MORDENITE AND ASSOCIATED MINERALS FROM NEARCHALLIS, CUSTER COUNTY, IDAHO.
By Clarence S. Ross,Geologist, United States Geological Survey,ANDEarl V. Shannon,Assistant Curator of Geology, United States National Museum.
INTRODUCTION.In 1917 Mr. Milton A. Brown, of Challjs, Idaho, sent a specimenof a very fine cottony mineral of snow white color to the UnitedStates National Museum for identification. This mineral, which wasthought to be asbestus, was said to be available in considerablequantity. Upon examination in the Museum laboratory this wasfound to be a fibrous zeolite, and optical examinations by Wherryand Larsen showed it to have properties similar to those of therare mineral ptilolite. Chemical examination by Koch supportedthis conclusion, although the mineral was so mixed with smallgrains of quartz that it was not suitable for analysis. A short noteon the occurrence was published by Koch.^In 1921 a box of some 50 pounds of specimens of the zeolite andassociated minerals from this locality was received at theMusemn.This material has been subjected to a detailed examination, theresults of which are here presented.NOMENCLATURE.Dr. Waldemar T. Schaller has recently critically reviewed thegroup of zeolites which includes mordenite, ptilolite, and flokite.In addition to the existing literature, which includes recent workby Walker and Boggild, Schaller has considered new analyses madeby himself and the analytical data presented below together withnumerous new optical determinations. One of the results of thisadmirable work was to show that the Idaho mineral here consideredis not ptilolite, as at first supposed, but is, in reality, mordenite.^
1 Louis A. Koch. A new occurrence of ptilolite. American Mineralogist, vol. 2, p. 143, 1917.* Waldemar T. Schaller. Unpublished paper presented before the first meeting of the MineralogicalSociety of Washington, February, 1923. See abstract in Amer. Mineralogist, vol. 8, p. 93, 1923.No. 2509.
?
Proceedings U. S. National Museum, Vol. 64. Art. 19.1
2 PROCEEDINGS OF THE IJATIONAL MUSEUM. vol.64LOCALITY.The specimens were received from Dr. Charles L. Kirtley, a physi-cian of ChalHs, who is the original discoverer of the occiu-rence.Some years ago Mr. Brown had found, in the low hills south ofChallis, some pebbles of colored corundum of such a quality as toindicate the possible occurrence of gem material. \Miiie searchingfor the corundum locality, which has not been rediscovered, DoctorKirtley found the mordenite. The writers wish to gratefully acknowl-edge their indebtedness to Doctor Kirtley for his generosity insupplying these interesting materials together with notes on theirplace and mode of occurrence.Challis is an interior community in Salmon River Valley 58 milesby stage northwest of Mackay, the nearest railroad point knd theterminus of the Lost River branch of the Oregon Short Line Railroad.The zeolites and other minerals are from a small area of amygdaloidalrock in the low hills south of Challis in the wedge formed by SalmonRiver and Antelope Valley and a short distance west of the Challis-Mackay road. GEOLOGY.As described by Doctor Kirtley, the zeolite bearing formationoccupies an area 3 acres in extent, surrounded by low hills of yellowishclay underlain by soft shale, the stratification being inclined about20?. In this area the large masses of mordenite may be seen atmany places, although the specimens collected all came from asingle shallow cut. The shale hills are strewn wdth small pebble-like masses of translucent quartz and chalcedon}^ A map accom-panying a reconnaissance report by Umpleby ^ shows this area isoccupied by lacustrine sediments with included tuffs and flows ofandesite and basalt of probable Ivliocene age. The following descrip-tion is quoted from this report.Lacustrine deposits occur in the northeastern and eastern portions of thearea studied. The northern area comprises an irregular belt about 10 milesacross with tongues reaching up the several valleys tributary to the basin inwhich Challis is situated. To the south, west, and north the lake beds giveway to lava flows, but to the south they connect across a low divide at thehead of Antelope Valley with similar deposits in the valley of Big Lost River.The area east of Bay Horse is of peculiar shape but seems to have an outwardcontinuation south of Antelope Ridge.Perhaps the best exposures of these beds are along the east bank of SalmonRiver below Challis. Here the bluffs are about 200 feet high and, as seen fromacross the river, are made up of light gray sandstones, probably tuffaceous, anddove-colored shales. Conglomeratic members are sparsely and irregularlydistributed through the section. The beds vary in thickness from less than aninch to several feet. In bluflfs north of Challis, tuffaceous material, generally
' Joseph B. Umpleby. Some ore deposits in northwestern Custer County, Idaho. U. S. Geol. SurveyIJuU. 539, p. 27, 1913.
ABT.19. MORDENITE?BOSS AND SHANNON. 6
very fine grained and of green, gray, or white color, predominate. It is nicelybedded, in places thinly laminated. , ^, , ..The hurried observations afforded little idea of the thickness of the deposits,although it must unquestionably be hundreds of feet.Fossils were not found in this area, but the topographic relations of thesebeds are similar to those at Salmon, which were assigned to the Miocene onfloral evidence. -.atJulien * has published petrographic descriptions of several volcanic tuffs fromChallis.It seems most probable that the bed containing the mordenite andother minerals here described is an exceedingly vesicular andesiteflow interbedded with the Miocene lake sediments. The larger geodalmasses of the zeohtes have some inclosing rock attached, but this isfriable and sandy textured and falls to pieces except where it has beensilicified-by processes attending the deposition of the minerals in thecavities One spechnen which was labeled "country rock" consistsof very numerous small pebblehke amygdules fiUed with mordenite,heulandite, and diabantite or a related chlorite, in a friable greenishmatrLx. A thin section of this rock, which below is frequentlyreferred to as " the amvgdaloid," shows the matrix to consist of pheno-crysts of augite in a glass base. FuUy half of the volume of the rockis occupied by the filled vesicles, as shown in the two middle illustra-tions of plate 2. The forms of these smaU amygdules vary fromspherical to irregular, but show them unquestionably to be gas cavi-ties rather than replacement nodules in a tuffaceous rock.MORDENITE.Mordenite was originaUy described from Nova Scotia, where it hasbeen found at several locaHties, but only in relatively small amountat each, as compact masses with confused fibrous structure fiUingsmall cLvities m lavas. No finely fibrous or cottony material isknown from Nova Scotia, although, as shown by Schaller in the paperpreviously cited, the flokite from Iceland, which has such structure,is identical with mordenite. The mineral seems to be present as finerspecunens and m greater abundance at the Idaho locahty than at anyother known occurrence. In the Idaho spechnens exammed themordenite occurs in a variety of forms ranging from fine cottony feltedaggregates through radial-fibrous masses to dense compact porce-laneous materials. Most abundant and conspicuous are the largegeodal masses of fine cottony material which reach a diameter of afoot or more. These vary from round to irregular in cross sectionand consist usuaUy of a thin shell of chalcedonic material fiUed withthe fight tufted fibers of the mordenite. Although this cottony ma-teriafappears entirely homogeneous to the unaided eye, it aU con-tains gritty particles which can be felt when it is crushed, and upon
? Alexis A. Julien. Volcanic tuffs from ChaUis, Idaho. Trans. N. Y. Acad. Sci., vol. 1, pp.49-53, 1882.20183?25?Proc.N.M.vol.64 29
4 PKOCEEDINGS OF THE NATIONAL MUSEUM. vol. 64
microscopic examination these are found to be minute spherulitic orspindle-shaped grains of quarts with radial structure. Koch selectedwhat he thought was a very pure material for analysis and estimated,after the analysis had shown an abnormal amount of silica, that theanalyzed material had contained about 33 per cent of quartz. Schal-ler's recalculation shows that about 40 per cent would be nearer thetruth. The small quartz spherulites are pinned through by the mor-denite fibers and have clearly developed subsequent to the mordenite.None of the fine cottony material was found to be free from quartz, al-though the compact varieties are frequently free from any impurity.Practically all of the geodes of the cottony type have a very thin outershell of heulandite which forms the first lining in the cavity. Withinthis there is usually a thin layer of chalcedony upon which the mor-denite was deposited, filling the interior of the cavity with loosely ag-gregated fibers. A typical specimen of this sort, although one of thesmallest received, is that shown in plate 1, upper left. Many of thecavities have a thicker crust which is found upon microscopic exami-nation to consist of chalcedony including mordenite fibers which havethe same attitude as those which occur free in the middle portion ofthe cavity. This chalcedony appears to be later than the mordeniteand to have formed a crust by saturating the pore space of the loose-textured mordenite. In some geodes of moderate size only a smallarea remains in the center where the mordenite is not saturated withsilica, the specimen shown in plate 2, lower, being an example, whilein extreme cases, noted only in the smaller masses, the entire contentof the cavity has been saturated with chacedony, or, more rarely,granular quartz. One striking specimen illustrated in plate 3, lower,has silky bundles of short fibers of mordenite attached to masses ofsmall quartz spherulites which project from platy (argentine) calcite.Analcite occurs as flattened crystals between the calcite plates. Boththis spheruhtic quartz and the calcite are probably younger than themordenite and the calcite has apparently, in part, been dissolvedaway. The microscope shows the mordenite fibers of this specimento be individually coated with a thin film of calcite. In other speci-mens the fibers were similarly incrusted with quartz. A part of thelight cottony vaiiety contained many of the small radial grains ofquartz but was free from the incrustations. This was separated fromits included quartz grains by alternate rubbing between hardwoodboards and screening, the quartz grains when freed from the fiber,passing through the meshes of the sieve while the tufts of mordeniteremained behind. This process was continued until no more grittyparticles could be felt in the sample when it was examined opticallyand found to be pure and free from quartz. Upon analysis this sam-ple gave the following results and ratios.
MORDENITE?ROSS AND SHANNON.Analysis (1) and ratios of cottony mordenite.
Constituent.
SiOjAI2O3CaOMgONaaOK2OH2O+110? C.H2O-110? C.Total..
Per cent. Ratios.0.119 X 9
.118 X 1
.126 X 1
.137 X 6
1.00 X 9
.99 X 11.06 X 11.15 X 6
The material of this analyzed sample was biaxial and negative withthe orientation XAc = 3?40', Y = a, Z = h, the indices of refractionbeing a = 1 .470, jS = 1 .475, 7 = 1 .475, 7 - a = 0.005.The second sample analyzed came from a pure nodule about 8centimeters in maximum diameter which has evidently been weath-ered upon the surface for some time, as it is weather stained; theoriginal shell is bleached, and the mordenite is in places stained greenby algae. The nodule is red on its original surface from a thin coatingof heulandite, succeeded inward by a thin shell of chalcedony, insidewhich the mass consists of pure mordenite in fibers extending fromall sides toward the center. The fibers are straight and close packedand reach a maximum length of 2 centimeters. The color is creamywhite and the luster, on fresh fracture, is faintly silky. The analysisgave the following results
:
Analysis {2) of weathered fibrous mordenite nodule.
Constituent.
Si02AI2O3CaOMgONazOK2OH2O+110?CH2O-110?CTotal
Ratios.
.093
.121
.0631
.012
.057
.006J
.5041
.240/
0.121 X 9
.121 X 1
.138 X 1
.124 X 6
1.00 X 91.00 X 11.13 X 11.02 X 6
The refractive indices of this sample, measured on the analyzedmaterial are : a= 1 .472, ^ = 1 .475, 7 = 1 .476, 7 - a= 0.004. All of thefibers which are thick enough to have appreciable birefringence werefound to have a small inclined extinction.Many of the specimens received are amygdules free from matrixwhich apparently were gathered, like so many pebbles, from the
b PROCEEDINGS OF THE NATIONAL MUSEUM. vol.64
surface. These vary in size and are very irregular in form butapparently they represent casts of the interiors of very irregularbubble cavities probably formed by the sudden expansion of steamin viscous lava. They all preserve the original outside crust of redcolor which resembles a stain of iron oxide, but which, upon moreminute examination, is found to be the thin skin of heulandite whichformed the first lining of the cavity. All of these heulandite coatedmasses are hard and compact, but they vary somewhat in internalstructure and composition. Some of them are entirely filled withbluish chalcedony and quartz, others are mixtures of mordenite andchalcedony or of mordenite and quartz, while a majority contain onlypure mordenite The mordenite of the latter varies from distinctlyradial fibrous to very compact in structure The third analysis wasmade upon such a nodule which was exceedingly tough and difficultto break. Under a lens this shows a confused and interwovenfibrous structure and when examined under the microscope itsstructure is finely felted fibrous. The mean index of refraction, theonly one determinable, is 1.473. This description applies equallywell to a type specimen of How's original mordenite from NovaScotia in the Roebling collection which has a mean index of refractionof 1.473 and a birefringence of about 0.005. The material from thisnodule which was analyzed was shown to be pure and free fromextraneous substances by optical examination. The analysis gavethe following results
:
Analysis (S) and ratios of compact nodule of mordenite.
Constituent.
ART. 19. MORDENITE?EOSS AND SHANNON.
rare thin fibers attached to heulandite crystals lining hollow flesh-redmasses of heulandite. The snow-white masses show on fracturethe texture of porcelain and are lusterless and exceedingly tough.These, when ground for analysis were found to be pure with a veryfinely felted fibrous structure. The mean index of refraction of theanalyzed material was 1.473. The analysis gave the followingresults and ratios:Analysis (4) of small pebblelike masses of mordenite.
Constituent.
SiOsAI2O3CaOMgOK2ONa20H2O+110? CHjO-llO" CTotal.
Ratios.
1.099 0.122 X 9 1.00 X 9
.116 .116 X 1 .95 X 1
.0491?Oil .133 X 1 1.09 X 1!066|
:280} -128 X 6 1.05 X 6
The four analyses of the Idaho material agree in giving the silicaratio as 9 instead of 10 molecules and the water as 6. For con-venience in comparing the silica content, the several analyses arebrought together and their average compared with the values cal-culated to suit the two formulas (Na2,Ca)O.Al203.9Si02.6H20. and(Na,,Ca)O.Al2O3.10SiO2.6H2O in the following table, a 1 : 1 ratiobetween CaO and NajO being assumed in calculating the theoreticalcompositions
:
Comparison of mordenite analyses with formula percentage values.
8 PKOCEEDINGS OF THE NATIONAL MUSEUM. VOL. 64the Colorado ptilolite (Ca,K2,Na2)O.Al2O3.10SiO2.7H2O, and 1molecule of silica and ? molecule of water less than the formulacommonly given for mordenite, based upon Pirsson's Wyomingmaterial (Ca,K2,Na2)O.Al2O3.10SiO2.6?H2O. Boggild^ has recentlyconcluded that flokite is identical with ptilolite, while Walker ^suggests that ptilolite and mordenite are probably identical. Schaller,in the paper above referred to, however, has shown that mordeniteand flokite are identical, but that mordenite, which is monoclinic,is distinct, chemically as well as crystallographically, from ptilolitewhich is orthorhombic. The material from Wyoming described asmordenite by Pirsson, which is distinct from either mordenite orptilolite (clinoptilolite, Schaller), was instrumental in confusing thespecies mordenite. The best available data on the refractive indicesof these minerals is assembled below:Refractive indices of mordenite and related minerals.
Mineral.
MordeniteDoDoDoDoPtilolite....Mordenite.DoFlokite
Locality.
IdahoIdaho (1).Idaho (2)
.
Idaho (3).
_-.do(4)-.Colorado.NovaScotia.
._-doIceland _
Authority.KochRoss
._-do
...do
...doLarsen
..-do ?..-Ross *CalHsen..
Refractive indices.
1.475 1.477 1.4781.470 1.475 1.4751.472 1.475 1.4761.4731.4731.476 1.480 1.4801.4731.471 1.4751.472 1.474
Extinction.
Undetermined.IncUned.Do.Not determin-able.Do.Parallel.Not determin-able.Do.Inclined.
1 Specimen of How's original material from Morden, Nova Scotia, owned by Col. W. A. Roebling. Thevalue given originally (1 465) in Larsen's tables (Bull. U. S. Geol. Surv. No. 079, p. 113, 1921) has been foundto be an error.
' Specimen of recent collection by Walker and Parsons from Hall's Harbor, Kings County, Nova Scotia(Cat. U. S. N. M. 94, 552). Data determined by Ross and not pre\iously published.Two fibrous minerals, one orthorhombic with parallel extinction,and the other monoclinic, with inclined extinction, are representedby ptilolite and mordenite which differ one molecule in silica and onemolecule in water. These minerals are very similar, however, bothin composition and properties, and very careful work is necessary todistinguish between them. It may be mentioned, in passing, thaterionite has the same formula type, the same habit, and similar op-tical properties, although the refractive indices are much lower.This mineral has the formula (Ca,K2,Na2)O.Al203.6Si02.6H20. MissCallisen made flokite monoclinic on the basis of a constant extinc-tion inclination of 5? on (010) combined with twinning on (100).These observations Boggild set aside as probably due to inexact ob-
' O. B. Boggild. Dansk. Videnskab. Selskab. Mathemat.-Phys., vol. 4, No. 8, p. 19, 1922.? Thos. L. Walker. Contributions to Canadian Mineralogy, 1922. Univ. Toronto Geol. Series, No. 14,p. 59, 1922.
ABT.19. MORDENITE?BOSS AND SHANNON. 9
servation despite the fact that Thugutt had described a "ptilolite"from Seiser Alp which showed inclined extinction. With this inmind the Idaho material was very carefully examined, its composi-tion and refractive indices being identical with those given for fiokite.Since the fibers are very thin and the birefringence is very low anextinction obliquity of several degrees might readily evade observa-tion. Such a small angle might also be attributed to lack of adjust-ment between the cross hau-s and nicols of the microscope. Afterspecially adjusting the microscope the Idaho material was found togive, as the mean of about 20 good measurements, an extinctioninclination of 3^?, measured from the elongation of the fibers. In-stead of containing errors of observation, Miss Callisen's work onflokite appears to be a very commendable and highly exact researchwhich has furnished the key for the interpretation of the group.The Idaho mordenites are interesting especially as furnishing acomplete gradation between the physically very dissimilar compactmordenite from Nova Scotia and the cottony fibrous ''fiokite" fromIceland. HEULANDITE.Heulandite occurs in the specimens as a complete filling of smallvesicles in the rock, as drusy linings of cavities, and as thin crustsforming the first lining of cavities later filled with mordenite,chalcedony, or quartz. The characteristic flesh-red masses in theamygdaloid consist of fine transparent heulandite, and many of thesesmall masses are hollow and present fine terminations of acute wedge-shaped crystals, the habit being as shown in figures 1 and 2. Theseshow very well developed cleavage parallel to the pinacoid b(OlO) aswell as some other less perfect cleavages at right angles to this plane.Usually two or several crystals are grown together in more or lessparallel position, but the tendency to parallel growth is not so con-spicuous as usual in the heulandite of this type. A mp,jority of thesepebbles are not hollow, but are solid masses of the heuandite. Thosewhich do show open centers often have a few fine fibers of mordeniteextending across the central cavity. Other heulandite-lined vesiclesin the same rock are completely filled with closely packed cottonymordenite and the faintly pink to salmon colored masses which fillmany of the larger bubble cavities are compact mordenite with athin outer skin of heulandite. The numerous irregular masses whichhave weathered free from the inclosing rock, and which consist ofcompact mordenite, of mordenite saturated with quartz or withchalcedony, or of quartz or chalcedony alone, all have a brick-redouter coating which resembles a stain of ferric oxide, but which,when examined with a lens, is found to consist of a thin outer crustof crystalline heulandite. All of the larger geodal masses of mor-denite, of whatever type, have a thin deposit of heulandite next the
10 PROCEEDINGS OF THE NATIONAL MUSEUM. VOL. 64inclosing rock. Optical examination of the clear crystals from hol-low heulandite amygdules showed them to be transparent and color-less and devoid of any visible pigmenting material to which the redcolor might be attributed. Optically the crystals lining the smallcavities or making up the flesh-red amygdules of the amygdaloidrock are biaxial positive ( + ) with 2V = 52?. The indices of refrac-tion are a =1.482, /3 = 1.485, 7 = 1.489, 7- ? = 0.007. The acutebisectrix (Z) is perpendicular to the 6(010) face, while the extinction,measured from the c crystallographic axis averages 35?, the opticplane being nearly parallel (measured 5? to 8?) to the trace of one ofthe terminal clinodomes, either t{101) or s(TOl). The crystals asseen between crossed nicols are frequently not entirely simple, being
1. 2. 3. 4.Figs, l to 4.?Heulandite Crystals from Challis, Idaho.divided by sutures into areas which differ slightly in extinction angle.Thus the crystal measured and illustrated by the drawing (fig. 2),was found to be divided vertically by a straight line, on one side ofwhich the extinction, measured from the vertical cleavage lines,gave 38^? while on the other side it was 44^? and a wedge enteredat the termination of the crystal in which the extinction was 33?,the latter being about the normal extinction angle of homogeneouscrystals. In addition to this peculiarity the crystals frequentlyshow a very narrow outer border of a material of decidedly higherindex of refraction.The crystals of this heulandite which were measured did not giveoverly satisfactory angles, mainly because of the poor development,
AET. 19. MORDENITE?ROSS AND SHANNON. 11bulging, or parallel growth in the vertical zone rendering accurateorientation on the goniometer very difficult. The average degree ofagreement may be shown by giving, in detail, the measurementsobtained upon two crystals, shown in figures 1 and 2, which were asfollows
:
Measurements obtained on heulandite (fig. 1).
Letter
.
12 PKOCEEDINGS OF THE ZsTATIOFAL MUSEUM. VOL. 64
essentially pure and homogeneous by optical study. The percentagesand ratios obtained are as follows:Analysis and ratios of heulandite.
Constituent.
ART. 19. MOEDENITE?ROSS AND SHANNON. 13
rock or filling gashes in the mordenite which probably result fromthe solution of plates of calcite as shown in plate 1, upper right.Individual crystals of the analcite reach an extreme diameter of 14millimeters. Where free from inclusions the analcite is clear andcolorless, but the crystals commonly inclose fibers of mordenite andspherulites of quartz, giving them a mUky appearance. They arefor the most part simple trapezohedrons of the form qill2), whichis a common form for analcite as shown in figure 5. Occasionallyhowever, one shows faces of the cube a(lOO) and of the trigonaltrisoctahedron w{233) which, though previously known, are rareforms on analcite. A modified crystal is shown in figure 6. Certainvery interesting crystals occur in the platy argentine calcite. Theseare attached to the calcite plates by the equatorial plane and, wherespace was available, the top halves have developed completely.Where the space between the calcite plates was thin the analcites
5. 6.Figs. 5 and 6.?Analcite crtstals from mordenite mass, Challis, Idaho.
are very much flattened and are bounded only on their edges bynarrow crystal planes. Both undeformed and greatly flattenedcrystals are shown attached to calcite in plate 3, right. The disk-like crystals may reach 1 centimeter in diameter with a thickness ofonly 0.2 miUimeter, although most of them are tliicker than this,with a smaller diameter. Selected ones of these are ideal thinsections, and show admirably the birefringence and division intosectors Such a crystal photographed in polarized light betweencrossed nicols is shown in plate 3, upper. The flattened analcitesadhere to the quartz films which coat the- platy calcite and usuallyshow very numerous quartz spicules attached to their surface. Thisquartz was ground off of the crystal shown in the photograph. Opti-cally the analcite is biaxial and negative with large axial angle anda mean index of refraction of 1.486 The biaxial interference figureis extremely sharp and definite and this, together with the geometri-cal perfection of the division into sectors and the sharpness of the
14 PROCEEDINGS OF THE NATIONAL MUSEUM. VOL. 64lamellae, makes it hard to accept the idea that the mineral owes itsnonisotropic character to anomalies explained by internal strain.A sample for analysis was selected carefully and its homogeneouscharacter and purity were established microscopically. The analysisgave the following composition and ratios:Analysis and ratios of analcite.
Constituent.
ART. 19. MORDENITE?ROSS AND SHANNON. If)and incloses mordenite and spherulitic quartz. No other mineral isdeposited on the analcite, and it is probably the youngest of theminerals represented in the Challis specimens.CHLORITE.About one-fourth of the total number of small pebblelike fillingsin the amygdaloid rock consist of a soft, foliated, brownish-greenchlorite. These chlorite fillings, which reach an extreme diameter of7 millimeters, are easily freed from the surrounding matrix and whencrushed yield a homogeneous sample for analysis. This chlorite doesnot occur in the same vesicles with any of the zeolitic minerals, butit is sparingly distributed in the silicified portions of the inclosingrock near the large zeolite-filled cavities. Under the microscope thechlorite is seen to consist of yellow-green irregular folia. These arebiaxial and optically negative ( ? ) with 2V near 15?. The indices ofrefraction, which are high for a chlorite are, ?= 1.615, i8= 1.637,7=1.638, 7 ? a! = 0.023. The mineral is strongly pleochroic with Xand Y olive-green and Z brownish-yellow. This material wasanalyzed yielding the following results and ratios:Analyses and ratios of brownish-green chlorite.
Constituent. Per cent. Ratios.
Si02-..AI2O3FezOsFeOMnOCaOMgOH2O+110?CH2O-110?CTotal.
30.7612.129.1222.761.24Trace.12.369.761.8099.92
.307j
.542\
.100/
0. 170 X 3
.176 X 1
.160 X 4
.160 X 4
1 00 X 31.03 X 1
.94 X 4
.94 X 4
This analysis yields as the formula, 4RO.R2O3.3SiO2.4H2O, withR0 = M20 : FeO = l : 1 and R203 = Fe203 AL0,= 1 : 2. This cannot be definitely assigned to any member of the chlorite group but ismost closely related to diabantite and delessite, chlorites occurringcharacteristically as amygdule fillings in basic igneous rocks. Thecontent of ferric iron is noteworthy, and this doubtless explains therelatively high index of refraction and birefringence of the mineral.In a recent description of diabantite from Nova Scotia, Walker ^ ob-tained from an analysis a formula slightly different from that ofThomas L. Walker. Univ. Toronto Geol. Series No. 14, p. 47, 1922.
16 PROCEEDINGS OF THE NATIONAL MUSEUM. VOL, 64Hawes, and points out that both his andHawes's minerals can beinterpreted as isomorphous mixtures of two compounds, namely(a) RO.RA-2SiO2.2H2O, and (6) 2RO.SiO2.H2O. This is distinctlynot true of the present analysis in its present form. However, if theferric iron be calculated back to ferrous iron and the water below110? be excluded, the formula obtained is 4Fe0.3MgO.Al203.5Si02.5H2O, which may be regarded as composed of Walker's (a) x 1 withRO = MgO plus (6) X 1 with RO = MgO plus (h) x 2 with RO = FeO.The analysis, recalculated on this basis, is given in column 1 below,while in column 2 is given the calculated composition of a mixtureof the above formula:
ABT. 19. MOEDENITE?^ROSS AND SHANNON. 17
18 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.64with the Challis argentine in appearance and like it has the calciteplates coated with thin reticulate layers of quartz. The Washingtonargentine is associated with laumontite.The large argentine masses clearly are later, in age, than both themordenite and the spherulitic quartz and the calcite is older thanthe reticulated quartz and the analcite, which occurs as flattenedcrystals between the calcite plates. In the specimen shown in plate1, lower, however, the calcite plates are embedded in the mordeniteand thin sections show that the calcite plates of this specimen notonly certainly are earlier than the mordenite, but that they are olderthan the early heulandite deposit which lines the cavity and coatsthe free plates of calcite on both sides. Elsewhere in the largemordenite specimens there are flat gashes, now lined with analcite,which are apparently casts left by the removal by solution of earlycalcite plates. There are obviously two generations of the calcitein the specimens, both of which have the unusual argentine habit.The habit of the calcite may be due to the fact that the solutionsfrom which it crystallized were highly saturated with silica and thecalcite plates and their surfacing films of quartz are more or lesscontemporaneous, the silica presumably having deposited in colloidalform. CONCLUSION.The assemblage of minerals here described occurs as fillings ofcavities, often of very irregular form and large size, in what is prob-ably a flow of rather glassy andesite with a porphyritic developmentof augite, poured out in a Miocene lake basin. The cavities withtheir fillings may be so closely spaced as to make up a large propor-tion of the bulk of the rock. In general, the augite phenocrysts ofthe andesite are quite unaltered and there is no evidence of extensivealteration of the groundmass, so that it is evident that the mineralsof the cavities were not derived by the breaking down of any of theconstituents of the immediately adjacent rock. So far as can bedetermined from a study of the specimens available the order ofdeposition has been as follows, beginning with the earliest mineraldeposited: Chlorites (diabantite, etc.), chalcedony, calcite (argen-tine), heulandite, mordenite, quartz (spherulitic), calcite (argentine),quartz (films), analcite.The deposition of both quartz and chalcedony was apparentlyrepeated or was continuous over a considerable period. Of theminerals, mordenite, a comparatively rare zeolite, is by far the mostabundant. The area deserves further study and may be expectedto produce other minerals of interest if carefully explored.
ABT. 19. MORDENITE?ROSS AND SHANNON. 19EXPLANATION OF PLATES.Plate 1.LTpper left. Geodal mass of cottony type of mordenite with chalcedonic shell.Upper right. Analcite crystals in cottony mordenite.Lower. Argentine calcite inclosed in compact fibrous mordenite. Note thinshell of heulandite which lines the cavity. The small black specks on or nearthe calcite plates are small dendrites of manganese oxide. All natural size.Plate 2.Upper. Small pure mass of compact, felted-fibrous mordenite filling gascavity in andesite.Center. Two specimens of amygdaloidal andesite containing amygdules ofmordenite, heulandite, and chlorite.Lower. Mordenite-filled cavity, all except the central part of the loose-textured mordenite being saturated with granular quartz. All natural size.Plate 3.Upper. Natural thin flattened analcite crystals from between plates ofargentine calcite. Photographed between crossed nicols to show anomalousbirefringence and division into sectors. X20.Right. Argentine (calcite) coated with quartz spicules and bearing smallflattened analcites. Natural size.Lower. Tufts of mordenite fibers attached to groups of quartz spicules pro-jecting from platy argentine calcite. Natural size.

U. S. NATIONAL MUSEUM PROCEEDINGS, VOL. 64. ART. 19 PL. I
/4
MORDENITE, ANALCITE, AND ARGENTINE,For explanation of plate see page 19.
U. S. NATIONAL MUSEUM PROCEEDINGS, VOL. 64, ART. 19 PL. 2
^W^-^- 't^h}
MoRDENiTE Masses and Amygdaloidal Rock, Challis, Idaho.For explanation of plate see page 19.
U.S. NATIONAL MUSEUM PROCEEDINGS. VOL. 64. ART. 19 PL. 3
4^^s?i.-
Analcite, Argentine Calcite. Mordenite. and Quartz.FOR EXPLANATION OF PLATE SEE PAGE 19.