom left side of the photograph shows posterior fusion
crust and spalled areas of crust. The bottom center area
shows a broken surface and extensive interior veining.
for scientific study and preservation. He also can-
vassed the surrounding area and made numerous
inquiries about the possibility of other pieces hav-
ing been found. Only the one specimen was re-
covered, and no observations were reported that
would allow fixing the time of fall more precisely.
The meteorite as found was largely covered with
a dark fusion crust. Areas where the fusion crust
had spalled late in flight covered 10 to 15 percent
of the surface. The largest single surface covering
nearly half of the specimen, was smooth and
rounded and appeared to have been a leading
surface during atmospheric flight. The other sur-
faces were more irregular and the opposite surface
appears to have been a posterior surface. Its fusion
crust is lighter in color, more vesicular, and it has
small areas of the darker anterior fusion crust de-
posited on it. This area of posterior fusion crust is
shown in the central area of Figure 3.
Wethersfield is a gray, highly recrystallized chon-
drite with black veins and patches of black material
(Figure 3). This black material appears to have
been produced by shock. Relict chondrules range
up to 2 mm in diameter, with a few large ones
approaching 5 mm. Fifteen microprobe grain analy-
ses show the silicates to be homogeneous, with an
olivine composition of Fa25 and hypersthene of
Fs2i. Other minerals identified were kamacite,
taenite, troilite, chromite, and whitlockite. It is
classified as an L6 chondrite.
70 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
Literature Cited
Bogard, D. D., M. A. Reynolds, and L. A. Simms
1973. Noble Gas Concentrations and Cosmic Ray Exposure
Ages of Eight Recently Fallen Chondrites. Geochi-
mica et Cosmochimica Acta, 37:2417-2433.
Cressy, Philip J. Jr.
1970. Multiparameter Analysis of Gamma Radiation from
the Barwell, St. Severin and Tathlith Meteorites.
Geochimica et Cosmochimica Acta, 34:771-779.
Cressy, P. J., Jr., and L. A. Rancitelli
1974. The Unique Cosmic-Ray History of the Malakal
Chondrite. Earth and Planetary Science Letters, 22:
275-283.
Dawoud, A. S., and J. R. Vail
1971. Malakal Meteorite, Sudan. Nature: Physical Science,
229:212-213.
El Rabaa, S. M., A. M. Daminova, M. I. Dyakonova, L. G.
Kvasha, L. K. Levskii, and A. V. Fisenko
1974. The Results of the Investigation of the Malakal
Chondrite. Meteoritika, 33:83-89.
Fireman, E. L., and R. Goebel
1970. Argon 37 and Argon 39 in Recently Fallen Meteor-
ites and Cosmic-Ray Variations. Journal of Geo-
physical Research, 75:2115-2124.
Fireman, Edward L., and Gert Spannagel
1971. Fresh Meteorites in 1970 and the Cosmic-Ray
Gradient. Chemie der Erde, 30:83-101.
Fleischer, R. L., E. Lifshin, P. B. Price, R. T. Woods, R. W.
Carter, and E. L. Fireman
1970. Schenectady Meteorite. Icarus, 12:402-406.
Heimann, M., P. P. Parekh, and W. Herr
1974. . A Comparative Study on ^Al and MMn in Eighteen
Chondrites. Geochimica et Cosmochimica Acta, 38:
217-234.
Hutchison, R., D. E. Ajakaiye, C. J. Elliott, and F. A. Fry
1973. The Kabo, Nigeria, Meteorite Fall. Mineralogical
Magazine, 39:340-345.
Jarosewich, E.
1966. Chemical Analyses of Ten Stony Meteorites. Geo-
chimica et Cosmochimica Acta, 30:1261-1265.
Reynolds, M. A., D. D. Bogard, and C. M. Polo
1971. Dwaleni?A New Gas-Rich Chondrite (Abstract).
E.O.S., 52:269.
Sensi, Guglielmo
1953. [Letter from Professor Sensi to the Smithsonian
Institution.] Accession file 202751, Office of the
Registrar, Smithsonian Institution.
Smith, S., and E. L. Fireman
1973. Ages of Eight Recently Fallen Meteorites. Journal of
Geophysical Research, 78:3249-3259.
Tobailem, Jacques, and Claude Lalou
1972. Age d'exposition de la meteorite Kiffa. Comptes
Rendus des Seances de UAcadimie des Sciences,
Series B, 274 (21): 1185-1187.
Van Schmus, W. R., and J. A. Wood
1967. A Chemical-Petrologic Classification for the Chon-
dritic Meteorites. Geochimica et Cosmochimica Acta,
31:747-765.
List of Meteorites in the National Museum of Natural History,
Smithsonian Institution
Compiled by Brian Mason
ABSTRACT
A brief history of the meteorite collection in the
National Museum of Natural History is followedby a list with data of locality, type, and weight of
each specimen.
Introduction
Meteorites formed part of the original collections
of the Smithsonian Institution. James Smithson's
bequest, besides the endowment fund, included
(Goode, 1897:305) "a cabinet, which . . . proves to
consist of a choice and beautiful collection of min-
erals, comprising probably eight or ten thousand
specimens. . . . The cabinet also contains a valuable
suite of meteoric stones, which appear to be suites
of most of the important meteorites which have
fallen in Europe during several centuries." In 1854
the then Secretary, Joseph Henry (1854:18), re-
corded: "The laboratory of the Institution during
the past year has been used by Professor J. Lawrence
Smith in the examination of American minerals.
. . . He also made a series of analyses of meteorites,
among which were fourteen specimens belonging to
the cabinet of James Smithson." Unfortunately, the
specific meteorites from the Smithson collection
are not identified, and the meteorite collection was
apparently destroyed, along with the rest of Smith-
son's cabinet, in the 1865 fire.
In 1880 G. P. Merrill, in a manuscript report,
Brian Mason, Department of Mineral Sciences, National
Museum of Natural History, Smithsonian Institution, Wash-
ington, D.C. 20560.
noted only six meteorites in the catalog: Imilac,
Vaca Muerta, New Concord, Parnallee, Searsmont,
and Cold Bokkeveld. The Tucson iron, which
should have formed the chief attraction, was not
cataloged, although brought to Washington in
1863. The Casas Grandes iron was also received
around 1880, being transferred to the Smithsonian
Institution along with much other material from
the Philadelphia Centennial Exhibition. However,
the collection grew rapidly during the next decade,
largely through the efforts of F. W. Clarke, who
was appointed honorary curator in 1883. In 1889
he published a catalog of the meteorite collection
as of October 1888, in which he lists 128 distinct
falls and finds; in addition, there were 217 mete-
orites in the Shephard collection, which was de-
posited in the museum by Shephard's son in 1886
(and ultimately bequeathed to the Smithsonian
Institution in 1915). Considerable duplication ex-
isted between the two collections, and many of the
specimens in both collections were small fragments.
Since that time the meteorite collection has
grown steadily. In 1902, W. Tassin brought the
Clarke catalog up to date; the Shephard meteorites
were listed along with the museum's collection,
and the combined collections contained 348 falls
and finds. The most recent published catalog is
that by G. P. Merrill, "Handbook and Descriptive
Catalogue of the Meteorite Collections in the
United States National Museum." The collection
at that time comprised 412 independent falls and
finds. In the Annual Report of the Smithsonian
Institution, E. P. Henderson (1948) noted that the
collection had grown to 767 meteorites.
The present collection contains representative
material of some 1300 meteorites, or about two-
thirds of all known meteorites. The current catalog
71
72 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
in the Department of Mineral Sciences lists over
5700 individual specimens. This collection ranks in
size and comprehensiveness with that of the British
Museum (Natural History).
ACKNOWLEDGMENTS.?I wish to thank Mr. R. S.
Clarke, Jr., and Mr. A. F. Noonan for much assis-
tance in compiling this list, and Dr. W. R. Van
Schmus for some unpublished data on the classifi-
cation of individual chondrites.
Literature Cited
Clarke, F. W.
1889. The Meteorite Collection of the U.S. National Mu-
seum: A Catalogue of Meteorites Represented No-
vember 1, 1886. Report of the Smithsonian Institu-
tion, 1885-86, part 2:255-265.
Goode, G. B.
1897. The Smithsonian Institution, 1846-1896. Washing-
ton.
Henderson, E. P.
1948. American Meteorites and the National Collection.
Annual Report of the Board of Regents of the
Smithsonian Institution, pages 257-268.
Henry, J.
1854. Report of the Secretary. Ninth Annual Report of
the Board of Regents of the Smithsonian Institu-
tion. 463 pages.
Merrill, George P.
1916. Handbook and Descriptive Catalogue of the Mete-
orite Collections in the United States National
Museum. United States National Museum Bulletin,
94.
Tassin, W.
1902. Descriptive Catalogue of the Meteorite Collection in
the United States National Museum to January 1,
1902. Report of the United States National Mu-
seum for 1900, pages 671-698.
CHONDRITES:
(Type, where
known, is in-
dicated by a
following
digit)
ACHONDRITES:
Enstatite
Bronzite
Hypersthene
Amphoterite
Carbonaceous
Aubrite
Diogenite
Chassignite
Ureilite
Angrite
Nakhlite
Howardite
Eucrite
E
H
L
LL
C
Ae
Ah
Ac
Au
Aa
An
Aho
Aeu
Abbreviations
STONY-IRONS:
IRONS:
Pallasite
Mesosiderite
Siderophyre
Lodranite
HexahedriteOctahedrite
Coarsest octahedrite
Coarse octahedrite
Medium octahedrite
Fine octahedrite
Finest octahedrite
Plessitic octahedrite
Ataxite
Anomalous
P
M
S
Lo
Hx
O
Ogg
og
Om
Of
Off
Opl
D
Anom
List
Weights are given in grams unless otherwise indicated;
"s" indicates that only a thin or polished section is present
Name
Aarhus
Abancay
Abee
Abernathy
Accalana
Achilles
Achiras
Adams County
Adargas
ad-Dahbubah
Adelie Land
Adhi Kot
Admire
Locality
Denmark
Peru
Canada
Texas
Australia
Kansas
Argentina
Colorado
Mexico
Saudi Arabia
Antarctica
Pakistan
Kansas
Class
H
Of
E4
L6
L3
H5
L6
H5
Om
H5
L5
E3
P
Weight
s
9.8
2950
83
14
546
6.8
569
24
40 kg
69
22
22 kg
Name
Adrian
Agen
Aggie Creek
Aguada
Ahumada
Ainsworth
Akpohon
Akron (1961)
Alais
Alamogordo
Alamosa
Albareto
Albin
Locality
Texas
France
Alaska
Argentina
Mexico
Nebraska
Canada
Colorado
France
New Mexico
Colorado
Italy
Wyoming
Class
H4
H5
Om
L6
P
Ogg
Om
L6
Cl
H5
L6
L4
P
Weight
1090
163
984
8.0
834
1635
64
70
0.6
905
1784
1.7
3200
NUMBER 14
Name
Aleppo
Alessandria
Alexander County
al-Ghanim (iron)
al-Ghanim (stone)
Alfianello
Algoma
Alikatnima
Allegan
Allende
Al Rais
Alt Bela
Altonah
Amates
Ambapur Nagla
Amherst No. 1
Amherst No. 2
Anderson
Andover
Andura
Angelica
Angers
Angra dos Reis (stone)
Ankober
Anoka
Anthony
Antofagasta
Apoala
Appley Bridge
Arapahoe
Arcadia
Archie
Arispe
Arlington
Arltunga
Armel
ar-Rakhbah
Arriba No. 1
Arroyo Aguiar
Artracoona
Asheville
Ashfork
Ashmore
ash-Shalfar
Assam
Assisi
as-Su'aydan
Aswan
Atarra
Atemajac
Athens
Atlanta
Atoka
Atwood
Auburn
Augustinovka
Aumale
Aumieres
Aurora
Locality
Syria
Italy
North Carolina
Saudi Arabia
Saudi Arabia
Italy
Wisconsin
Australia
Michigan
Mexico
Saudi Arabia
Czechoslovakia
Utah
Mexico
India
Nebraska
Nebraska
Ohio
Maine
India
Wisconsin
France
Brazil
Ethiopia
Minnesota
Kansas
Chile
Mexico
Great Britain
Colorado
Nebraska
Missouri
Mexico
Minnesota
Australia
Colorado
Saudi Arabia
Colorado
Argentina
Australia
North Carolina
Arizona
Texas
Saudi Arabia
India
Italy
Saudi Arabia
Egypt
India
Mexico
Alabama
Louisiana
Oklahoma
Colorado
Alabama
USSR
Algeria
France
New Mexico
Class
L6
H5
Og
O?
H6
L6
Om
D
H5
C3
C2
Om
Of
Og
H5
L6
L6
P
L6
H6
Om
L6
Aa
H4
Of
H5
P
Om
LL6
L5
LL6
H6
Og
Om
D
L5
O
L5
H5
L6
Om
Og
H5
H5
L5
H5
H5
Om
L4
L6
LL6
E5
L6
L6
Ogg
Om
L6
L6
H4
Weight
162
37
12
500
3755
680
22
32
18 kg
380 kg
132
10
10 kg
14
45
8017
473
48
2791
9.0
331
0.7
8.5
6964
132
8711
16 kg
324
545
422
1676
3760
178 kg
6140
979
301
395
1142
2.9
375
6.2
39
16
935
7
29
5530
11
37
8
206
235
432
161
235
99
81
16
78
Name
Ausson
Avanhandava
Babb's Mill
(Troost's Iron)
Bachmut
Bacubirito
Bahjoi
Bald Mountain
Baldwyn
Balfour Downs
Bali
Ballinger
Ballinoo
Bandong
Bansur
Banswal
Baquedano
Barbotan
Barea
Baroti
Barranca Blanca
Barratta
Bartlett
Barwell
Barwise
Bath
Bath Furnace
Bear Creek
Beardsley
Bear Lodge
Beaver Creek
Beddgelert
Beenham
Bella Rocca
Belle Plaine
Bells
Bellsbank
Benares
Bencubbin
Bendego
Benld
Bennett County
Benton
Bereba
Beuste
Bhola
Bholgati
Bialystok
Bielokrynitschie
Billings
Billygoat Donga
Binda
Bingera
Bir Hadi
Bischtiibe
Bishop Canyon
Bishopville
Bishunpur
Locality
France
Brazil
Tennessee
USSR
Mexico
India
North Carolina
Mississippi
Australia
Cameroon
Texas
Australia
Java
India
India
Chile
France
Spain
India
Chile
Australia
Texas
Great Britain
Texas
South Dakota
Kentucky
Colorado
Kansas
Wyoming
Canada
Great Britain
New Mexico
Mexico
Kansas
Texas
South Africa
India
Australia
Brazil
Illinois
South Dakota
Canada
Upper Volta
France
Bangladesh
India
Poland
USSR
Missouri
Australia
Australia
Australia
Saudi Arabia
USSR
Colorado
South Carolina
India
Class
L5
H5
D
L6
Off
Og
L4
L6
Og
C3
Og
Opl
LL6
L6
L5
Om
H5
M
L6
Anom
L4
Om
L5
H5
H4
L6
Om
H5
Om
H4
H5
L5
Om
L6
C2
Hx
LL6
Anom
Og
H6
Hx
LL6
Aeu
L5
L
Aho
Aho
H4
Om
L6
Aho
Hx
H5
Og
Of
Ae
LL3
73
Weight
223
162
133
9
997
497
222
0.5
911
154
459
1415
56
0.5
5.7
864
354
132
6.6
84
4256
670
1463
290
1086
388
275
926
3265
731
17
4166
666
90
0.5
343
88
7021
1980
6.3
8135
225
15
3
116
1.6
16
56
434
0.6
228
197
525
3358
226
630
42
74 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
Name
Bitburg
Bjelaja Zerkov
Bjurbole
Black Moshannon Park
Black Mountain
Blackwell
Blanket
Blansko
Blithfield
Bluff
Bocas
Bodaibo
Boelus
Boerne
Bogou
Boguslavka
Bolrumilitz
Bolivia
Bondoc Peninsula
Bonita Springs
Boogaldi
Bori
Borkut
Bowden
Bowesmont
Boxhole
Brady
Brahin
Braunau
Breece
Breitscheid
Bremervorde
Brenham
Brewster
Bridgewater
Briggsdale
Briscoe
Bristol
Britstown
Broken Bow
Brownfield (1937)
Brownfield (1964)
Bruderheim
Bruno
Budulan
Bununu
Burdett
Burgavli
Bur-Gheluai
Burkett
Burlington
Burnabbie
Burrika
Buschof
Bushman Land
Bushnell
Bustee
Butler
Butsura
Locality
Germany
USSR
Finland
Pennsylvania
North Carolina
Oklahoma
Texas
Czechoslovakia
Canada
Texas
Mexico
USSR
Nebraska
Texas
Upper Volta
USSR
Czechoslovakia
Bolivia
Philippines
Florida
Australia
India
USSR
South Africa
North Dakota
Australia
Texas
USSR
Czechoslovakia
New Mexico
Germany
Germany
Kansas
Kansas
North Carolina
Colorado
Texas
Tennessee
South Africa
Nebraska
Texas
Texas
Canada
Canada
USSR
Nigeria
Kansas
USSR
Somalia
Texas
New York
Australia
Australia
USSR
South Africa
Nebraska
India
Missouri
India
Class
Of
H6
L4
L
Og
L5
L6
H6
E6
L5
L6
Of
LL6
H6
Og
Hx
Og
Og
M
H5
Of
L6
L5
H5
L6
Om
L6
P
Hx
Om
H5
H3
P
L6
Om
Om
L5
Of
Opl
H4
H3
H5
L6
Hx
M
Aho
H5
Og
H5
Og
Om
H5
L6
L6
Of
H4
Ae
Opl
H6
Weight
168
10
3509
29
18
2362
1790
2.3
129
8585
2
261
675
537
3107
185
325
20 kg
28 kg
34 kg
79
136
8.4
6.5
116
7346
60
20
241
541
s
33
89 kg
986
54
60
260
126
19
283
1182
168
5515
17
97
331
203
130
3513
1173
1325
65
0.6
14
2865
50
78
639
18
Name
Cabezo de Mayo
Cabin Creek
Cacaria
Cachari
Cadell
Calico Rock
Calliham
Cambria
Campbellsville
Campo del Cielo
Camp Verde
Canellas
Cangas de Onis
Canon City
Canton
Canyon City
Canyon Diablo
Cape Girardeau
Cape of Good Hope
Caperr
Cape York
Carbo
Cardanumbi
Carlton
Caroline
Carraweena
Carthage
Cartoonkana
Casas Grandes
Casey County
Cashion
Casilda
Casimiro de Abreu
Castalia
Castine
Cavour
Cedar (Kansas)
Cedar (Texas)
Cedartown
Cee Vee
Central Missouri
Cereseto
Chainpur
Chamberlin
Chandakapur
Channing
Chantonnay
Charcas
Charlotte
Charsonville
Chassigny
Chateau-Renard
Chaves
Chebankol
Cherokee Springs
Chesterville
Chico
Chico Mountains
Chicora
Locality
Spain
Arkansas
Mexico
Argentina
Australia
Arkansas
Texas
New York
Kentucky
Argentina
Arizona
Spain
Spain
Colorado
Georgia
California
Arizona
Missouri
South Africa
Argentina
Greenland
Mexico
Australia
Texas
Australia
Australia
Tennessee
Australia
Mexico
Kentucky
Oklahoma
Argentina
Brazil
North Carolina
Maine
South Dakota
Kansas
Texas
Georgia
Texas
Missouri
Italy
India
Texas
India
Texas
France
Mexico
Tennessee
France
France
France
Portugal
USSR
South Carolina
South Carolina
New Mexico
Texas
Pennsylvania
Class
L6
Om
Om
Aeu
L6
Hx
L6
Of
Om
Og
Og
H5
H5
H5
Om
Om
Og
H6
D
Om
Om
Om
L5
Of
H5
L3
Om
L6
Om
Og
H4
H5
Om
H5
L6
H6
H6
H4
Hx
H5
Ogg
H5
LL3
H5
L5
H5
L6
Om
Of
H6
Ac
L6
Aho
Og
LL6
Hx
L
Hx
LL6
Weight
207
34
165
1650
71
539
230
351
9521
196 kg
541
6.5
1103
55
409
284
1675 kg
4
209
36
8178
24 kg
0.6
1486
13
44
1399
s
1317 kg
92
71
203
212
17
0.2
9285
191
2181
9940
130
5072
65
114
98
32
1530
371
182
182
83
23
727
s
57
7885
103
s
173
272
NUMBER 14
Name
Chihuahua City
Chilkoot
Chinautla
Chinga
Chinguetti
Chulafinee
Chupaderos
Cincinnati
Clareton
Clark County
Clayton ville
Cleveland
Clover Springs
Clovis No. 1
Coahuila
Cobija
Cockarrow Creek
Cockburn
Cocklebiddy
Cocunda
Colby
Cold Bay
Cold Bokkeveld
Coldwater (iron)
Coldwater (stone)
Colfax
Collescipoli
Colomera
Comanche (iron)
Concho
Cookeville
Coolac
Coolamon
Coolidge
Coomandook
Coonana
Coon Butte
Coopertown
Coorara
Cope
Copiapo
Cortez
Cosby's Creek
Costilla Peak
Cotesfield
Covert
Cowra
Coya Norte
Crab Orchard
Cranberry Plains
Cranbourne
Cranganore
Cratheus
Credo
Crescent
Cronstad
Cross Roads
Crumlin
Cruz del Aire
Locality
Mexico
Alaska
Guatemala
USSR
Maurctania
Alabama
Mexico
Ohio
Wyoming
Kentucky
Texas
Tennessee
Arizona
New Mexico
Mexico
Chile
Australia
Australia
Australia
Australia
Wisconsin
Alaska
South Africa
Kansas
Kansas
North Carolina
Italy
Spain
Texas
Texas
Tennessee
Australia
Australia
Kansas
Australia
Australia
Arizona
Tennessee
Australia
Colorado
Chile
Colorado
Tennessee
New Mexico
Nebraska
Kansas
Australia
Chile
Tennessee
Virginia
Australia
India
Brazil
Australia
Oklahoma
South Africa
North Carolina
Ireland
Mexico
Class
Anom
Om
Of
D
M
Om
Om
Hx
L6
Om
L5
Om
M
H3
Hx
H6
L6
L6
H5
L6
L6
P
C2
O
H5
Om
H5
Anom
Og
L6
Og
Og
L6
C4
H6
H4
L6
Og
L6
H5
Anom
H6
Og
Om
L6
H5
Opl
Hx
M
Of
Og
L6
Of
L6
C2
H5
H5
L5
Of
Weight
363
5524
162
1613
409
. 119
2516
32
208
2069
198
819
1041
284 kg
147 kg
202
3.2
1850
119
s
3906
287
37
4768
5472
312
127
133
550
138
459
2067
s
203
11
s
199
1149
7.6
4147
174
54
1375
1849
17
2966
61
4850
2047
7.9
700 kg
10
42
7.4
s
12
6.5
35
351
Name
Cuernavaca
Cubertson
Cullison
Cumberland Falls
Cumpas
Cushing
Cynthiana
Dalgaranga
Dalgety Downs
Dalhart
Dalton
Dandapur
Daniel's Kuil
Davis Mountains
Dayton
Deal
Deelfontein
Deep Springs
Delegate
Del Rio
Densmore
Den ton County
Denver
Deport
Descubridora
Dexter
Dhurmsala
Dimboola
Dimitrovgrad
Dimmitt
Dingo Pup Donga
Dix
Djati-Pengilon
Dokachi
Doroninsk
Dorrigo
Douar Mghila
Doyleville
Drake Creek
Dresden (Canada)
Dresden (Kansas)
Drum Mountains
Duchesne
Duel Hill (1854)
Dungannon
Durala
Duruma
Dwaleni
D wight
Dyarrl Island
Eagle Station
Edjudina
Edmonson
Edmonton (Canada)
Edmonton (Kentucky)
Efremovka
Ehole
Locality
Mexico
Nebraska
Kansas
Kentucky
Mexico
Oklahoma
Kentucky
Australia
Australia
Texas
Georgia
India
South Africa
Texas
Ohio
New Jersey
South Africa
North Carolina
Australia
Texas
Kansas
Texas
Colorado
Texas
Mexico
Texas
India
Australia
Yugoslavia
Texas
Australia
Nebraska
Java
Bangladesh
USSR
Australia
Morocco
Colorado
Tennessee
Canada
Kansas
Utah
Utah
North Carolina
Virginia
India
Kenya
Swaziland
Kansas
New Guinea
Kentucky
Australia
Texas
Canada
Kentucky
USSR
Angola
Class
Om
H4
H4
Ae
Om
H4
L4
M
L5
L5
Om
L6
E6
Om
Off
L6
Og
D
Om
D
L6
Om
L6
Og
Om
Om
LL6
H5
Om
H4
Au
L
H6
H5
H6
Opl
LL6
H5
L6
H6
H5
Om
Of
Of
Og
L6
L6
H6
L6
M
P
H4
L6
Hx
Of
C3
H5
Weight
738
12
2421
8640
2212
519
643
159
125 kg
65
49 kg
57
5
117
24 kg
3
105
318
202
237
42 kg
8
208
4455
6021
212
681
46
219
6285
4.0
8905
464
16
7.5
39
0.5
17
152
27
212
522 kg
2780
138
1009
59
1.5
115
174
8
345
0.8
281
584
8277
269
1528
76 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
Name
Eichstadt
Ekeby
Elba
Elbogen
El Burro
El Capitan
Elenovka
Eli Elwah
Ellemeet
Ellerslie
Elm Creek
El Perdido
Elsinora
Elyria
Emery
Emmaville
Emmitsburg
Enigma
Enon
Ensisheim
Ergheo
Erie
Erxleben
Essebi
Estacado
Estherville
Esu
Eustis
Eva
Faith
Farley
Farmington
Farmville
Farnum
Faucett
Fayetteville
Felix
Fenbark
Filomena
Finmarken
Finney
Fisher
Fleming
Floyd
Follinge
Forest City
Forest Vale
Forksville
Forrest
Forrest Lakes
Forsyth
Forsyth County
Fort Pierre
Four Corners
Franceville
Frankfort (iron)
Frankfort (stone)
Franklin
Locality
Germany
Sweden
Colorado
Czechoslovakia
Mexico
New Mexico
USSR
Australia
Holland
Australia
Kansas
Argentina
Australia
Kansas
South Dakota
Australia
Maryland
Georgia
Ohio
France
Somalia
Colorado
Germany
Zaire
Texas
Iowa
Sudan
Florida
Oklahoma
South Dakota
New Mexico
Kansas
North Carolina
Nebraska
Missouri
Arkansas
Alabama
Australia
Chile
Norway
Texas
Minnesota
Colorado
New Mexico
Sweden
Iowa
Australia
Virginia
Australia
Australia
Georgia
North Carolina
South Dakota
New Mexico
Colorado
Kentucky
Alabama
Kentucky
Class Weight Name
H5 1 Freda
H4 18 Fremont Butte
H5 94 Frenchman Bay
Om 94 Fukutomi
Ogg 34 kg Futtehpur
Om 4831
L5 219 Gambat
L6 55 Garland
Ah s Garnett
L5 28 Garraf
H4 1185 Geidam
H5 39 Georgetown
H5 87 Ghubara
Om 234 Gibeon
M 1002 Gifu
Aeu s Gilgoin
Om 7 Girgenti
H4 94 Giroux
Anom 44 Gladstone (iron)
LL6 225 Gladstone (stone)
L5 1082 Glasgow
L 1275 Glorieta Mountain
H6 31 Gnadenfrei
C2 61 Goalpara
H6 10 kg Gobabeb
M 10 kg Goodland
H s Goose Lake
H4 480 Grady (1933)
L5 300 Grady (1937)
Grand Rapids
H5 434 Grant
H5 930 Grant County
L5 2620 Grassland
H4 5941 Great Bear Lake
L5 150 Greenbrier County
H5 6209 Gressk
H4 161 Gretna
C3 1286 Grosnaja
H5 4.0 Grossliebenthal
Hx 20 kg Gruneberg
P 977 Gruver
L5 277 Guarena
L6 6500 Guffey
H3 68 Guibga
H4 256 Guidder
Off 14 Gun Creek
H5 36 kg Gundaring
H4 147 Gunnadorah
L6 1995 Gutersloh
H5 1.7
LL5 s Haig
L6 9.5 Hainholz
Hx 1121 Hale Center No. 1
Om 99 Hale Center No. 2
Om 7255 Hallingeberg
Om 297 Hamilton
Om 1638 Hamlet
Aho 4.7 Hammond
H5 207 Hammond Downs
Locality Class Weight
North Dakota D 255
Colorado L4 162
Australia H3 82
Japan L5 9.7
India L6 19
Pakistan L6 29
Utah Ah 102
Kansas H4 286
Spain L6 53
Nigeria H5 5.5
Colorado H6 687
Saudi Arabia L5 9
South-West Africa Of 263 kg
Japan L6 18
Australia H5 4285
Italy L6 8939
Canada P 2381
Australia Og 21 kg
New Mexico H6 3250
Kentucky Om 4047
New Mexico Om Hkg
Germany H5 0.8
India Au 163
Southwest Africa H4 25
Kansas L4 174
California Om 1165 kg
New Mexico L 1768
New Mexico H3 772
Michigan Om 3526
New Mexico Om 481 kg
Kansas L6 696
Texas L4 114
Canada H6 40
West Virginia Om 473
USSR Hx 1935
Kansas L5 32 kg
USSR C3 4.7
USSR L6 49
Poland H4 16
Texas H4 1256
Spain H6 416
Colorado D 8200
Upper Volta L5 250
Cameroon LL5 0.4
Arizona Om 1204
Australia Og 237
Australia H5 1.2
Germany H4 22
Australia Om 450
Germany M 382
Texas L5 188
Texas H4 324
Sweden L3 144
Australia L6 201
Indiana LL4 1543
Wisconsin Om 293
Australia H4 15 kg
NUMBER 14
Name
Haraiya
Hardwick
Harleton
Harriman
Harrison County
Harrisonville
Havana
Haven
Havero
Haviland
Hawk Springs
Hedeskoga
Hedjaz
Helt Township
Hentmry
Hendersonville
Hermitage Plains
Hessle
Hex River Mountains
Hildreth
Hill City
Hoba
Hobbs
Hokmark
Holbrook
Holland's Store
Holly
Holyoke
Homestead
Honolulu
Hope
Hopper
Horace
Horse Creek
Hraschina
Huckitta
Hugoton
Huizopa
Hungen
Hvittis
Ibbenbiiren
Ibitira
Ider
Idutywa
Ilimaes (pallasite)
Ilinskaya Stanitza
Imilac
Imperial
Indarch
Indianola
Indian Valley
Indio Rico
Inman
Ipiringa
Iquique
Iredell
Iron Creek
Locality
India
Minnesota
Texas
Tennessee
Indiana
Missouri
Illinois
Kansas
Finland
Kansas
Wyoming
Sweden
Saudi Arabia
Indiana
Australia
North Carolina
Australia
Sweden
South Africa
Nebraska
Kansas
South-West
Africa
New Mexico
Sweden
Arizona
Georgia
Colorado
Colorado
Iowa
Hawaii
Arkansas
Virginia
Kansas
Colorado
Yugoslavia
Australia
Kansas
Mexico
Germany
Finland
Germany
Brazil
Alabama
South Africa
Chile
USSR
Chile
California
USSR
Nebraska
Virginia
Argentina
Kansas
Brazil
Chile
Texas
Canada
Class
Aeu
L4
L6
Om
L6
L6
Of
H6
Au
H5
H5
H5
L
Ogg
Om
L5
L6
H5
Hx
L5
Of
D
H4
L4
L6
Hx
H4
H4
L5
L5
Og
Om
H5
Hx
Om
P
H5
Of
H6
E6
Ah
Aeu
Om
H5
P
Om
P
H4
E4
L5
Hx
H6
L4
H6
D
Ogg
Om
Weight
2.2
1091
4400
12 kg
19
10 kg
12
21
1.8
84
88
26
107
135
227 kg
2695
130
1863
582
173
10 kg
4203
224
12
8454
54
37
222
5616
32
885
25
464
140
1.0
403
258
2790
21
190
9
1.6
90 kg
50
155
184
19 kg
4
2840
225
489
4.7
14
10
168
97
122
Name
Itapicuru-Mirim
Ivanpah
Ivuna
Jackalsfontein
Jackson County
Jajh deh Kot Lalu
Jamestown
Jelica
Jenkins
Jenny's Creek
Jerome
Jhung
Joe Wright Mountain
Johnson City
Johnstown
Jonzac
Judesegeri
Juromenha
Juvinas
Kaalijarv
Kaba
Kabo
Kainsaz
Kakangari
Kalaba
Kaldoonera Hill
Kalkaska x
Kamiomi
Kamsagar
Kandahar
Kansas City
Kapoeta
Kappakoola
Karakol
Karatu
Karee Kloof
Karkh
Karoonda
Kaufman
Kearney
Keen Mountain
Kelly
Kendall County
Kennard
Kenton County
Kerilis
Kermichel
Kernouve'
Kesen
Keyes
Khairpur
Khanpur
Kharkov
Khohar
Khor Temiki
Kiel
Kielpa
Locality
Brazil
California
Tanzania
South Africa
Tennessee
Pakistan
North Dakota
Yugoslavia
Missouri
West Virginia
Kansas
Pakistan
Arkansas
Kansas
Colorado
France
India
Portugal
France
USSR
Hungary
Nigeria
USSR
India
Zaire
Australia
Michigan
Japan
India
Afghanistan
Missouri
Sudan
Australia
USSR
Tanzania
South Africa
Pakistan
Australia
Texas
Nebraska
Virginia
Colorado
Texas
Nebraska
Kentucky
France
France
France
Japan
Oklahoma
Pakistan
India
USSR
India
Sudan
Germany
Australia
Class
H5
Om
Cl
L6
Om
E6
Of
LL6
Og
Og
L4
L5
Om
L6
Ah
Aeu
H6
D
Aeu
Og
C3
H5
C3
C3?
H4
H6
Om
H4
L6
L6
H5
Aho
H6
LL6
LL6
Og
L6
C4
L5
H5
Hx
LL4
Anom
H5
Om
H5
L6
H6
H4
L6
E6
LL5
L6
L3
Ae
L6
H5
Weight
9.7
3115
122
30
46
18
329
231
55 kg
14
166
22
571
1955
1028
0.2
16
16
172
4.3
0.7
80
26
0.5
11
1.4
759
5
9.6
123
283
0.9
0.7
4
57
27
49
123
168
1008
5629
3512
2044
156
2691
2.4
42
1135
2105
1781
27
30
26
92
90
6.2
351
78 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
Name
Kiffa
Killeter
Kimble County
Kingfisher
Kingoonya
Kingston
Kissij
Kittakittaooloo
Klein-Wenden
Klondike
Knowles
Knyahinya
Kodaikanal
Kopjes Vlei
Koraleigh
Krahenberg
Kramer Creek
Krasnoi-Ugol
Krasnoyarsk
Kress
Krymka
Kuleschovka
Kulnine
Kumerina
Kunashak
Kuttippuram
Kyancutta
Kybunga
Kyle
Kyushu
La Becasse
Laborel
La Caille
Ladder Creek
Lafayette
La Grange
L'Aigle
Lakangaon
Lake Bonney
Lake Brown
Lake Grace
Lake Labyrinth
Lake Murray
Laketon
Lakewood
La Lande
Lance
Langon
Landes
Langhalsen
Lanton
La Porte
La Primitiva
Las Vegas
Laundry East
Laundry Rockhole
Laundry West
Laurens County
Locality
Mauritania
Ireland
Texas
Oklahoma
Australia
New Mexico
USSR
Australia
Germany
Canada
Oklahoma
USSR
India
South Africa
Australia
Germany
Colorado
USSR
USSR
Texas
USSR
USSR
Australia
Australia
USSR
India
Australia
Australia
Texas
Japan
France
France
France
Kansas
Indiana
Kentucky
France
India
Australia
Australia
Australia
Australia
Oklahoma
Texas
New Mexico
New Mexico
France
France
West Virginia
Sweden
Missouri
Indiana
Chile
Nevada
Australia
Australia
Australia
South Carolina
Class
H4
H6
H6
L5
L4
Om
H5
H4
H6
D
Om
L5
Off
Hx
L6
LL5
L4
L6
P
L6
LL3
L6
L6
Opl
L6
L6
Om
L5
L6
L6
L6
H5
Om
L6
An
Of
L6
Aeu
L6
L6
L6
LL6
Ogg
L6
L6
L5
C3
H6
O
L6
Om
Om
Hx
Og
H3
H5
L4
Of
Weight
25
2.5
4130
2260
2.4
333
22
10
67
158
58
1575
283
64
s
4.4
38
5.5
910
113
136
11
170
339
175
120
754
94
253
216
81
17
107
5700
637
2168
448
10
s
0.2
335
1085
80
124
523
158
185
73
492
76
558
1491
93
2791
1.0
3.1
2.9
12
Name
Lawrence
Leedey
Leeds
Lenarto
Leon
Leoville
Le Pressoir
Les Ormes
Lesves
Lexington County
Lick Creek
Lillaverke
Lime Creek
Limerick
Lincoln County
Linville
Linwood
Lissa
Little Piney
Little River
Livingston (Montana)
Livingston (Tennessee)
Lixna
Locust Grove
Lodran
Logan
Lombard
Lonaconing
Long Island
Loomis
Loop
Loreto
Los Reyes
Lost City
Losttown
Lowicz
Loyola
Lua
Lubbock
Lucky Hill
Luis Lopez
Lumpkin
Lundsgard
Luotolax
Lutschaunig's Stone
Macau
Macibini
Madoc
Madrid
Mafra
Magura
Mainz
Malakal
Malotas
Malvern
Manbhoom
Mangwendi
Mantos Blancos
Locality
Kansas
Oklahoma
Canada
Czechoslovakia
Kansas
Kansas
France
France
Belgium
South Carolina
North Carolina
Sweden
Alabama
Ireland
Colorado
North Carolina
Nebraska
Czechoslovakia
Missouri
Kansas
Montana
Tennessee
USSR
Georgia
Pakistan
Oklahoma
Montana
Maryland
Kansas
Nebraska
Texas
Mexico
Mexico
Oklahoma
Georgia
Poland
doubtful
India
Texas
Jamaica
New Mexico
Georgia
Sweden
Finland
Chile
Brazil
South Africa
Canada
Spain
Brazil
Czechoslovakia
Germany
Sudan
Argentina
South Africa
India
Rhodesia
Chile
Class
L6
L6
Og
Om
H5
C2
L6
L5
L6
Og
Hx
H5
D
H5
L6
D
Og
L6
L5
H6
Om
Om
H4
Hx
Lo
H5
Hx
Og
L6
L6
L6
Om
Om
H5
Om
M
L5
L5
L5
Om
Om
L6
L6
Aho
L6
H5
Aeu
Om
L6
L4
Og
L5
L5
H5
Aho
LL6
LL6
Of
Weight
328
97
20
208
17 kg
482
3.6
0.3
1.5
2617
9.7
150
2.9
210
71
15
33 kg
118
78
170
294
123
58
2166
13
336
62
24
2696
229
149
89 kg
377
16 kg
71
82
1772
25
390
40
174
31
97
0.5
73
72
15
11
3.5
9
672
2
160
206
19
15
17
356
NUMBER 14
Name
Mapleton
Maria Elena (1935)
Marion
Marjalahti
Marshall County
Marsland
Mart
Matatiele
Mauerkirchen
Mayday
Mayodan
Maziba
Mbosi
McKinney
Mejillones
Melrose
Menow
Merceditas
Mem
Mertzon
Merua
Mesa Verde Park
Messina
Metsakyla
Mezo-Madaras
Miami
Mighei
Milena
Miller (Arkansas)
Miller (Kansas)
Mills
Milly Milly
Minas Gerais
Mincy
Misshof
Misteca
Moab
Mocs
Modoc
Mokoia
Molina
Molong
Monahans
Monroe
Monte Milone
Monze
Moonbi
Mooranoppin
Moore County
Mooresfort
Moorleah
Morito
Morland
Mornans
Morradal
Morrill
Morristown
Morven
Mosca
Locality
Iowa
Chile
Iowa
Finland
Kentucky
Nebraska
Texas
South Africa
Austria
Kansas
North Carolina
Uganda
Tanzania
Texas
Chile
New Mexico
Germany
Chile
Denmark
Texas
India
Colorado
Italy
Finland
Romania
Texas
USSR
Yugoslavia
Arkansas
Kansas
New Mexico
Australia
Brazil
Missouri
USSR
Mexico
Utah
Romania
Kansas
New Zealand
Spain
Australia
Texas
North Carolina
Italy
Zambia
Australia
Australia
North Carolina
Ireland
Australia
Mexico
Kansas
France
Norway
Nebraska
Tennessee
New Zealand
Colorado
Class
Om
Om
L6
P
Om
H5
Of
Og
L6
H
Hx
L6
Om
L4
Hx
L5
H4
Om
L6
Om
H5
Om
L5
H4
L3
H
C2
L6
H5
H
H6
Om
L
M
H5
Og
O
L6
L6
C3
L5
P
Opl
H4
L
L6
Om
Og
Aeu
H5
L6
Om
H6
H5
D
Om
M
H5
L6
Weight
3728
11kg
1717
452
67
244
454
69
73
5
13 kg
390
779
5208
12 kg
2782
110
724
220
3319
197
3439
3.4
s
335
1160
620
99
53
41
367
283
10
308
158
344
253
3346
5623
6.4
1189
864
913
252
0.6
265
286
149
460
161
s
636
517
32
58
87
1757
112
6136
Name
Mosquero
Mossgiel
Moti-ka-nagla
Motpena
Motta di Conti
Mount Ayliff
Mount Browne
Mount Dooling
Mount Dyrring
Mount Edith
Mount Egerton
Mount Joy
Mount Magnet
Mount Morris
Mount Ouray
Mount Padbury
Mount Sir Charles
Mount Stirling
Mount Vernon
Muddoor
Mulga (North)
Mulga (South)
Mundrabilla
Mungindi
Muonionalusta
Murchison
Murfreesboro
Murnpeowie
Muroc Dry Lake
Murphy
Murray
Muzzaffarpur
Nagaria
Nagy-Vazsony
Nakhla
Nakhon Pathom
Nanjemoy
Naoki
Nardoo No. 1
Nardoo No. 2
Narellan
Naretha
Narraburra
Nas
Nashville (Kansas)
Nashville (North
Carolina)
Navajo
Nazareth (iron)
Nedagolla
Needles
Neenach
Negrillos
Nejed
Nejo
Nelson County
Nenntmannsdorf
Neptune Mountains
Locality
New Mexico
Australia
India
Australia
Italy
South Africa
Australia
Australia
Australia
Australia
Australia
Pennsylvania
Australia
Wisconsin
Colorado
Australia
Australia
Australia
Kentucky
India
Australia
Australia
Australia
Australia
Sweden
Australia
Tennessee
Australia
California
North Carolina
Kentucky
India
India
Hungary
Egypt
Thailand
Maryland
India
Australia
Australia
Australia
Australia
Australia
Sweden
Kansas
North Carolina
Arizona
Texas
India
California
California
Chile
Saudi Arabia
Ethiopia
Kentucky
Germany
Antarctica
Class
H4
L4
H6
L6
H
Og
H6
Ogg
P
Om
Ae
Ogg
Opl
E6?
Om
M
Of
Og
P
H5
H6
H4
Om
Of
Of
C2
Om
Anom
L
Hx
C2
Opl
Aeu
Og
An
L6
H6
H6
H5
L6
L6
L4
Om
LL6
L6
O
Ogg
Om
Anom
Of
L6
Hx
Om
L6
Ogg
Ogg
Og
Weight
1629
560
77
98
2.4
95
486
370
672
1577
3731
2833
100
557
84
8147
21
383
134 kg
26
278
0.4
217 kg
573
419
30 kg
125
191
43
582
5830
5
s
36
644
852
79
227
75
109
33
2.7
321
17
23 kg
2676
517
315
28
10 kg
2210
15 kg
2931
183
1207
15
997
80 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
Name
Nerft
Ness County
Netschaevo
New Almelo
New Baltimore
New Concord
New Leipzig
Newport
New Westville
Ngawi
N'Goureyma
Nikolskoe
N'Kandhla
Nobleborough
Nocoleche
Nogoya
Nora Creina
Norcateur ?
Nordheim
Norfolk
Norfork
Norristown
North East Reid
North Forrest
North Haig
North Reid
North West Forrest
Norton County
Novo-Urei
Nuevo Laredo
Nulles
Nyirabrany
Oak
Oakley (iron)
Oakley (stone)
Oberlin
Obernkirchen
Ochansk
Odessa
Oesede
Oesel
Ogallala
Ogi
Ojuelos Altos
Okano
Okechobee
Okirai
Olivenza
Ollague
Olmedilla de Alarcon
Orange River (iron)
Orgueil
Orlovka
Ornans
Oro Grande
Oroville
Orvinio
Oscuro Mountains
Locality
USSR
Kansas
USSR
Kansas
Pennsylvania
Ohio
North Dakota
Arkansas
Ohio
Java
Mali
USSR
South Africa
Maine
Australia
Argentina
Australia
Kansas
Texas
Virginia
Arkansas
Georgia
Australia
Australia
Australia
Australia
Australia
Kansas
USSR
Mexico
Spain
Hungary
Australia
Idaho
Kansas
Kansas
Germany
USSR
Texas
Germany
USSR
Nebraska
Japan
Spain
Japan
Florida
Japan
Spain
Bolivia
Spain
South Africa
France
USSR
France
New Mexico
California
Italy
New Mexico
Class
L6
L6
Om
L5
Anom
L6
Og
P
Of
LL3
Anom
L4
Om
Aeu
Anom
C2
L4
L6
D
Om
Om
Om
H5
H4
Au
LL5
E6
Ae
Au
Aeu
H6
LL5
L5
Og
H6
LL5
Of
H4
Og
H5
L6
Og
H6
L
Hx
L4
H5
LL5
P
H
Om
Cl
H
C3
H5
Om
L6
Og
Weight
71
5810
267
1557
3216
8700
17 kg
222
3318
103
1519
32
146
0.1
203
4.5
8.5
133
1094
110
60
196
1.9
3.8
8.2
4.7
11
1797
93
132
124
s
0.8
111kg
1614
195
302
1668
128 kg
137
28
118
40
191
20
968
31
938
296
314
120
183
127
3.8
9.1
401
59
102
Name
Osseo
Otis
Ottawa
Oubari
Ovid
Owens Valley
Ozona
Ozren
Pacula
Palo Blanco Creek
Pan de Azucar
Pantar
Para de Minas
Paragould
Paranaiba
Parnallee
Pasamonte
Patos de Minas
Patrimonio
Patwar
Pavlodar
Pavlograd
Pavlovka
Peace River
Peckelsheim
Peck's Spring
Peetz
Pena Blanca Springs
Penokee
Perpeti
Perryville
Persimmon Creek
Pervomaisky
Pesyanoe
Petersburg
Petropavlovka
Pevensey
Phillips County
Piancaldoli
Pickens County
Piedade do Bargre
Pierceville (iron)
Pierceville (stone)
Pillistfer
Pima County
Pine River
Pinnaroo
Pinon
Pipe Creek
Pitts
Plains
Plainview (1917)
Plantersville
Pleasanton
Plymouth
Pohlitz
Ponca Creek
Portales No. 3
Locality
Canada
Kansas
Kansas
Libya
Colorado
California
Texas
Yugoslavia
Mexico
New Mexico
Chile
Philippines
Brazil
Arkansas
Brazil
India
New Mexico
Brazil
Brazil
Pakistan
USSR
USSR
USSR
Canada
Germany
Texas
Colorado
Texas
Kansas
Pakistan
Missouri
North Carolina
USSR
USSR
Tennessee
USSR
Australia
Colorado
Italy
Georgia
Brazil
Kansas
Kansas
USSR
Arizona
Wisconsin
Australia
New Mexico
Texas
Georgia
Texas
Texas
Texas
Kansas
Indiana
Germany
Nebraska
New Mexico
Class
Og
L6
LL6
L4
H
Om
H
Og
L6
Aeu
Og
H5
Of
LL5
L6
LL3
Aeu
Hx
L
M
P
L6
Aho
L6
Ah
L5
L6
Ae
H
L6
Opl
Opl
L6
Ae
Aeu
H
LL5
P
L3
H
Om
Om?
L
E6
Hx
Om
M
D
H6
Of
H5
H5
H6
H5
Om
L5
Ogg
L
Weight
37 kg
215
12
151
251
157 kg
1658
17
44
s
101
112
223
31kg
322
520
1585
827
14
592
12
97
8
551
2.2
629
810
457
3552
35
14 kg
3519
446
67
42
118
14
368
s
67
398
808
183
66
150
297
1132
1240
312
1022
363
24 kg
1862
204
619
0.1
29
156
NUMBER 14
Name
Port Orford
Potter
Prairie Dog Creek
Prambanan
Pribram
Pricetown
Providence
Puente del Zacate
Pulaski County
Pulsora
Pultusk
Puquios
Puripica
Putinga
Putnam County
Quartz Mountain
Queen's Mercy
Quenggouk
Quillagua
Quinn Canvon
Raco
Rafruti
Rakovka
Ramsdorf
Ranchapur
Rancho de la Pila
Rancho de la Presa
Rangala
Ransom
Ras Tanura
Rawlinna
Reager
Red River
Reed City
Reid
Reliegos
RenazzO
Rhine Villa
Richardton
Richland
Richmond
Rich Mountain
Rifle
Rio Loa
Rio Negro
Rochester
Rodeo
Roebourne
Rolla (1936)
Rolla (1939)
Rosario
Rosebud
Rose City
Rowena
Rowton
Roy (1933)
Roy (1934)
Ruff's Mountain
Locality
Oregon
Nebraska
Kansas
Java
Czechoslovakia
Ohio
Kentucky
Mexico
Georgia
India
Poland
Chile
Chile
Brazil
Georgia
Nevada
South Africa
Burma
Chile
Nevada
Argentina
Switzerland
USSR
Germany
India
Mexico
Mexico
India
Kansas
Saudi Arabia
Australia
Kansas
Texas
Michigan
Australia
Spain
Italy
Australia
North Dakota
Texas
Virginia
North Carolina
Colorado
Chile
Brazil
Indiana
Mexico
Australia
Kansas
Kansas
Honduras
Texas
Michigan
Australia
Great Britain
New Mexico
New Mexico
South Carolina
Class
P
L6
H3
Off
H5
L6
Om
Om
Og
H5
H5
Om
Hx
L6
Of
Om
H6
H4
Hx
Om
H5
D
L6
L6
H4
Om
H
L
H4
H
H5
L
Om
Og
H
L
C2
Og
H5
Hx
L5
L6
Og
Hx
L4
H
Om
Om
H
H
Og
H
H5
H6
Om
L5
L6
Om
Weight
24
1930
314
2.3
140
3.0
5492
2332
112
5.8
3263
90
15 kg
962
2736
57
424
35
741
45
133
22
23
20
59
54
12
26
295
5.9
2.8
37
79
1877
4.0
26
26
114
5155
1026
18
112
4367
940
66
53
1998
4018
414
10
190
5
3587
78
32
8607
4528
4762
Name
Rupota
Rush County
Rush Creek
Russel Gulch
Sacramento Mountains
St. Caprais-de-Quinsac
St. Francois County
St. Genevieve County
St. Germain-du-Pinel
St. Lawrence
St. Louis
St. Mark's
St. Mary's County
St. Mesmin
St. Michel
St. Peter
Saint-Sauveur
St. Severin
Salaices
SalinaSaline
Salla
Salles
Salta
Salt Lake City
Salt River
Sams Valley
San Angelo
San Cristobal
Sanderson
Sandia Mountains
Sandtown
San Emigdio
San Francisco
Mountains
San Jose
San Juan Capistrano
Santa Apolonia
Santa Catharina
Santa Cruz
Santa Isabel
Santa Luzia
Santa Rosa
Santa Rosalia
Santiago Papasquiero
Sao Jose do Rio Preto
Sao Juliao de Moreira
Saratov
Sardis
Sarepta
Savannah
Schenectady
Schonenberg
Schwetz
Scott City
Scottsville
Scurry
Seagraves
Locality
Tanzania
Indiana
Colorado
Colorado
New Mexico
France
Missouri
Missouri
France
Texas
Missouri
South Africa
Maryland
France
Finland
Kansas
France
France
Mexico
UtahKansas
Finland
France
Argentina
Utah
Kentucky
Oregon
Texas
Chile
Texas
New Mexico
Arkansas
California
Arizona
Mexico
California
Mexico
Brazil
Mexico
Argentina
Brazil
Colombia
Mexico
Mexico
Brazil
Portugal
USSR
Georgia
USSR
Tennessee
New York
Germany
Poland
Kansas
Kentucky
Texas
Texas
Class
L6
H
L6
Om
Om
L
Og
Of
H
LL6
H4
E5
LL3
LL6
L6
L5
E4
LL6
H4
Om
H5
L
H6
P
H5
Opl
Om
Om
D
Om
Ogg
Om
H4
Of
H5
H6
Om
D
C2
L6
Ogg
Anom
P
Anom
L5
Ogg
L4
Og
Og
Om
H5
L6
Om
H5
Hx
H5
H4
81
Weight
975
4020
3170
72
6550
0.4
236
1380
53
100
2.4
258
s
86
231
549
7.8
2463
395
221
897
318
50
24 kg
2.5
112
19
3443
136
471
1672
100
490
1334
142
18
5689
11kg
s
29
23 kg
5751
63
1176
7.5
2150
513
800 kg
254
27 kg
30
8.8
165
228
1042
26
13 kg
82 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
Name
Searsmont
Seelasgen
Segowlie
Seguin
Seibert
Seligman
Selma
Semarkona
Seminole
Sena
Seneca
Seneca Township
Seres
Serre de Mage
Sevrukovo
Seymour
Shalka
Shallowater
Sharps
Shaw
Shelburne
Shergotty
Shingle Springs
Shrewsbury
Shytal
Siena
Sierra Gorda
Sierra Sandon
Signal Mountain
Sikhote-Alin
Silver Bell
Silver Crown
Silverton
Simdndium
Sindhri
Sinnai
Sioux County
Sitathali
Ski
Sleeper Camp
Slobodka (1818)
Slobodka (1838)
Smith Center
Smithland
Smithonia
Smith's Mountain
Smithsonian Iron
Smithville
Social Circle
Soko-Banja
Somervell County
Soper
Soroti
South Bend
South Byron
South Dahna
South Oman
Spearman
Springwater
Locality
Maine
Poland
India
Kansas
Colorado
Arizona
Alabama
India
Texas
Spain
Kansas
Michigan
Greece
Brazil
USSR
Missouri
India
Texas
Virginia
Colorado
Canada
India
California
Pennsylvania
Pakistan
Italy
Chile
Chile
Mexico
USSR
Arizona
Wyoming
Australia
South Africa
Pakistan
Italy
Nebraska
India
Norway
Australia
USSR
USSR
Kansas
Kentucky
Georgia
North Carolina
Tennessee
Georgia
Yugoslavia
Texas
Oklahoma
Uganda
Indiana
New York
Arabia
Arabia
Texas
Canada
Class
H5
Og
L6
H
H
Og
H4
LL3
H4
H4
H
Of
H4
Aeu
L5
Og
Ah
Ae
H3
L6
U>
Aeu
D
Om
L6
LL5
Hx
Og
Of
Ogg
Ogg
Og
L6
M
H5
H
Aeu
H5
L6
L6
L4
L
L6
D
Hx
Om
Ogg
Og
Of
LL4
P
Anom
Off
P
D
O
E
Om
P
Weight
90
407
61
49
131
147
157
245
457
167
265
9372
2.0
25
23
1320
88
1017
401
1596
735
270
32
395
27
15
17 kg
72
41
4107
192
170
69
9.1
46
17
154
13
1.2
3.3
3.5
4.0
91
133
590
57
3400
3708
22 kg
337
80
1122
154
2.8
465
142 kg
13
870
2647
Name
Stalldalen
Stannern
Staunton
Stavropol
Steinbach
Stonington
Sublette
Success
Suchy Dul
Sultanpur
Supuhee
Susuman
Sutton
Suwahib (Buwah)
Sweetwater
Sylacauga
Tabor
Tadjera
Taiban
Tamarugal
Tambo Quemado
Tane
Tanokami Mountain
Tatahouine
Tathlith
Tawallah Valley
Tazewell
Tell
Temple
Tenham
Tennasilm
Ternera
Texline
Thiel Mountains
Thomson
Thoreau
Thule
Thunda
Thurlow
Tiberrhamine
Tieraco Creek
Tieschitz
Tilden
Timochin
Tirupati
Tjabe
Tjerebon
Tlacotepec
Tobe
Tocopilla
Toluca
Tomatlan
Tombigbee River
Tomhannock Creek
Tonganoxie
Tonk
Torrington
Toubil River .
Locality
Sweden
Czechoslovakia
Virginia
USSR
Germany
Colorado
Kansas
Arkansas
Czechoslovakia
India
India
USSR
Nebraska
Saudi Arabia
Texas
Alabama
Czechoslovakia
Algeria
New Mexico
Chile
Peru
Japan
Japan
Tunisia
Saudi Arabia
Australia
Tennessee
Texas
Texas
Australia
USSR
Chile
Texas
Antarctica
Georgia
New Mexico
Greenland
Australia
Canada
Algeria
Australia
Czechoslovakia
Illinois
USSR
India
Java
Java
Mexico
Colorado
Chile
Mexico
Mexico
Alabama
New York
Kansas
India
Wyoming
USSR
Class
H5
Aeu
Og
L6
S
H
L
L6
H5
L6
H6
Om
H5
H3
H5
H
H5
L5
L5
Om
Om
L5
Og
Ah
L6
D
Off
L6
L
L6
L4
D
H5
P
L6
Og
Om
Om
Om
L6
Om
H3
L6
H5
H
H
L5
D
H4
Hx
Og
H6
Hx
H5
Om
Cl
H
Om
Weight
192
497
6101
52
427
100
62
3503
1.8
71
89
74
305
5
1018
1682
65
69
438
17 kg
970
586
91
14
246
1267
1642
180
169
5535
4838
31
646
28 kg
217
294
449
209
403
100
4038
31
903
64
16
39
s
2400
5454
1467
109 kg
168
3001
31
231
3
116
175
NUMBER 14
Name
Tourinnes-la-Grosse
Travis County
Trenton
Trenzano
Treysa
Trifir
Troup
Tryon
Tucson
Tulia
Turtle River
Twin City
Tysnes Island
Uberaba
Ucera
Udei Station
Ularring
Ulysses
Umm Ruaba
Union
Union County
Ute Creek
Utrecht
Utzenstorf
Uwet
Uwharrie
Vaca Muerta
Valdinizza
Valkeala
Valley Wells
Vavilovka
Vengerovo
Vera
Veramin
Verkhne Udinsk
Vernon County
Victoria West
View Hill
Vigarano
Vincent
Vishnupur
Vouille
Wabar
Waconda
Waingaroraia
Waldo
Waldron Ridge
Walker County
Wallapai
Walltown
Walters
Warrenton
Waterville
Wathena
Weatherford
Weaver Mountains
Locality
Belgium
Texas
Wisconsin
Italy
Germany
Mali
Texas
Nebraska
Arizona
Texas
Minnesota
Georgia
Norway
Brazil
Venezuela
Nigeria
Australia
Kansas
Sudan
Chile
Georgia
New Mexico
Holland
Switzerland
Nigeria
North Carolina
Chile
Italy
Finland
California
USSR
USSR
Argentina
Iran
USSR
Wisconsin
South Africa
New Zealand
Italy
Australia
India
France
Saudi Arabia
Kansas
New Zealand
Kansas
Tennessee
Alabama
Arizona
Kentucky
Oklahoma
Missouri
Washington
Kansas
Oklahoma
Arizona
Class
L6
H5
Om
H6
Om
L
L6
L
Anom
H5
Om
D
H4
H5
H5
Om
L6
H
L5
Hx
Ogg
H4
L6
H
Hx
Om
M
L6
L6
L6
LL6
H5
L4
M
Om
H
Of
Om
C3
L5
LL6
L
Om
L6
Om
L
Og
Hx
Of
L
L6
C3
Anom
Hx
Anom
D
Weight
69
1480
391kg
285
195
3.0
114
2028
911kg
4917
387
3400
104
103
205
202
1.2
334
64
172
84
177
40
105
1105
1287
1449
850
s
11
3.0
34
50
52
36
9.3
34
1883
2051
s
24
161
299
1189
600
94
70
291
303 kg
3.7
25 kg
36
197
429
1216
773
Name
Webb
Weekeroo
Weldona
Welland
Wellington
Wellman
Wessely
Western Arkansas
West Forrest
Weston
West Reid
Weathersfield
Whitman
Wichita County
Wilbia
Wildara
Wiley
Willamette
Willaroy
Williamstown
Willow Creek
Willowdale
Wilmot
Wiluna
Wingellina
Winona
Witchellina
Wold Cottage
Wolf Creek
Wonyulg-unna
Woodbine
Wood's Mountain
Woodward County
Woolgorong
Wynella
Yambo No. 1
Yambo No. 2
Yandama
Yanhuitlan
Yardymly
Yatoor
Yayjinna
Yenberrie
Yonozu
York (iron)
Youndegin
Ysleta
Yurtuk
Zaborzika
Zacatecas (1972)
Zacatecas (1969)
Zavid
Zebrak
Zhovtnevyi
Zomba
Zsadany
Zvonkov
Locality
Australia
Australia
Colorado
Canada
Texas
Texas
Czechoslovakia
Arkansas
Australia
Connecticut
Australia
Connecticut
Nebraska
Texas
Australia
Australia
Colorado
Oregon
Australia
Kentucky
Wyoming
Kansas
Kansas
Australia
Australia
Arizona
Australia
Great Britain
Australia
Australia
Illinois
North Carolina
Oklahoma
Australia
Australia
Zaire
Zaire
Australia
Mexico
USSR
India
Australia
Australia
Japan
Nebraska
Australia
Texas
USSR
USSR
Mexico
Mexico
Yugoslavia
Czechoslovakia
USSR
Malawi
Romania
USSR
Class
L6
Og
H
Om
H
H5
H
Of
H5
H4
H6
L5
H
Og
H5
H5
Opl
Om
H3
Om
Og
H
H6
H4
H4
E6?
H4
L6
Om
Om
Of
Of
H4
L6
H4
H
L3
L6
Of
Og
H5
L6
Og
H4
Om
Og
Anom
Aho
L
Anom
Om
L6
H5
H
L6
H
H
Weight
3.1
11kg
1070
213
4224
543
0.2
1672
2.3
181
0.8
290
21
2251
3.8
s
189
2721
187
1223
4866
74
133
284
s
224
156
57
350 kg
577
48 kg
1786
44 kg
362
431
4.0
3.2
377
1200
81
222
14
3836
44
30
5756
19
33
4.1
444
6600
234
73
703
270
13
40
Chemical Analyses of Two Microprobe Standards
Eugene Jarosewich
ABSTRACT
Chemical analyses are given for two materials foruse as microprobe standards: volcanic glass from
the Mid-Atlantic Ridge and hornblende fromArenal Volcano, Costa Rica.
There has been increasing need for well-analyzed
microprobe standards because, as with any instru-
mental method, a standard compositionally similar
to an unknown is the most likely to give reliable
results. The key requirements for microprobe stan-
dards can be found in the literature in general and
some are briefly summarized in an earlier paper
(Jarosewich, 1972).
The chemical analyses were performed employing
analytical methods described by Hillebrand, et al.
(1953) and Peck (1964) except for the determination
of sodium and potassium. The latter elements were
determined by flame photometry. Data for two new
microprobe standards follow.
ARENAL HORNBLENDE USNM 111356.?A green
hornblende xenocryst from an andesitic basalt
scoria was collected by Dr. W. Melson from the
September 1968 eruption of Arenal Volcano, Costa
Rica. The sample was crushed and sieved and the
60-100 mesh fraction purified using a heavy liquid
(D 3.19-3.21) followed by Franz magnetic separa-
tion. The purified sample contains less than 0.5
percent grains with adhering groundmass. An elec-
tron microprobe check indicates the sample to be
Eugene Jarosewich, Department of Mineral Sciences, National
Museum of Natural History, Smithsonian Institution, Wash-
ington, D. C. 20560.
TABLE 1.?Chemical analyses of two microprobe
standards (analyst: E. Jarosewich)
Constituent
SiO2
A12O3
Fe2O3
FeO
MgO
CaO
Na2O
K2O
H2O+
H2O-
TiO2
P2O5
MnO
Total
Total Fe
Basaltic Glass
Arenal Hornblende Juan de Fuca Ridge
(USNM 111356) (USNM 111240152)
41.46
15.47
5.60
6.43
14.24
11.55
1.91
0.21
1.21
<0.01
1.41
<0.01
0.15
99.64
8.91
50.81
14.06
2.23
9.83
6.71
11.12
2.62
0.19
n.d.
0.02
1.85
0.17
0.22
99.83
9.20
n.d. = not determined.
homogeneous within the limits of analytical error.
BASALTIC GLASS USNM 111240/52.?This sample
was dredged from the median, flat-floored valley of
the southern part of the Juan de Fuca Ridge (44?
40'N, 130?30'W, 2195-2220 m). Approximately 2
percent of the sample is composed of inclusions,
primarily plagioclase and olivine. Because the aver-
age combined chemical composition of these inclu-
sions is similar to the composition of the glass and
because they are present in small amount, the bulk
chemistry of the sample is essentially the same as
85
86 SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
the glass. Electron microprobe analyses shows the 1953. Applied Inorganic Analysis. 2nd edition, 1034 pages.
glass to be extremely homogeneous in major and New York: J?hn Wile? and Sons-
. , / s> J Jarosewich, E.
minor element content. .??? ?, . , . , , r. ... , , ,,. ,
1974. Chemical Analyses or Five Minerals tor microprobe
Standards. Smithsonian Contributions to the Earth
Literature Cited Sciences, 9:83-84.
Peck, L. C.
Hillebrand, W. F., G. E. F. Lundell, H. A. Bright, and J. I. 1964. Systematic Analysis of Silicates. U. S. Geological Sur-
Hoffman vey Bulletin, 1170:66.
Preparation of Multiple Microprobe Samples
Grover C. Moreland and Richard Johnson
ABSTRACT
Several microprobe samples may be prepared simul-taneously using a new technique described here.
With judicious choice of samples the same tech-nique may be used to prepare microprobe standard
discs.
hardener. This mixture is stirred on the 40? to 45?
C hot plate until it reaches a watery consistency.
Each sample is now inserted in its appropriately
numbered hole. Then using a dissecting needle,
place some of the epoxy mixture on the plexiglass
disc and with the needle push some of the mixture
over the edge of each hole. The epoxy will run
down the wall of the hole, hit the cover glass, and
A preparation technique has been developed
whereby it is possible to prepare several probe
samples simultaneously. The bask: mount is a
plexiglass disc y4 inch thick cut from a 1-inch rod.
Holes of any convenient size are drilled through
the disc and the holes numbered with a scriber. As
many as 19 samples can be mounted and prepared
simultaneously (Figure 1).
A white 3X5 inch card is fastened to a hot plate
with masking tape; this card provides a visual con-
trast against which the sample can be easily seen.
A carefully cleaned one inch cover glass is placed
on the white card, and the sample disc placed on
the cover glass. A weight is added to insure good
contact between the disc and the cover glass, and
the hot plate temperature adjusted to 40? to 45? C
(100?-115? F). Rubber cement is now painted
around the edge of the coverglass-plexiglass disc
sandwich, and the mount is left on the hot plate
for 5 to 10 minutes, or until cured as shown by a
brown color.
A mixture of epoxide and expoxide hardener
is now prepared (Buehler 20-8130 epoxide and
20-8132 hardener or equivalent). A convenient
amount is 4.8 gms of epoxide and 0.8 gm of
Grover C. Moreland and Richard Johnson, Department of
Mineral Sciences, National Museum of Natural History,
Smithsonian Institution, Washington, D. C. 20560.
E D
FIGURE 1.?Schematic of plexiglass disc. (A = Dissecting needle,
B = epoxy mixture, C = cross-section of plexiglass disc, D =
coverglass, E = mineral grain, F = hole drilled in plexiglass
disc.)
rise. This procedure should leave the specimen
in contact with the cover glass, with epoxy around
and above it, but very little if any beneath it. One
can check under a stereomicroscope to see that the
sample is resting on the cover glass (i.e., at the bot-
tom of the hole) and that no air bubbles have
formed. If these conditions are not met then the
samples can be pushed down and air bubbles freed
87
SMITHSONIAN CONTRIBUTIONS TO EARTH SCIENCES
using a small dissecting needle. One then covers
the entire upper (numbered) side of the disc with
epoxy, thus insuring that the numbers will not be
accidentally obliterated during the subsequent steps
of preparation and during later use of the disc.
The epoxy mixture should be allowed to dry on
the hot plate for about 30 minutes at 50? C (125?
F). The card-coverglass-plexiglass disc assembly is
then placed in a 50? C oven for about 10 hours. The
assembly is then removed from the oven and the
card separated from the remainder of the unit.
The samples are exposed by grinding off the cover
glass on an 850 rpm lap with a slurry of water and
$:600 aluminum oxide. Being careful to keep the
lap damp, grind the disc for 2 to 3 minutes or
until the samples are exposed. Also grind the num-
bered side, being careful not to grind off the num-
bers. The disc should now be washed thoroughly
with soap and water and dried.
The samples are now brought to a rough polish
by covering a 600 rpm wheel with a double thick-
ness of brown wrapping paper that has been
covered with polishing oil and 3 micron diamond
compound. (The diamond compound regularly used
is VB-3, produced by the Glennel Corp. of West
Chester, Pennsylvania.) The disc should be moved
counter to the rotation of the wheel; a lapping time
of about 5 minutes is usually required. After wash-
ing and drying the polish is examined under a re-
flected light microscope. If necessary repeat the
lapping until the only scratches visible are those
made by the 3 micron polishing compound. The
numbered side of the disc should also be polished to
make the numbers visible.
Next, the disc is lapped on a 150 rpm wheel
covered with a polishing cloth. The polishing cloth
is covered with water to which 0.05 micron alumina
is sparingly added until a damp-dry condition is
reached. The disc is moved for 30 turns counter to
the rotation of the wheel. The disc, same side down,
is rotated 180? and given another 30 turns on the
wheel. Once again it is washed and dried.
Finally the disc is lapped on a 50 rpm wheel
covered with a clean polishing cloth, giving it only
5 turns with light pressure, counter to the rotation
of the wheel. This final polishing is to remove any
possible contamination due to smearing.
The disc is now washed and dried and examined
under a reflected light microscope at about 80
magnification. The 3 micron scratches should no
longer be visible and one should see only highly
polished, optically fiat grains.
One should exercise care in choosing specimens
to be mounted in one disc as some very soft
material may smear upon polishing and hence
cause contamination of the rest of the specimens.
Tests were made here on eight samples of varying
hardness (metallic zinc, gold, magnesium, alumi-
num, silver, platinum. Ni-Fe-Co alloy, and lead
sulfide (galena)), and mounted and polished on a
single disc. A check was made by electron micro-
probe to determine any effects of smearing from
one sample to another due to the polishing tech-
nique. For each sample (sulfur in the case of the
galena) the background level above and below the
peak was determined. Each sample was then probed
for Zn, Au, Mg, Al, Ag, Pt, Fe, Co, Ni, and S. Sulfur
was the only element that showed a measurable
concentration in some of the other samples on the
disc, hence the galena did contaminate the other
samples.
Preliminary tests have shown that with judicious
choice of samples, microprobe standard discs may
be prepared using this technique, thus effecting a
considerable saving in time and materials over the
one-at-a-time method of polishing microprobe stan-
dards.
ACKNOWLEDGMENT.?We wish to thank Charles
Obermeyer and William Potts for their assistance in
the preparation of this paper.
U.S. GOVERNMENT PRINTING OFFICE: 197S 546-367/32