Ufa off. SMITHSONIAN MISCELLANEOUS COLLECTIONSVOLUME 143, NUMBER 4 Cfjarles! B. anb JWarp "«Taux OTalcott&eaearcf) Jfunb COMPARISON OF TEKTITE SPECIMENSFROM EMPIRE, GEORGIA, ANDMARTHA'S VINEYARD,MASSACHUSETTS(With Six Plates) ByROY S. CLARKE, Jr.ANDMAXWELL K. CARRON (Publication 4465) CITY OF WASHINGTONPUBLISHED BY THE SMITHSONIAN INSTITUTIONAUGUST 24, 1961 SMITHSONIAN MISCELLANEOUS COLLECTIONSVOLUME 143, NUMBER 4 Cfjarles S. anb Jfflarp Uaux OTalcott&eseard) Jftmb COMPARISON OF TEKTITE SPECIMENSFROM EMPIRE, GEORGIA, ANDMARTHA'S VINEYARD,MASSACHUSETTS(With Six Plates) ByROY S. CLARKE, Jr.ANDMAXWELL K. CARRON (Publication 4465) CITY OF WASHINGTONPUBLISHED BY THE SMITHSONIAN INSTITUTIONAUGUST 24, 1961 PORT CITY PRESS, INC.BALTIMORE, MD., U. S. A. Cfjarlea 5©. anb illarp \Ta«x Maltott &esfearc|j JfunbCOMPARISON OF TEKTITE SPECIMENS FROMEMPIRE, GEORGIA, AND MARTHA'SVINEYARD, MASSACHUSETTS 1By ROY S. CLARKE, Jr.,* and MAXWELL K. CARRON 8(With Six Plates)INTRODUCTIONThe recent find of a tektite at Gay Head, Martha's Vineyard, Mass.,has been reported by Kaye, Schnetzler, and Chase (1961). This speci-men, representing a possible new occurrence of tektites, was gener-ously submitted by the finders to us for laboratory study. TheMartha's Vineyard tektite (USNM 2082) arrived when we werecompleting study of a tektite from Empire, Ga. (USNM 1396), whichhas been at the U. S. National Museum since 1938. A similarity be-tween these two specimens was immediately suggested by their closeagreement in color, density, and magnetic properties. Further studyof the Martha's Vineyard tektite established that a truly remarkablesimilarity does exist. This report presents new physical and chemicaldata and photographs for both of these specimens. The possible sig-nificance of the unexpected nature of these data and the ambiguousconclusion to which they lead are discussed. Either this new groupof tektites has much more uniform properties than would be expected,or else there is room to doubt their authenticity.ACKNOWLEDGMENTSThe authors are indebted to a number of their colleagues who havesupported and contributed to the studies presented here. Most ofthese workers are cited at appropriate places in the text. Frank E.Senftle, Irving Friedman, and E. C. T. Chao, of the U. S. GeologicalSurvey, should receive special mention, as they have had an active in-terest in this work since its inception. Their suggestions, criticisms,and experimental observations are included below. Paul D. Low- 1 Publication authorized by the Director, U. S. Geological Survey.2 U. S. National Museum, Smithsonian Institution.8 U. S. Geological Survey, Washington 25, D. C. SMITHSONIAN MISCELLANEOUS COLLECTIONS, VOL. 143, NO. 4 2 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I43 man, Jr., of the National Aeronautics and Space Administration, andE. P. Henderson, of the Smithsonian Institution, have generously-shared their knowledge of tektite specimens and literature with thesenior author in many helpful discussions.PREVIOUS WORKThe initial identification of tektites from Georgia was made by E. P.Henderson of the Smithsonian Institution's Division of Mineralogyand Petrology. Two specimens (USNM 1396) were submitted forexamination and an identification was made during 1938. Confirma-tion of Georgia as an area of tektite occurrence has been reportedby Barnes and Bruce (1959). Bruce (1959) has published a generaldiscussion of tektite finds in Georgia, and included in his paper arephotographs of several specimens. Cohen (1959) has discussedGeorgia tektites with particular reference to their similarity tomoldavites and bediasites. His paper includes a compilation of physi- cal properties and spectrochemical data. Senftle and Thorpe (1959)have measured the magnetic susceptibility and intensity of mag-netization for the Georgia tektite and for a number of other tek-tites, and have discussed the significance of these measurements.Reynolds (i960) and Gentner and Zahringer (i960) have meas-ured potassium-argon ages for the major tektite groups. These datashow that Georgia tektites and bediasites are of similar age, butthat moldavites are much younger. Stair (1955a, 1955b, 1956) haspublished the absorption spectra and a photomicrograph of this sameGeorgia specimen. He also gives a photograph of a second Empire,Ga., tektite. The measurements reported in the literature on Georgiatektites by all the workers cited above have been made on portions ofone specimen, USNM 1396.The only previous experimental work using material from theMartha's Vineyard tektite other than a chemical analysis reported byKaye et al. (1961) is that of Pinson and Schnetzler (i960). Theseauthors have determined rubidium and strontium contents and stron-tium isotope ratios.MORPHOLOGY AND INTERNAL STRUCTUREThe two tektite specimens with which we are particularly concernedhave both similarities and striking differences in gross morphology.Plate 1 is a direct-light photograph of (A) the Empire, Ga., and (B)the Martha's Vineyard, Mass., tektites. Plates 2 and 3 are photo-graphs of these specimens after ammonium chloride smoking to bring NO. 4 TEKTITE SPECIMENS—CLARKE AND CARRON 3 out surface features. All photographs of the Martha's Vineyard tek-tite show the complete object before removal of material for analysis.Over half of the Empire, Ga., specimen has been consumed in experi-mental studies, and plate I, A, shows the remaining portion of thisspecimen. Plate 2 shows the front and back surfaces of this specimenafter it had been cut to remove a slice for study. The distance betweenhalves approximates the material that has been removed.Plates 5 and 6 are previously unpublished photographs of otherGeorgia tektites. They were furnished to the authors by E. P. Hen-derson and are included here as background material. Plate 5, A andB, are photographs of a second Empire, Ga., specimen (also havingcatalog number USNM 1396), a complete individual that is preservedin the collection of the U. S. National Museum. A tektite from Plain-field, Ga., belonging to G. A. Bruce is shown in plate 5, C and D.Plate 6, B, shows a tektite found near Osierfield, Ga., lent by A. S.Furcron, of the Georgia Geological Survey. Dimensions of thesetektites are given in table 1. Table i.—Approximate she and weight of specimensLengthLength of perpendicular Maximumlongest axis to longest axis thickness WeightLocality cm. cm. cm. g. Illustrated in — Empire, Ga.a ~6.5 ^3.5 ~i.o >2S PI. 2Empire, Ga 3-3 2.7 1.4 13.4 PI. 5, A, BPlainfield, Ga 3.5 2.9 0.9 11.2 PI. 5, C, DOsierfield, Ga 4.7 4.4 0.6 17.8 PI. 6, BMartha's Vineyard,Mass 5.3 3-9 1.0 17-8 PI. 3 a Lengths given were estimated from photographs of cut specimen, and thickness was meas-ured on remaining portion of specimen. The most striking feature of the four Georgia tektites is theirdisklike shape ; three are nearly circular. These specimens are ratheruniformly covered with many shallow pits and grooves which producegenerally smooth surfaces and edges. The disk shapes and generalsurface features are suggestive of the moldavites. Disk shapes areknown among moldavite specimens but are rare among the other tek-tite groups (Suess, 1900; Barnes, 1940; Baker, 1959).The Martha's Vineyard specimen appears to be a sector of aroughly circular disk about 3 inches in diameter. The smooth fracturesurfaces on the sides of the specimen imply that it has been brokenfrom a parent mass after formation of its surface features. The deeplyserrated edge of the Martha's Vineyard specimen is different from 4 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I43 anything that has been observed on Georgia tektites,* and it is an un-common feature of tektites in general. The surface relief is also muchmore pronounced for this specimen. It has sharp ridges on the topand bottom surfaces and particularly on the serrated edge. Thesesharp, relatively unabraded features imply that the Martha's Vineyardtektite has not been transported far by normal geologic processes sub-sequent to sculpturing. An unsual feature of this specimen is that theedge pattern appears to be radial, while the surface pattern on the in-terior of the disk appears to be concentric (pi. 3, A).There is a remarkable similarity between the Martha's Vineyardspecimen and a photograph of a moldavite published by F. E. Suess(1900, pi. V, fig. 5b). Our specimen appears, at first glance, to be apart of this specimen studied long ago by Suess. However, this ap-parent duplication is due to the fact that Suess's photograph is en-larged. His figures 5a and 5c show this tektite at natural size. It isobvious that the Martha's Vineyard tektite must have come from aparent of greater diameter than Suess's specimen.It has been stated above that the Martha's Vineyard tektite is ap-parently a part of a larger disk-shaped object, probably 3 inches(7.6 cm.) in diameter. If this assumption is valid, the parent body ofthis specimen was larger than any disk-shaped tektite of which we areaware. The hypothetical parent tektite would have a diameter-to-thick-ness ratio of J.6, which is greater than that of any tektite known tous. Even if a 2-inch diameter is assumed, this tektite would still havea very high ratio, approximately 5. The Osierfield, Ga., tektite (pi. 6,B, and table 1), with a ratio value of 7, is the only other tektite weknow of in this range.The internal structure and inclusions in the Empire, Ga., andMartha's Vineyard specimens are shown in the accompanying photo-micrographs. Plate 4, A, is a photomicrograph taken with white trans-mitted light of a slice 0.25 cm. thick cut radially from the Martha'sVineyard tektite. Plate 4, B, is of the same area using plane polarizedlight, crossed nicols. Plate 4, C and D, are photographs of a slice0.07 cm. thick of the Empire, Ga., tektite. If allowance is made forthe differences in thickness between the two sections, the similarity inpattern and character of inclusions is apparent. Some of these inclu-sions are well outlined and are of lower index of refraction than thesurrounding glass. They show wavy extinction and have not beenpositively identified. Barnes (1940) has proposed that similar inclu-sions in bediasites are lechatelierite. Sparsely distributed small round 4 Bruce, G. A., personal communication, i960. no. 4 TEKTITE SPECIMENS—CLARKE AND CARRON and elongated bubble cavities are also present, appearing in the photo-micrographs as dark spots.Pronounced flow structure, or flow lines, indicative of inhomo-geneity within the glass, appears in both specimens. This structure isrevealed by variation in index of refraction resulting presumably from B Fig. i.—Flow structure diagram prepared from slice of (A) Empire, Ga., tektite, and (B)Martha's Vineyard, Mass., tektite. X 3- minor compositional differences (pi. 4, A and C). Strain is alsopresent in these glasses and is associated both with the flow structureand inclusions. This strain is evident from the anisotropism that isobserved in the sections with plane polarized light, crossed nicols(pi. 4, BandD).The flow structure of both specimens here studied in detail conformsquite well to the surface of the specimens. In figure 1 are given flowstructure drawings prepared from a projected image of the sections 6 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I43 used in making the photomicrographs (pi. 4). Figure 1, A, is of asection from the flat end of the Georgia tektite specimen (pi. 1, A).The slice from which figure 1, B, was prepared was taken several milli-meters in from the left broken edge of the Martha's Vineyard speci-men positioned as in plate 1, B, extending approximately two-thirdsof the way into the specimen and parallel to that edge. This relation-ship of surface to flow structure is unusual for tektites in general(Barnes, 1940; Baker, 1959).The present external surfaces of these specimens are essentiallysecondary features due largely to chemical etching. Indeterminant fac-tors such as the original shape of the specimen, the susceptibility of itsvarious parts to chemical attack, the nature of the chemical environ-ment and the time through which it has acted, and mechanical effects,combined to produce the present surface features of these tektites.The main surface features, pitting and grooving, have no obviousrelation to the internal structure of the material. Tektite surface pitsare sometimes referred to as bubble cavities, but it is unlikely thatbubbles within the glass were responsible for the pitting on thetektites we studied. It has been mentioned above that the bubblespresent in the sections were small and sparsely distributed (pi. 4).Their concentration in the medium and their individual diameters areboth minute when compared to the surface pits.The internal flow structure, however, is related directly to delicate striae that are readily observable as a secondary surface feature onthese specimens. The striae frequently occur where the flow structureis truncated by the specimen surface and undoubtedly result fromslight differences in susceptibility to chemical attack. The left-handpiece of the Empire, Ga., tektite shown in plate 2, A, exhibits striationwhich is of particular interest because it indicates the extent to whichflow structure conforms to the surface of the specimen. The striaefollow the edge of the specimen and suggest that the flow structure pat-tern based on the section (fig. 1, A) holds in a general way for thecomplete specimen. The arrows in figure 1, A, indicate areas wherethe U-shaped striae on the surface of the specimen (fig. 1, A)terminate.Striae are obvious on the surfaces of the Martha's Vineyard tektite(pi. 3, A and B) and especially on the serrated edge (pi. 3, C). Theconcentric external pattern is consistent with the flow structure illus-trated from the section (fig. 1, B). The second Empire, Ga., tektite(pi. 5) is a striking example of surface expression of internal struc-ture. The more irregular pattern on this tektite probably indicatesa more contorted flow structure. Surface striation of this type NO. 4 TEKTITE SPECIMENS—CLARKE AND CARRON J is also present on the Plainfield (pi. 5, C and D) and Osierfield(pi. 6, B), Ga., specimens and can be seen in the photographs.Plate 6, A, is an enlargement of a small area of the surface of thePlainfield, Ga., specimen. It shows several features that are commonto all the specimens with which we are concerned and one feature thatis peculiar to this specimen. The latter is an apparently glassy mass,or protuberance, that projects from the bottom of a surface cavity(pi. 6, A; and slightly to left of center in pi. 5, C). This protuber-ance is firmly attached to the body of the specimen and apparentlyresulted from chemical attack on a volume of glass containing aninclusion or inhomogeneity of more resistant composition. No meas-urements of properties or composition of this protuberance werepossible as the owner desired to maintain the specimen intact.All these tektites show what appear to be several generations ofsurface pits, a feature particularly apparent on close examinationof plate 6, A. Around the top edge of the cavity containing theprotuberance there are four outlined depressions, apparently the rem-nants of previous pits that have grown together and been largely ob-literated by the younger central pit. The photograph also shows nu-merous examples of pits within pits, and pits overlapping pits. Aparticularly interesting pattern can be seen in the lower right-handcorner of plate 6, A. A raised, rather white area is surrounded byfive distinctly outlined grayish areas that seem to have been formedas a result of enlargement of pits. This feature and the glassy pro-tuberance described above provide direct evidence that the internalcomposition of the material has at least a limited control on the sur-face features that develop. A number of very small pits possibly couldhave resulted from bubbles within the glass, but it is impossible toidentify any of these from the photograph.To summarize: Study of the detailed morphology of these speci-mens supports the idea that chemical weathering, controlled to aslight extent by variations in composition of the material, is the mainagent responsible for the formation of these surface features. Wefind no evidence either in the gross shapes or on the surfaces of thesespecimens that suggests a history of aerodynamic shaping.PHYSICAL PROPERTIESA comparison of some of the physical properties of the two speci-mens is given in table 2. All the properties listed are remarkablysimilar.Density measurements were made by weighing the suspended speci-mens in air and in carbon tetrachloride of accurately known density 8 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I43 at the temperature of measurement. The resulting bulk density figuresof 2.330 for the Georgia tektite and 2.332 for the Martha's Vineyardtektite agree within the limit of error of the measurement (estimatedto be 0.002 g./cm. 3 ). These density figures are slightly lower thanthe lowest specific gravity figure (2.334) given by Barnes (1940)for bediasites and are in the middle of the range of density figures(2.303 to 2.367 g./cm. 3 ) he gives for moldavites. Conversion ofBarnes's specific-gravity values to densities is not possible because ofinsufficient data. The difference between our values and his lowest Table 2. — Comparison of physical properties of the Martha's Vineyard andGeorgia tektites Martha's VineyardGeorgia tektite tektiteProperty (USNM 1396) (USNM 2082)Color Light olive green Light olive greenWeight (g.) ii.4a i7-76bIndex of refraction i.485±0.003c i.4852±o.ooo4dDensity (g./cm.3 ) 2.330 2.332Magnetic susceptibility (e.m.u./g.) . . 3.6X10-*' 3.90X10""Magnetization o* o*Remaining portion of specimen.b Complete specimen.c Determined by I. Friedman, U. S. Geological Survey.d Bulk index determined by E. C. T. Chao, U. S. Geological Survey. • From Senftle and Thorpe (1959).f Determined by A. Thorpe, U. S. Geological Survey. bediasite value could be more apparent than real. The close agree-ment of density values for the two specimens under study combinedwith their chemical compositions (table 4) confirms the impressionobtained from transparent sections that bubble size and distributionin the two materials are the same.The index of refraction of both the Martha's Vineyard and Georgiatektites is 1.485. This value is slightly less than the smallest value(NNa=i-488) given by Barnes (1940) for bediasites and in the mid-dle of the range (NNa= 14798 to 1.4961) he gives for moldavites.Barnes (1940, pp. 522-523) has used the Gladstone and Dale rela-tionship to plot index of refraction and density data to show relation-ships of these data for the various tektite groups and other naturalglasses. The data for the Martha's Vineyard and Georgia tektitesgive a specific refractivity of 0.208, which falls in the moldavite areaof Barnes's plot, outside of the area where moldavites and bediasitesoverlap.Magnetic data for a number of tektites and other glasses have beenobtained and discussed by Senftle and Thorpe (1959). The magnetic NO. 4 TEKTITE SPECIMENS—CLARKE AND CARRON 9 susceptibility values depend both on the total amount of iron presentand the proportion of oxidized to reduced iron. The Martha's Vine-yard and Georgia tektites contain approximately the same total iron(table 4). The slightly higher proportion of oxidized iron in theMartha's Vineyard tektite is consistent with the slightly higher mag-netic susceptibility value observed for this specimen. The magneticsusceptibility values for the Georgia and Martha's Vineyard tektitesfall in the range between the highest moldavite value Q.oxio-6e.m.u./g.) and the lowest bediasite value (4.2 X io~6 e.m.u./g.) re-ported by Sen ftie and Thorpe (1959). The zero magnetization value,a value which is typical for tektites in general, is interpreted to meanessentially complete solution of iron in the tektite glass. These ob-servations are indicative of a history of high-temperature treatmentduring formation of the glass. ABSORPTION SPECTRAThe spectral transmission of a number of tektites, including theEmpire, Ga., specimen, in the ultraviolet, visible, and near infraredregions of the spectrum (300 to 5,000 millimicrons) has been reportedby Stair (1955a, 1955b, 1956). Cohen (1958) has given absorptionspectra for a number of tektites in the region 300 to 2,600 millimicrons.He points out that his curves and Stair's are in agreement for theregion they treat in common, and that the Empire, Ga., tektite curveagrees particularly well with that of moldavites. Cohen (1958) in-terprets these curves as being consistent with the high ferrous toferric iron ratio observed in chemical data on tektites (table 4), whileStair (1955a) tentatively interprets them as indicative of high ferriciron.A new determination of the absorption spectrum of the Empire,Ga., tektite, along with that of the Martha's Vineyard spectrum, isgiven in figure 2. These curves are directly comparable to those ofCohen and were obtained by using a Cary Model 14 recording spec-trophotometer. 5 Highly polished specimen slice surfaces were pre-pared, using o- to 2-micron diamond powder followed by magnesiumoxide. 6 The Georgia tektite slice used for the photomicrograph inplate 4, C and D, was further polished and used for the absorptionmeasurement. Masks with identical light transmission areas slightly 5 Dr. Walter Shropshire, Jr., Division of Radiation and Organisms, Smith-sonian Institution, did the instrumental work in obtaining these curves.6 Grover C. Moreland, Division of Mineralogy and Petrology, U. S. NationalMuseum, Smithsonian Institution, prepared the polished slices. IO SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I43 smaller than the smallest specimen were prepared for use in the sam-ple and reference beam of the spectrophotometer. A blank correctionwas determined by measuring the absorbance with the masks in posi-tion previous to mounting the specimens. 2.0- 1.5- 1.0- < 0.5- Absorption Spectra Martha's Vineyard Tektite Georgia Tektite 1000 1500Wavelength in Millimicrons 2500 Fig. 2.—Absorption spectra of the Empire, Ga., and Martha's Vineyard, Mass.,tektite specimen. The curves in figure 2 are essentially identical. The greater ab-sorption of the Martha's Vineyard slice can be attributed to samplethickness, suggesting that these two materials adhere to Lambert'slaw and have the same extinction coefficient.SPECTROGRAPHIC AND CHEMICAL ANALYSESSemiquantitative spectrographic analyses of both tektites are shownin table 3. Assuming that this type of analysis is within a factor ofonly 2 of the correct value, one can say that the analyses of both speci- no. 4 TEKTITE SPECIMENS—CLARKE AND CARRON II mens are essentially the same with the exception of boron, lead, beryl-lium, and yttrium. The high boron and beryllium contents are un-doubtedly due to contamination. The mortar in which the Martha'sVineyard tektite was ground had previously been used for grinding Table 3. — Semiquantitative spectrographic analysis ' of the Georgia andMartha's Vineyard tektites Martha's VineyardGeorgia tektite » tektite bElement percent percentSi M MAl 7 7Fe i-5 1.5Mg 0.3 0.3Ca 0.3 0.3Na 0.7 0.7K 1.5 1.5Ti 0.15 0.3Mn 0.07 0.07Ag 0.00007 0.00007B 0.003 0.015Ba 0.03 0.03Be 0.0003 0.003Co 0.0015 0.0015Cr 0.0007 0.0007Cu 0.0007 0.0007Ga 0.0003 0.0003Nb 0.0015 0.0007Ni 0.0015 0.0015Pb 0.00015 0.007Sc 0.0007 0.0007Sn 0.0007 0.0007Sr 0.007 0.007V 0.007 0.007Y 0.0015 0.007Yb 0.00015 0.00015Zr 0.015 0.015 « Figures are reported to the nearest number in the series 7, 3, l.S, 0.7, 0.3, 0.15, etc., inpercent. These numbers represent midpoints of group data on a geometric scale. Comparisonot this type of data with that obtained by quantitative methods shows that the assigned groupincludes the quantitative value about 60 percent of the time.6 Analyst: Helen W. Worthing, U. S. Geological Survey. hambergite, Be2(OH)B03 , and this probably accounts for the highvalues for these elements. The high lead and yttrium values could notbe accounted for. The following elements were looked for and notfound: As, Au, Bi, Cd, Ce, Dy, Er, Eu, Gd, Ge, Hf, Ho, In, Ir, La,Li, Lu, Mo, Nd, Os, Pd, Pr, Pt, Re, Rh, Ru, Sb, Sm, Ta, Tb, Th,Tl, Tm, W, Zn. Only elements to which the method is sensitive in 12 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 143 amounts of o.oi percent or less are included in this list. The analyti- cal procedure used has been described in detail by Waring and Annell(1953)-The chemical analyses of both specimens are essentially the same(table 4). The analysis reported here for the Martha's Vineyardtektite also agrees equally well with the independent analysis of adifferent part of the same specimen given by Kaye et al. (1961). Table 4. — Chemical analyses of the Georgia and Martha's Vineyard tektites Martha's VineyardGeorgia tektite * tektite bElemental oxide percent percentSiOa 80.54 80.6AlaOs 11.21 II.3FeaOa O.33 0.4FeO 2.40 2.2CaO 0.61 0.7MgO 0.65 0.7MnO 0.05 0.05NaaO 1.16 1.1KaO 2.38° 2.4HaO" None '.> - u +-> (Jr'SbCT3 O .2 o CQ *#. '> ' #v - • - p.. . ;•»