ACNE

J
© wy

U. S. NATIONAL MUSEUM BULLETIN 149, FRONTISPIECE

GEORGE PERKINS MERRILL
BORN, MAY 31, 1854. DIED, AUGUST 15, 1929

SMITHSONIAN INSTITUTION
UNITED STATES NATIONAL MUSEUM

Butuetin 149

COMPOSITION AND STRUCTURE
OF METEORITES

BY

GEORGE P. MERRILL
Head Curator of Geology, United States National Museum

4

APR 4 1939 a

%
e .
Zona muse

UNITED STATES
GOVERNMENT PRINTING OFFICE
WASHINGTON : 1930

For sale by the Superintendent of Documents, Washington,D.C. - - - - - - = Price 40 cents

ADVERTISEMENT

The scientific publications of the National Museum include two
series, known, respectively, as Proceedings and Bulletin.

The Proceedings, begun in 1878, is intended primarily as a medium
for the publication of original papers, based on the collections of the
National Museum, that set forth newly acquired facts in biology,
anthropology, and geology, with descriptions of new forms and revi-
sions of limited groups. Copiés of each paper, in pamphlet form, are
distributed as published to libraries and scientific organizations and
to specialists and others interested in the different subjects. The
dates. at which these separate papers are published are recorded in
the table of contents of each of the volumes.

The Bulletins, the first of which was issued in 1875, consist of
a series of separate publications comprising monographs of large
zoological groups and other general systematic treatises (occasionally
in several volumes), faunal works, reports of expeditions, catalogues.
of type-specimens, special collections, and other material of similar
nature. The majority of the volumes are octavo in size, but a
quarto size has been adopted in a few instances in which large plates
were regarded as indispensable. In the Bulletin series appear
volumes under the heading Contributions from the United States
National Herbarium, in octavo form, published by the National
Museum since 1902, which contain papers relating to the botanical
collections of the Museum.

The present work forms No. 149 of the Bulletin series.

ALEXANDER WETMORE,
Assistant Secretary, Smithsonian Institution.
WasuHinetTon, D. C., January 2, 1930.

Bag

FOREWORD

The paper here presented was written to form a part of a series to
be issued by the Smithsonian Institution, but on completion was
adjudged too technical for that particular publication. Rather than
attempt its popularization the author withdrew it, substituting in
its place a few pages of more easily comprehended generalities.
Since then, on considering the matter, and in view of the fact that
there is nothing in English covering the same ground, it has been
thought advisable to further elaborate and publish as here presented.
Students of meteorites in America are at this moment not too
abundant, and anything that will excite interest, or be of help, is
surely worthy of publication.

That many of the views expressed are the author’s own, and per-
haps not generally accepted, is recognized. It is thought, however,
that this is made sufficiently clear to avoid any misunderstanding.’

[Owing to the sudden death of Dr. George P. Merrill in Auburn,
Me., on August 15, 1929, while on his summer vacation, he never saw
the proof of this bulletin. This fact will serve as an explanation for
any scientific errors that may be found in the text, and conspicuously
the references to the plates, which would have been more ample, had
he seen the reproductions after they had been made.

A copy of his latest photograph has been inserted in this bulletin
as a frontispiece.—Ep1Tor.]

1 For a general treatise on the subject the reader is referred to Meteorites, by Dr. O. C. Farrington,
Chicago, 1915.

IIl

in

NMA Poti.

\
: ‘ : j :
;
ar ;
; } j
4 :
7 t vi
ce
¥ '
f
4 ; ve
7 A
5
A
J r«
7 ’
ee ed
’ a
: ‘
i
t
r Lata ,
i. /
t
‘ 7
2
" 1 LAG
i 4
‘ ; vit
iL) ‘th :
rok) A
' b = aie etx
; (wee tae 3
f ee
gains tae
{ ; j Yh... os PRUE 2
P Tore,
i a
j
' fy 3
: irk
v4
; ;
w 5

é

CONTENTS

Page
RORONU.G Gls eR ak. 4 F8F crt Mp te a A Mea ee RE CES 2 RUA BELA SEN ep Ill
mMinmentalbcomposition, of. meteonmtes..- 4 8 hk oe eee nee a 1
Mineral composition of meteorites... 22.248 son. 2 ee ee ee ek 1
Ghemical composition of meteorites...314 3-42-2422 404 ool ag3- ke 15
External and internal features of meteorites____....__._...-.---.-.---- 17
Chonarue,-its mature and origina =. 42 2b boot ek Re ee 29
Origin of other types of structure in meteoric stones______._----_------ 39
Medamaorphism: im meteorites: 25 4-421 wi lket |e ce a ee ee eS 40
Meteorites compared with terrestrial rocks._...........-.--.----------- 45
BUSS ecti tse yr ete ae i ad re Meal Sie eas ote cl fa as oe 47
Combustible meteorites____.____..--- pS a LE eh oe eR 48
Ghondritic structures.in terrestrial rocks. Sit 327522. ke ek 51
Methods of analyses of stony meteorites._.......2...--.~--.---.-2---- 57

LIST OF PLATES

Frontispiece. George Perkins Merrill.
Puate 1. External form of Allegan and Bath Furnace stones; and of Mazapil
iron.
2. (1) Carbon nodule in Canon Diablo meteoric iron; (2) Re-fused feld-
spar in Estherville meteoric stony iron.
3. (1) Maskelynite in Troup meteoric stone; (2) Merrillite in New Con-
cord meteoric stone.
4. (1) Fragmental twinned pyroxenes in Johnstown meteorite; (2) Frag-
mental pyroxenes in Estherville meteorite.
5. (1) Widmanstatten figures, enlarged; (2) Schreibersite in Arispe and
(lower) in Tombigbee irons.
6. (1) Meteoric iron from Dungannon, Va., showing partial granulation;
(2) The same greatly enlarged.
7. (Upper) Swathing kamacite in Admire pallasite; (lower) Octahedral
structure and Reichenbachian lines in Carleton iron.
8. (Upper) Neuman lines in Braunau iron; (lower) Granular structure in
Mejillones iron.
9. (Upper) Octahedral iron with silicate inclosures, Four Corners,
N. Mex.; (lower) Persimmon Creek, N. C., iron.

10. (Upper) Brecciated hexahedrite, Kendall County, Tex.; (lower)
Coarsest or kamacite octahedrite, Ainsworth, Nebr.

11. (1) Swollen kamacite in Mesa Verde iron; (2) Structure of Four Corners
iron, enlarged; (3) Quartz crystals in St. Marks stone; (lowest)
Chondrules in Sharps, Va., stone.

12. Grahamites from Morristown and Crab Orchard stony irons.

13. (Upper) Estherville stony iron; (lower) Same with inclosure.

14. Pallasite from Brenham, Kans.

15. (Upper) Section of the Admire pallasite; (lower) An iron-rich portion
of Brenham pallasite.

16. Microstructure of (1) El Nakhla stone and (2) of the Shergotty stone.

v

VE

PLATE 17.

18.

19.
20.

21.

22.

28.

29.

30.

31.

32.

CONTENTS

(Upper) Brecciated structure of the Cumberland Falls, and (lower),
St. Michel stones.

Chondrules and chondritic structure of Selma, Ala., and Cedar, Tex.,
stones.

(Upper) Microstructure of the Bluff and (lower) Estacado stones.

Microstructure of (1) Indarch stone, and (2) of Cullison stone, show-
ing a collar of metal about a fragmental chondrule.

(Upper) Slice of Anthony meteorite showing troilite with border of
metal, and (lower) Section of dark inclosure of Cumberland Falls
stone showing distribution of metal.

Chondrules in Barratta, Cullison, Elm Creek, Hessle, and Parnallee
stones.

(Upper) Olivine chondrule in Beaver Creek and Cullison stones;
(lower left) Enstatite chondrule in Cullison; (lower right) Twinned
enstatite chondrules in Parnallee stone.

. Chondrules in (1) Parnallee; (2) Tennasilm; and (8) Elm Creek

stones.

. (1 and 2) Chondrules and chondroidal forms from the Bjurbdle stone.

(3) Broken surface of Bjurbéle stone, about natural size.

. Toluca iron (1) before, and (2) after roasting.

(1) Chondrule in Parnallee stone with secondary glass border; (2)
Same showing effects of crushing; (3, 4, and 5) Chondrules from
Hendersonville, Troup, and Ensisheim stones showing effects of
crushing.

(1) Section through crust of Allegan stone; (2) Section through black
vein in Bluff stone.

(Upper) Faulted meteoric iron, New Baltimore, Pa.; (lower) Shat-
tered stone from Cedar, Fayette County, Tex.

(Upper) Broken surface of “‘ kugelngriinstein,’’ Schemnitz, Hungary,
showing chondroidal forms; (lower) Cut and polished surface of
the same.

(1) Microsection of phonolite tuff, Hegau, Germany (after Cushing) ;
(2) Chondroidal form in basaltic tuff (after Foye).

(1) Chondroidal forms in nepheline basalt, Hussenberges, Westphalia
(after Rinne); (2) Pseudochondritic forms of olivine in peridotite,
Raton, N. Mex.

COMPOSITION AND STRUCTURE OF METEORITES

By Grorce P. MerRRILu

Head Curator of Geology, United States National Museum

ELEMENTAL COMPOSITION OF METEORITES

A meteorite is a body of more than immediate mineralogical or
petrographical interest. It furnishes tangible evidence of the nature
of materials existing in remote regions of our solar system and per-
haps beyond, and affords, aside from the spectroscope, the only clue
to the matter of which celestial bodies are composed and the condi-
tions,under which they originated, a fact recognized by Humboldt
many years ago. It is, therefore, of interest to compare the materials
that are now coming from space, or have come within historic times.
with those forming the rocks of the earth’s crust.

The most abundant of the meteoric elements are, named in alpha
betical order: Aluminum, calcium, carbon, iron, magnesium, nickel,
oxygen, phosphorus, silicon, and sulphur. In smaller quantities are
found chlorine, chromium, cobalt, copper, hydrogen, iridium, lithium,
manganese, nitrogen, palladium, platinum, potassium, ruthenium,
sodium, titanium, and vanadium, probably also argon and helium.
The presence of antimony, arsenic, gold, lead, strontium, tin, and
zinc have from time to time been reported, but recent investigation
throws doubt upon the correctness of the determinations.' Tests
for fluorine have thus far yielded only negative results.

MINERAL COMPOSITION OF METEORITES

Though the elemental matter of meteorites may be the same as in
terrestrial rocks, the proportional amounts and forms of combina-
tion are at times radically different and of a nature to indicate that
they came about under conditions quite unlike those existing on the
earth to-day, and particularly so with reference to the presence of
free oxygen and moisture. It is for this reason in part that the study
of meteorites is so fascinating.

The following list comprises the meteoric minerals which are also
constituents of terrestrial rocks: Olivine; the orthorhombic pyroxenes
enstatite, bronzite, or hypersthene; the monoclinic pyroxenes diopside

1 Merrill, Geo. P., On the Minor Constituents of Meteorites. Amer. Journal of Science, vol. 35, 1913,
p. 509.
1

2 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

and augite; the plagioclase feldspars anorthite, labradorite, or oligo-
clase; the phosphate apatite; the oxides magnetite, chromite, and
quartz; the sulphides troilite and pyrrhotite; rarely the carbonate
breunnerite and various forms of carbon including graphite and
diamond. Those meteoric constituents found rarely if ever in terres-
trial rocks are the various alloys of nickel and iron, to which the
names kamacite, taenite, and plessite have been given; the nickel and
iron phosphide schreibersite; the iron and chromium sulphide dau-
breelite; the iron protochloride lawrencite; the calcium and titanium
or zirconium oxysulphide osbornite; the calcitum-sodium phosphate
merrillite; the iron and nickel carbide cohenite; the carbon silicide
moissanite; an isotropic mineral believed to be a re-fused plagioclase
and called maskelynite; and a form of silica, asmanite. These are
described in some detail, in alphabetical order below.

Apatite——The phosphoric acid reported in the numerous analyses
of meteoric stones has in times past been considered a constitu ent
of the mineral apatite. As a matter of fact, crystals of this mimeral
in a meteorite have been actually observed only by Berwerth, in the
stony portion of the Kodaikanal, India, siderolite. If occurring at
all, it is usually in the form of microscopic granules, though Lacroix
has recently described the mineral in the form of microscopic needles
in the metal of the Saint Sauveur stone. Late investigations have
shown that the prevalent phosphatic mineral is not a normal apatite,
but a new mineral—a calcium-sodium phosphate—differing in its
crystallographic and optical properties, and to which the name
merrillite has been applied.? (See also p. 7.)

Asmanite-—This name was proposed by Maskelyne ® for a mineral
consisting essentially of silica, occurring in the meteorite of Breiten-
bach, of which it composed nearly one-third of the siliceous portion.
The mineral, when pure, is colorless, with a specific gravity of 2.245,
a hardness of 5.5, and is rhombic in crystallization. It is commonly
believed to be identical with the tridymite of terrestrial rocks.

Breunnerite.—This is the name given by Haidinger to a ferriferous
variety of magnesium carbonate occurring in terrestrial rocks and
in a single instance in a meteoric stone, that of Orgueil, France. It
is the only instance known of a carbonate compound occurring as an
original constituent of meteorites. Its original meteoric nature is
perhaps questionable.

Carbon.—Carbon as the gas carbon monoxide (CO) or dioxide
(CO,), or in the amorphous and crystalline form of graphite has been
recognized as a constituent of certain meteorites, particularly meteoric
irons, for many years. Berzelius recognized a carbon compound in
the stone of Alais as early as 1838. Wé6bler and Cloez in 1839 found

2 Wherry, Edgar T., Amer. Mineralogist, vol. 2, no. 9, 1917.
3 Philos. Trans. Royal Soc., London, 1871, p. 361.

COMPOSITION AND STRUCTURE OF METEORITES 3

compounds resembling residue from terrestrial organic substances in
the meteoric stone of Cold Bokkeveld, while the French chemist
Berthelot thought to have extracted hydrocarbons conformable with
the petroleum series from the carbonaceous meteoric stone that fell
in Orgueil, France, in 1864.4 The American chemist, J. Lawrence
Smith, and others since have reported repeatedly the presence of
carbon in both the amorphous and crystallized forms of graphite in
numerous analyses of stone and iron meteorites. Amorphous carbon
in the form of coal black spherical masses associated with troilite is
a common constituent of meteoric irons, and is often surrounded by
a halo of schreibersite as in that of Canon Diablo, Arizona (fig. 1,
pl. 2). A nodule of this nature 3 centimeters in diameter was
analyzed by Dr. J. E. Whitfield with the following results: Carbon,
38.97; iron, 37.26; sulphur, 20.69; phosphorus, 0.24; recalculated this
gives carbon 38.97 per cent; troilite 56.89 per cent, with traces of
schreibersite.

Haidinger in 1846, described a cubic form of graphite in the
meteoric iron of Arva (Magura), Hungary, as pseudomorphic after
pyrite, but which Rose suggested was pseudomorphic after diamond.
Fletcher in 1899 gave the name cliftonite to a cubical form of carbon
found by him in form of minute crystals in the meteoric iron of
Youndegin, Australia, and later in the irons of Cosby Creek, Smith-
field, and Toluca. Though at first thought to be a distinct species,
this, too, is now commonly regarded as pseudomorphic after the dia-
mond. In 1888 Jerofeieff and Latschinoff found carbon with the
hardness and form of the diamond in the Novo-Urei, Russia, me-
teoric stone. In 1889 was found the first colorless material, thought
from its hardness and its burning into carbon dioxide (CO2) to be
diamond in the Arva iron. In 1891 George A. Koenig, of Philadel-
phia, found a black vitreous substance, of a hardness beyond sapphire
and believed to be diamond, in the meteoric iron of Canon Diablo.
Material from this source was subsequently examined by O. W.
Huntington and found to contain unmistakable minute, colorless,
octahedral crystals of diamond.’ The French chemist Moissan
found in this same iron carbon in the amorphous form, as graphite
and as black diamond, or carbonado. Moissanite, a carbon silicide,
perhaps identical with artificial carborundum, was also reported by
this chemist in the Canon Diablo iron.®

Chromite and magnetite.—The oxides of chromium and iron, or of
iron alone, are common constituents of terrestrial rocks as well as of

‘ Doubt as to the correctness of this and other tests for hydrocarbons in meteorites has recently been
expressed by P. E. Spielman. Nature, August 23, 1924,

* Proc. Amer. Acad., vol. 29, 1894, p. 204.

6 The fact that meteoric irons are commonly sawn by crushed carborundum raises a doubt as to the actual
meteoric source of this material.

4 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

meteorites, and need no further mention here other than that they
occur as small, usually microscopic disseminated crystals and crys-
talline grains. Whether or not chromium enters into the composition
of the pyroxenes, as in terrestrial rocks, has not, as yet, been
determined.

Cohenite—This mineral was first described by Weinschenk in
1889, having previously been mistaken for schreibersite, which it
closely resembles and with which it is very commonly associated.
It differs, however, in being soluble in copper ammonium chloride
and practically infusible. It occurs in blebs and tabular masses
belonging to the isometric system. Chemically it is an iron carbide,
of the formula (FeNiCo),;C. Actual analyses of material from the
Canon Diablo (1) and Magura (II) irons yielded the results given
below:

|
I} | Ii 2 Average
ARO MEE oo SU ESE PE Cw. Pate Pee es Be oh a DE eS oe ee BE vs Be 22k 91.31 89. 81 90. 56
DINERO Se ace cee NR oe ie ath a Rs ree en Pa ee ee Let 3. 08 2. 425
BMotinlte ys OE Ee CaS Dale ey tie 18 SENS OEE SR, 2 5 xe BRN COU hs 2 525 .69 -47
MORE QTD se sees ee oe ae le ea Rar Oe ia ee RS es cen mo ee See 6. 67 6. 42 6. 545

100. 00 | 100. 00 100. 00

1 Recalculated after deducting 4.68 per cent schreibersite.
2 Recalculated after deducting 0.65 per cent schreibersite.

Daubreelite—In 1876 J. Lawrence Smith gave this name to a
black, lustrous, highly crystalline material found by him associated
with the troilite in the meteoric irons of Coahuila, Mexico. Incom-
plete analyses made at the time showed 36.48 per cent of sulphur,
some 10 per cent of iron, and a little carbonaceous matter, the unde-
termined portion being chromium. The true composition he an-
nounced as being, probably, sulphur 37.62 per cent; chromium 62.38
per cent.’ Later he was able to isolate the material in larger quantity
and greater degree of purity from the Coahuila iron, and in 1878,*° he
published new analyses and descriptions showing the mineral to have
the probable composition: Sulphur, 44.29 per cent; chromium, 36.33
per cent; iron, 19.38 per cent; or the formula FeS Cr, S;. Actual
analyses, however, showed: Sulphur, 42.69 per cent; chromium,
35.91 per cent; iron, 20.10 per cent; total 98.70 per cent.

Feldspars and maskelynite—From what is known regarding terres-
trial basic igneous rocks, the feldspars of meteorites would naturally
be assumed to belong to the more basic varieties, as labradorite and
anorthite (fig. 2, pl. 2). Not many actual and complete analyses
are available owing to its rare occurrence and the consequent diffi-
culty of securing a sufficient quantity of material in a fair degree of
purity. Those quoted below show that in at least two instances the

7 Amer. Jour. Sci., vol. 12, 1876, p. 109.
8 Tdem, vol. 16, 1878, p. 270.

COMPOSITION AND STRUCTURE OF METEORITES 5

feldspar is oligoclase, a form characteristic of rocks of intermediate
acidity, as the diorites. The name maskelynite, it should be said,
was given by Tschermak ° to an isotropic, colorless mineral abundant
in the Shergotty meteorite, and commonly considered a re-fused feld-
spar. With this most workers agree, regarding it a product of meta-
morphism.” (See pl. 3, fig. 1, also pl. 16, fig. 2.) The mineralogist
Groth, on the other hand, was inclined to believe it to be a species
alhed to leucite. The feldspars are common constituents of mete-
orites of the basaltic types, such as that of Juvinas, in France,
where they occur in elongated polysynthetically twinned forms as in
terrestrial rocks. In the chondritic types they occur as scattered
granules occupying the interspaces of the olivines and enstatites,
and often quite lacking in crystal outlines or twinning bands, in which
case their satisfactory determination is a matter of difficulty. In
many meteorites of the chondritic type, and in most pallasites,
feldspars are wholly lacking.

Analyses of meteoric feldspars

Sources

Constituents

wate Sher- Johns-
Hvittis!| Hessle ? gotty 3 | town 4

63. 5 64. 97 56. 3 43. 72
22. 2 22. 06 25. 7 35. 40
4.0 3. O1 11.6 16. 28
9.2 9. 96 5.1 1. 60
MGs |e eee Te) LS e leases
100. 0 100. 00 100. 0 97. 00

1 Borgstrom, Bull. Comm. Geol. Finlande.

2 Lindstrom, Ofv. Kongl. Vet. Akad. Forhandl., 1869, p. 723.
3 Tschermak, Sitz. Akad. Wiss. Wien, vol. 65, 1872, p. 130.

4 Shannon, Am. Museum Novitates, Nov. 30, 1925.

Gases.—The fact that hydrogen was given off when the Lenarto,
Italy, meteoric iron was heated in a vacuum, was first noted by
Thomas Graham in 1867. Prof. J. W. Mallet, in 1872, found that
the meteoric iron of Augusta County, Va., under similar circum-
stances yielded not merely hydrogen but also nitrogen, carbon mon-
oxide (CO), and carbon dioxide (CO,). Prof. A. A. Wright, in 1875
and 1876, showed that in the stony meteorites the gas was chiefly
in the form of the dioxide, or carbonic acid (CO,) as it is commonly
called, while in the irons the monoxide (CO) and hydrogen prevailed.
Doubts, if such there may have been, concerning these first announce-
ments would seemingly have been completely eliminated by the
later work of R. T. Chamberlin |! from whose paper have been taken
bodily the tables following.

8 Sitz. Akad. Wiss. Wien, vol. 65, 1872, p. 127
10 Merrill, Geo. P., On Metamorphism in Meteorites, Bull. Geol. Soc. America, vol. 32, 1921, p. 407.
4 The Gases in Rocks. Publ. No 106, Carnegie Institution of Washington, 1908.

6 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

Gases in stony meteorites

Meteorite H2S | CO: Co CH, He N2 Total Analyst
|
2 x z a —| toe |
Guersey nO nioes..----2 a2 bentseawecee 1.80 0.13 0. 06 0. 95 0. 05 2.99 | Wright.
Bultuske Polandece 442-2 c. 2 82 seeseee 1. 06 . 06 - 06 . 52 . 04 1.74 Do.
Parnaleesindigen so eon o. oe z 2.13 . 04 .05 . 36 - 04 2. 62 Do.
Weston, Conn---_-- 2. 83 - 08 . 04 . 46 . 08 3.47 Do.
Iowa County, lowa . 88 Opnlesaaeee = 1. 45 712 2. 50 Do.
Cold Bokkeveld__- 23. 49 . 61 . 82 -10 Sleeps Do.
Dhurmsala, India_ - 1.59 . 03 .10 me, . 03 2.47 | Dewar.
Pultusic: Polandiv 22-8 3 sya? 235525 2 2. 34 .19 2% . 64 - 09 3. 53 Do.
VEO CS eee se rca een | eee a 1,25 . 07 . 09 . 45 -07 1.93 Do.
Greqenle!. a) Sasa, { wee \ 7.405) “Teaad|\, eer .33 | 57.77 Do.
Allegan: sich: = 2222) -pesuenss ars -21 -19 - Ol . 08 Tr -49 | Chamberlin.
Estacado,’ Tex. = 22.2227 Pee ar. 24 25 - 03 31 -O1 . 84 Do.
Average of 12 analyses...-| 4.00 One . 24 . 20 . 50 - 09 8. 80 |

Gases in iron meteorites

Meteorite H2S COs CO CH, He N2 Total Analyst

GON ALLO! 2 ee ae ea a eae 0.13 O00s eae 2. 44 0. 28 2.85 | Graham.
Aousta County, Vaencenecusen|seneee es .3l D2 ees 1.14 ol 3.17 | Mallet.
Tazewell County, Tenne 2 222|. 3228S . 46 Iolo 1.35 . 05 3.17 | Wright.
Shingle springs: Calif. Sess eb Le | 13 23L:2) | eee . 67 05 .97 Do.
Cross bim bers ex. 28 oo) eee oul: EALO S| Seto 300) peso eee 1.29 Do.
Dickson) County; flexes 22s cle eo oe .29 SO 4s| Se tee RS Te eee ins 2. 20 Do.
AV PEL UIN DALY er ee een | ee eb O2ei S191) | Seems 8. 57 ion) cavalo Do.
Cranbourne, Australia.-.....--|_--.-.-- | - 04 1.13 0. 16 1. 63 . 63 3.59 | Flight.
Romton, Shropshire ==) 22-22 sone eo | .33 Adi | eee 4. 96 - 62 6. 38 Do.
Nolucas Mexico-—. =). 2 eee 32 Tr. eZ 1. 32 - 04 non - 10 1.85 | Chamberlin.

PAV OAL ett Se reais vie anal ey Pas, . 78 3. 80 . 02 2. 36 . 30 7. 26

Average omitting Arva |_-..-_-- 21 . 67 - 02 1. 67 . 24 2. 83

meteorite.

Omitting the high content of sulphur dioxide of the Orgueil stone,
which is obviously abnormal, and presumably due to the weathering
of the troilite, the total average volume of gas from the stony mete-
orites is 4.8 times the volume of the material used. In hke manner
the amount of CO, in the Arva iron is abnormal and probably due to
oxidation. An average in the irons of 2.83 is therefore considered
correct.

Sundry attempts at the determination of the radioactive properties
of meteorites have been made by use of photographic plates, but
with results by no means satisfactory. Strutt,” working by what is
known as the Emanation method, was the first to demonstrate the
presence of radium in determinable quantity in the stone of Dhurm-
sala. Later Messrs. T. T. Quirke and L. Finkelstein * examined a
considerable number of stones and have shown that “the average
stony meteorite is considerably less radioactive than the average
igneous rock, probably less than one-fourth as radioactive as an
average granite, and that the metallic meteorites are almost free
from radioactivity.” Sixteen meteoric stones were found to have an
average radioactivity of 7.39 by 10-% gram of radium to a gram of

12 Proc. Roy. Soc., vol. 77, March, 1916, p. 480.
13 Amer. Journ. Sci., vol. 44, 1917, pp. 237-242.

COMPOSITION AND STRUCTURE OF METEORITES fi

material. Two stony irons had an average of 6.88 by 10-* gram,
while of the seven irons examined but two were sufficiently active
for determination, the Toluca iron yielding 2.13 by 10~" grams, and
that of Coahuila 7.69 by 10-”.

Lawrencite—Ferrous chloride. The exudation of drops of ferrous
chloride from freshly cut or broken surfaces of meteoric iron was
early noted, but it was not until 1855 that J. Lawrence Smith found
the material in the condition of a soft solid of a green-brown color in
the meteoric iron of Tazewell County, Tenn.'* In 1877 he also
noted the occurrence of the substance in the iron of Rockingham
County, N.C. In this same year Daubree thought to note its occur-
rence in the terrestrial iron of Ovifak, Greenland,’® and proposed for
it the name lawrencite in honor of its first discoverer. The material
liquefies on exposure to the atmosphere, the iron passing over quickly
to the condition of sesquioxide. It is to this mineral that is due the
“sweating’’ and rapid disintegration of so many irons, and causes
the stone meteorites to become rust-brown or freckled with rust-
colored spots.

Merrillite—This mineral, as a common but minor constituent of
stony meteorites, was first noted and described by Merrill in 1915 ”
and the name proposed by Dr. E. T. Wherry in 1917.'8 Subsequent
more detailed investigations by Larsen and Shannon * have shown
it to be an entirely new compound of calcium, sodium, and phos-
phorus with the formula 3CaO, Na,O, P2Os.

Inasmuch as a description of this mineral is not to be found in the
literature in general, it may be given here in full. Occurrence spo-
radic, without crystal form, colorless and very brittle; cleavage for
the most part lacking, though sometimes imperfect and interrupted
showing angles of 60° and 120°. Optically uniaxial and negative;
birefringence weak, less than 0.005; indices of refraction w»=1.626,
e= 1.620; specific gravity 3.10. (See pl. 3, fig. 2.)

Metallic constituents; nickel-iron alloys ——These are essentially the
same in all meteorites. They occur in varying proportions from a
fraction of 1 per cent, as in the Bishopville stone, to upward of 90
per cent, as in the so-called iron varieties. In the stones the form
is that of disconnected drops or stringers; in the pallasites that of a
more or less disconnected mesh or sponge enfolding silicate minerals;
and in the metallic forms constituting nearly the entire mass.

14 Amer. Journ. Sci., vol. 19, 1855, p. 154.

18 Tdem, vol. 13, 1877, p. 214.

16 Compt. Rend., vol. 84, 1877, p. 66.

17 Proc. Nat. Acad. Sci., vol. 1, 1915, p. 302.

18 Amer. Mineralogist, vol. 2, No. 9, 1917, p. 119.

19 Amer. Journ. Sci., vol. 9, March, 1925, pp. 250-260.

2# Wrongly given as biaxial and positive in first publication. Amer. Journ. Sci., vol. 43, 1917, p. 324.

8 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

Etching by means of a weak acid, the polished surface of a meteoric
iron will in the majority of cases, as already noted, give rise to an
interesting series of markings known under the name of Widman-
statten figures, after a German chemist who first brought them to
public notice. They are due to the unequal solubility of the three
alloys of iron and nickel which make up the mass of the material.
Two of these alloys occur in the form of thin plates and are known
by the terms kamacite and taenite. A third alloy, or properly a
eutectic known as plessite, fills the space formed by the intersection
of these plates (pl. 5, fig. 1, a, b, and c). The composition of these,
as thus far determined, is somewhat variable owing to the difficulty
of separating them one from another, and it is considered probable
that the so-called plessite is but a mixture or intergrowth of the other
two. Davison gives the composition of the two first named as de-
termined on separations made from the Welland, Canada, iron as
follows:

Constituents

Kamacite}] Taenite

|
| Per cent | Per cent
74.78

7

| 6.69 24. 32
"20 "33

‘02 50

100. 00 99. 93

In both kamacite and taenite there are variations in the propor-
tionate amounts of iron and nickel, ranging in the first instance from
thirteen to eighteen to one, and in the second from one to seven to one.
According to Borgstrom,” who has given the most recent summary,
the kamacite of the octahedrites contains about 7 per cent Ni+Co;
the taenite about 38 per cent, and the plessite 14 per cent.

So far as may be judged from chemical analyses there is no essential
difference in the metalliferous portions of the stony meteorites and
those which are all metal. This is brought out in the selected series
of analyses tabulated below.”

That in. the oxidation of a meteoric iron the first product is not
limonite, but a highly lustrous material which crushes down readily
to a fine brown magnetic powder, was first noted by the present
writer.”

21 Bull. Soe. Geol. of Finland, vol. 45, 1925.

22 The frequent reported occurrences of platinum and other rare elements in meteorites and particularly
in meteoric irons led the present writer to undertake a series of investigations to determine their correct-
ness. Platinum in traces was found to be a matter of common occurrence; ruthenium and iridium
occurred less commonly. No confirmation could be discovered to the reported occurrences of gold, tin,
antimony, etc. See particularly A Meteoric Iron from Owens Valley, Calif. Mem. National Academy
of Sciences, vol. 19, 1922.

23 Shannon, E. V., The Oxidation of Meteoric Iron, Proc. U. S. National Museum, vol. 72, Art. 21, Oct.
1927.

COMPOSITION AND STRUCTURE OF METEORITES

Se SS ee 000600 Se

| ch 66

ae eae ee Siecgeecca|e uae |) aS SOE TcO serena: | 816 FeCOR ESS ae Se ee ee Se Os eIOA YY,
Sap re ee se fee ee Sales Se ee See ee aa #0" | 16° ZO'ZL | 80 '98 |---~ 777 e}Apuoy— oy1TnIeydg |-~~~~"--- BpuoaB AA
= EE pe en = ee oe gee en eee at SS Sac “OOBIT, | ‘SOB, 8Z° lLE°L CP I6 |~~- 777 84tpuoyo Oul[[BISAID apm Nihon g(eya) TOUS A
See eee |g ee [ele alin (as Beeelicae cee 8g" eg '8 €9 06 |-77 777777 -e}LApUOYo OyTINJeydg |~~~ ~~~" MOTAUTVT
ae a |e es [te ame eS ce = ed ee we 80 'T Z10° ‘O0BLL Cr 9€ ZI | 110 ‘48 | --~ e}Apuoyo OJBIpouTieJUyT | ~~~ ~~ plOoTOL) MON
oer aes eae ge (Cor age Se S| ee al ee “OOBIYT, | ‘O0B1, cl° cc’L 98 16 a eno T LUO, OuT[[BISAID QS eS ee MOTOTAL
Petree s a= Saba clo seen Sle eae FO all aa ee GO t) a tee 8I° 92 8 PROG Net a5 OIpUOYyD SNOsDVUOGIB) | qoiepuy
R3010) de 91‘0 | 6210 80'0 | e08%1L | 20°0 +0" 10S ° 102 ‘6 (OU SBSieee wepnaed eyIpuoyo oyTNIeqds | ~~- wOsTTTN)
See eee oe [ee a eee eae Real Se eae eee 9F0 0 990 18 °L GPIB |-777 77 OP ApMOY otpnseydsg -|--- ~~ > UeBOTTV

re ee | as ee ee a ae Oe Se ee SS SS
SALINOALAW ANOLS NI TVLAW
P66

a ge lee a ag [eC ga ee eee | Iz" 220° S19" CONOTa OGESS lee ee oe SS eles aon:
Se | ae ees le as seal pee ee jeaeAet ‘a0VLL | 062° PRE B) | OCPA0G Ic” a. ae eaqaouepig ooo poequeqerg
Se a ege pS alos ee Ge eS |e ee ae jecae se Selmer rear | ety 1238 CCU TBH eam cea = gS ae 2 Ss OF WO DIBOSGI In pean weed AOUTIAT
ee ial a aad ee ee ae ea See RL eS i ME |e ea Sel ee ee 19° 241 °6 LL 68 ee ee ee OL NODISUBO NNT GS settee tee ee COLLIE
exrseeer spies Sep |e gle seg g erie | came Gea ZOL* (| “e0B1T, | 069° (0 Tape SLO 26 | ee ae ee ee eLOpIsOseyy | - aT[TAIONISH
eo le ees bie cae ee “9UON “OUON “OORL TL, “OOBLL 60 ‘ZI 6G est |. Sa 9}LIBPISOSO AT = ae psleyoig qvuly
eS Sa se |r ae a || a ee “OUON | ‘OUON | ‘90081, Gilat | 06 ‘9T ZOOTGs\-< eee. eg ce. ee OTISBIISd |: ae. . uolTyBIS apse
ee le ee “OUON, | ‘OUON | CR0° “OUON 09° Zo 6 06 “68 oe oo a ee oo OMS a Tae ~--->-"""ysiplouseIy
*s00B1} “BA G0 | ‘e0e17, | 000 | 8080 | TST‘0 | 882° 06g 0 | FOI’ 6r6" PROT PIe acisee4l\ec =. aa ee ee eUSe[[ed | ~~~ WOUIEA JUNOT
ee | oe ee lie gee ale gid | aes ROO ie ce cman eCORO 190 CSlOT | =6PsS8al = Sarees sae ee ONISUIIED |= pe ee UU

SALINOALAW NOULANOLS NI TVLEW

0 ne SE ee SS Se ee SS SS oe ee eee

*s90R1] Ul ‘Uy pus “Ay “Bd “4d
“ay “Ay pus 4d

1] “Ay pus “4d
“0°90 “sed "[osuy

bs Tr0°
SILO rates ee ee
GON =) |B Pew
CO iee |e aS
10° c0°
£0° c0°

SNOSULTIOOSIJT

28 '86
_—_—_————____
20° F10° ‘e0BIL | &0° 261° 610° 89° 618 00 ‘06
a2 £00 ° “omON | 200° COE” 920° GPS * 99 6 CT 68
og 10° ‘e0BIT, | 10° G26 ° G0" 60 'T &@ “ET | 99% “$8
FOR Rhee elf sees ZO" £0° GO ° 0g” L6°L LE 16
NO eas eee anne 20° A ZO" 86° OF 6 $2 °S8
EOUOINT Sie sae ati ee cee ‘9OBLL, Ge GOs e9° 108 IZ ‘16
‘OUON | 10° “OUON | 620° 991 © G10° $09 ° G66 PF LP 06
See c£0'0 | ‘9UON | 80° 6ST * “eoVlT, | 9° cee lL | Sth 86
COLO alee fe Sih ae oe 800 93 0 ¥0'0 99 0 89'S -| 69 °€6
ID 1s uN 8 d nO 70) IN oA

SALIVOULAN NOUT

Be eS eee O}AIpPSYyLIO SOUT
Per Sore apeyeBjoO Soul
SoS nas Capa aUIpoyBO ouly
pet ses epee UINIpeW
a eras eqIpeyejoO UMIpEW,
ae eyIpeyejoO WINTPe TW
por eee OJIIPEYLIDG OSivOD
SS a aa ae ee a}IpoyexeH

pur

ie en 9Sv.10A V

SS eat mae OT[LAAII0T
<-d g5 cae BIMOD
"aE ts gee = AUB O
oT Sa re a OUIqOUR y

eS IPBISYOM
see sepurin seseo

Be BAe o[qviq, woUR,)
Soa 5 ees JOATY XOFT

aueN

10 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

Oldhamite—This name was given by Story-Maskelyne, in 1862, to
a calcium sulphide (CaS) found by him in the meteorite of Busti
and described in detail in the Philosophical Transactions of the Royal
Society of London for 1870. The mineral is of a pale, chestnut brown
color when pure, though often covered on the outer surface by a
gypseous oxidation product. It occurs in the form of rounded
granules, with cleavages essentially rectangular, imbedded in the
pyroxenic constituents. It is optically isotropic and is considered to
belong to the cubic or isometric system. Specific gravity 2.58.
Boiled in water it is decomposed, yielding a bright yellow solution of
calcium polysulphide and an insoluble residue. It readily undergoes
alteration into gypsum, hydrous calcium sulphate, which probably
accounts for its apparent rarity. As gypsum is itself readily soluble,
its possible presence may be determined by boiling the powdered
stone in water and then testing the solution for calcium and sulphur.

Olivine —A magnesium and iron silicate of the formula (MgFe)
Si0,; relative proportions of magnesia and iron are, however, some-
what variable, as shown in the following analyses.”

Locality SiO2 MgO FeO
Me PIrAaSnOlATSKN SIDOnL As ee see = eee eee eee Ae ee oe oe a eee a 40. 24 47.41 11. 80
2. KAO an COUD byer RanSASie oo be tes er eee eee a ae eee eee 40. 70 48. 02 10. 79
a. prahineg issiqs <2 ee bee oe oe ob bo eee es ees 39. 61 48. 29 11. 88

ge Nc Yets aay: van Ole (eae Se i. ES IE a iS Se a ree ee ee es ee — See ee 36. 92 43. 16 17. 21

The mineral rarely occurs in good crystal form except in the
porphyritic chondrules, though in the pallasites of Krasnojarsk
and Lodran, blebs with determinable crystal faces have been found.
It is of all meteoric minerals perhaps the most abundant and wide-
spread, sometimes, as in those of Warrenton, Mo., and Chassigny,
France, composing a very large proportion (75 per cent) of the mass
of stone. It israrely, if ever, wholly absent, even the iron meteorites
showing in most cases included granules. It is also a common and
widespread constituent of terrestrial igneous rocks. Its appearance
is easily determinable by the unaided eye in the pallasites and its
meteoric characteristics shown in the illustrations of chondritic
structures (pls. 22 and 23). The statement frequently made—first
I believe by Daubree—to the effect that meteoric olivines differ from
their terrestrial prototype in containing no nickel, needs confirmation.
Indeed many careful analyses could be quoted showing the direct
opposite.

Osbornite—This name is also one of Maskelyne’s proposal. The
mineral occurs in golden yellow microscopic octahedra associated
with the oldhamite in the Busti meteorite only, so far as now known.

2 Tscherwinsky gives the average composition of the olivine of pallasites as follows: SiOs, 39.35; FeO,
14.39; MgO, 45.98; MnO, 0.08; NiO, 0.02; CaO, 0.03; NazO, 0.02; Al:O3, 0.09; Fe203, 0.06; Sp. Gr. 3.38.

COMPOSITION AND STRUCTURE OF METEORITES 11

Crystals are brittle and insoluble in acids, even resisting the fluxes
potassium and sodium carbonates. Its composition is uncertain,
but it is commonly regarded as a titanium or zirconium oxychloride.”

Pyroxenes——Pyroxene is common in meteorites in both ortho-
rhombic and monoclinic forms.

1. Orthorhombic pyroxenes: enstatite, bronzite, and hypersthene.
These minerals, next to the olivines, are the most.common of the
meteoric silicate minerals. The composition is somewhat variable,
owing to the varying proportions of iron and magnesia, as in the
olivines. A typical enstatite corresponds to the formula MegSiQ,,
but through the assumption of iron this passes over into the bronzite
variety (MgFe) SiO;. So far as known, the highly pleochroic hyper-
sthene rarely occurs in meteorites, though in at least one instance—
that of Shalka, India—the percentage of iron is fully as high as in the
strongly pleochroic terrestrial mineral. The name clino-enstatite
has been proposed by Wahl * for a monoclinic variety with a smaller
extinction angle on clinopinacoidal sections than normal monoclinic
pyroxenes, and which is characterized further by a marked tendency
toward polysnythetic twinning. The varying composition of ensta-
tite and bronzite from some of the best known meteorites is given
below:

Meteorite SiO2 | MgO |} FeO | Na2O | K20 | CaO | Al:O;
Rsisbapwillotte: * =<. <.s ra ss ese ens nese 59.97 | 39.34 0:40-|5s-s2222]ee<= AEE) SO ee
SSI eae eres ei sew ek 8 Ne cn ah oo ae 58.44 | 38. 94 1.18] 0.36] 0.33 1h68 |e
Mio@eiranise wmtes 9s 'y enna Sel ashe le ass BOs Bb | S285 4) D1) fever ee a] ae se .58 0.60
ISR GVINeI ay afte 50 A ee ee ee es SGEL05 SOLS yi gad eee ah RCS ee ee
enna eeeere shone SRA ee he HSNO5i |) weoedOul tae Gauleeeeen ae See 2.73 3.19
ais Osea ged he Ue 59.05 | 37.10 90 FESili 47 . 98 1.09
Mp ince ori os esa sles 59.92 | 38.00 |_-_-___- bigest bse De bbb tee, oe
VEGlirsom re: Ory tote Ss) et eae 57.80 | 39.22 AGUS | etree s [ine a I ee 2. 07
Ripleeteeseee crs. 2 ee ee mee ee 55.55 | 27.73 | 16.53 goths ries £005 eee as oe
BITTORSCE A eee eu so eel! Theat Nie, 57.49 | 25.78 | 10.59 15459 |0 22s 2.12 2. 08
Wohustowrplee ess yes Ui ti Teese eer | 52.16 | 27.60] 13. 39 | prrire. jeatiao 1.97} 123.9%
|

1 Smith, J. L., Amer. Journ. Sci., vol. 38, 1864, p. 225.

1 Maskelyne, Philos. Trans. Roy. Soc. London, vol. 160, 1870, p. 206.
3 Tschermak, Sitz. Akad. Wiss. Wien, vol. 61, 1870, p. 467.

4 Maskelyne, Philos. Trans. Roy. Soc. London, vol. 161, 1871, p. 359.
5 Rammelsburg, Monatsber. Akad. Berlin, 1870, p. 314.

6 Borgstrom, Bull. Comm. Geol. Finlande, No. 14, 1903.

7 Teclu, Rammelsburg’s Mineralchemie, 1875, p. 382.

8 Meunier, Ann. Chem. Phys., vol. 17, 1869, p. 12.

§ Rammelsburg, Monatsber. Akad. Berlin, 1870, p. 319.

10 Winkler, Cohen’s Meteoritenkunde, Heft 1, 1894, p. 281.

11 Shannon, Amer. Museum Novitates, Nov. 30, 1925.

12 Also 0.69 Cr2Q3 and 0.56 MnO.

As with olivine the mineral rarely occurs in good crystal form
excepting in the porphyritic chondrules. A more common form, as

noted later, is in that of radiating and cryptocrystalline ‘kugels.”’
(See pl. 22.)

25 The fact seems not generally recognized that the oldhamite and osbornite occur only in that portion
of the Busti stone which is plainly an inclusion, and further that osbornite has as yet not been identified
in any other stone, although oldhamite is comparatively common.

26 Tschermak’s Min. u. Pet. Mitteilungen, vol. 26, 1907.

59587—30——2

12 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

Berwerth has described 7” under the name of ‘‘Netzbronzit” a
fibrous form of bronzite occupying the interstices of porphyritic
olivines in the Zavid stone, the fibers standing at right angles to the
olivine surfaces. These he considers due to a partial fusion and
recrystallization of the fine bronzite material in a chondritic tuff.

2. Monoclinic pyroxenes: Augite, diopside, and diallage. These
forms of pyroxene are, on the whole, less common in meteorites than
are the orthorhombic forms, though it is possible that they are in
reality more abundant than is generally supposed, their close resem-
blance in all but optical properties (which, owing to the small size
and poorly developed crystallization, can not always readily be de-
termined) rendering a sure discrimination somewhat difficult. The
composition is, presumably, fully as variable as that of the enstatites,
but few actual analyses of pure materials have been made, owing to
the difficulty in separating them from the associated minerals.
Of the following analyses, No. I is by Maskelyne * and II by
Tschermak.”®

Source
Constituents ee
. er-
I Busti gotty

Silica (SiO2)_-____- ENE 25 PREP RD ey pat Be oP elit SS Aa ENN ey OS EY he 55. 49 52. 34
AN aa ina (CAs Og) So. Se PE RS Ne IS DD IN TON LER EOS et eS BES TEDL OS See alt. See eee . 25
Hereicoxide(HeiOs) ee ee eee eee, eee ee eye eee et ee OOM sae
MELLOWUSTOXI GEN GHGO) eo eS BS nee 8 ae Se TR se A ee ee eel | eee ee 23.19
Misenesia (MgO) i eck 2 eee ee 8 2 pie ay Ra pas Re ee age SU Nee ele 23. 33 14. 29
Ten (OA@) sees see's PS AL on Sy eS ee ee ee ee 19. 98 10. 49
Sod ay( Nida) pc ese ee ee ne rae Leena eRe te pets ee ee ee eee ee ee OD al cesar we
99. 90 100. 56

SPs Garros ee ss Daa E NS a ON a) eee BAER Ce TE Se ee RS EA ee | Se eee 3. 46)

As with other silicate constituents, the monoclinic pyroxenes are
but poorly developed crystallographically, with irregular and very
imperfect cleavage, and are nearly colorless, though the diopside is
often of a bright green in tinge as in the iron of Four Corners, N. Mex.
(See pl. 9.) They are often intergrown with enstatites and are rarely
appreciably pleochroic. The so-called peckhamite of J. L. Smith is
but an altered bronzite.*®

Radium.—See under gases, p. 6.

Schreibersite (and Rhabdite)—This mineral, first described and
named by Haidinger in 1847 as a constituent of the Magura iron and
since found one of the commonest of the accessory meteoric constit-
uents, is a phosphide of nickel iron and cobalt, corresponding ap-
proximately to the formula (FeNiCo);P. It occurs commonly in thin ,

27 Wissenschafr. Mittheil. aus Bosnianu. der Hercegovina, vol. 7, 1901.
28 Philosophical Transactions, vol. 160, 1870.

20 Sitz. des k. Akad. des Wiss., vol. 65, 1872.

40 See Merrill, Proc. U. S. Nat. Mus., vol. 58, 1920, p. 634.

COMPOSITION AND STRUCTURE OF METEORITES 13

angular plates of a tin white color, either lying parallel with the
taenite-kamacite plates, or in angular, jagged masses as in the
Tombigbee iron (see pl. 5, fig. 3), and sometimes in dendritic forms
as in the iron of Arispe (see fig. 2 of the same plate). In the pallasites
it occurs in thin plates lying between the olivines and metallic mesh.
It is brittle, magnetic, and difficulty soluble in acids, but fuses readily.
The name rhabdite was given by Rose in 1863 to a phosphide found
in the Braunau, Seelasgen, and Misteca irons, occurring in the form of
minute prisms of the tetragonal system and commonly considered
identical with schreibersite. The following analyses are selected
from a large number available:

I II Il IV
rcye orem Bh Sk Pg Tye Ee! ad 2 bb Oba Be oh 70. 07 61.78 50. 52 54. 43
pie kes tere eee eect ae ees SEW ee a ee eee 14. 57 21.93 33. 90 29. 36
MORON Nae wee a PT i eT ee Bae Br ee ae . 43 . 38 . 62 . 67
ROSEN aE eee ee ee ee ee SON EO Ae ee . 03 521 22 . 34
TELNET Graig Re a Ne es Oe Oe Sees ees Sole ee eee 15. 80 15. 70 15. 68 15. 45

100. 90 100. 00 100. 94 100. 25

I. Schreibersite from the Sao Juliao de Moreira iron. Fahrenhorst, analyst.
II. Schreibersite from the Kendall County iron. Scher, analyst.

III. Schreibersite from the Magura iron. Fahrenhorst, analyst.

IV, Schreibersite from the Cosby Creek iron. Fahrenhorst, anaylst.

While the formula (FeNiCo)3P for this mineral is that commonly
accepted, a considerable list of analyses by reliable chemists could be
quoted from which formulas ranging from the above to (FeNiCQ),P,
(FeNiCO),P, and (FeNiCo),P could be calculated. Cohen would
account for these discrepancies on the ground of impure or small
quantities of material utilized.

The mineral has not been found in well-developed crystals. Some
small distorted and imperfect forms obtained by the writer from
the Ruff’s Mountain iron were submitted to Dr. E. T. Wherry for
examination, who reported as follows:*!

The erystals average about one-half millimeter in diameter and are irregularly
distorted, some of the faces being cavernous; the system of crystallization is not
evident on superficial examination. The faces yield, however, fairly good re-
flections, the positions of which can be located in many cases within 5 to 10 min-
utes, unquestionable tetragonal symmetry being exhibited by the angular rela-
tions. The forms observed are: c(001), a(100), m(110), 0(111), x(862). In
addition there are rounded or poorly developed faces of other pyramids and
prisms. All of the forms are incomplete, but there is hardly sufficient regularity
in the suppression of faces to justify the assignment of the crystals to any par-
ticular hemihedral class.

Below are given the angles observed, which compare closely with
those measured on artificial crystals by Mallard, Hlawatsch, and
Spencer.

31 Amer. Mineralogist, vol. 2, 1917, pp. 80-81; vol. 3, 1918, p. 184.

14 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

Tasie 1.—Measured and calculated angles of tron phosphide

[Tetragonal, c=0.346 0.001]

|

|
| |
|

Angles measured

No. | Letter Symbol Crystals pier |
| | ¢ p ¢ p
i | |
Gea eee aed 001 | 1 TSR SY ° 00’ eS 0° 00°
eres a’ O10 | 2 5 0° 00° 90° oo’— | 0° 00"! 90° 00”
fe ae m 110 2 5| 45° 00'415’ | 90° 00’— 45° 00° | 90° 00°
greene ND) 0 lil 2 5| 45° 00'460’ | 26° 05/415 | 45° 00’ | 26° 05"
Mb sis x 362 1 2} 26° 007460’ | 49° 00’60’ | 26° 34’| 49° 15°

The fact that the mineral never occurs in good crystal form, lends
color to the suggestion made elsewhere * to the effect that the
schreibersite is not a true mineral species but a solid solution of
rhabdite in varying proportions of nickel-iron.

Silica.—Silica in the form of quartz was doubtfully identified by
Wohler in the meteoric stone of Hainholz and by Klein in that of St.
Marks. All doubts concerning the latter are set at rest by the find-
ing, by Merrill, of well-defined crystal particles as shown in Figure 3 on
Plate 11. Itis to be noted that the crystals are imbedded in the me-
tallic portion. Silica in the form of tridymite (asmanite) constitutes
some 8.527 per cent of the pallasite of Steinbach, and Berwerth has
described both quartz and tridymite occurring imbedded in the augite
of the stones (eukrites) of Juvinas, Stannern, Jonzac, and Peramiho.
This he suggests may be a form of pyrometamorphism due to heat
when the meteor stream of which they formed a part came near the
sun. Nearly every detailed analysis of metallic meteorites shows
traces of silica, which, if reliance can be placed on the examination of
the ‘‘insoluble”’ residues of these irons, occurs in the form of crystal-
line granules and distinct crystals. The manner in which these
residues are obtained, it must be confessed, throws a doubt on some
of the determinations, but the occurrence of the mineral noted above
imbedded in the metallic portion of the St. Marks stone at least
insures the possibility.

Troilite—This name was given by Haidinger * to a monosul-
phide of iron first found in nodular masses in the meteorite of Al-
bareto and since shown to be an almost universal constituent of
meteorites (see fig. 1, pl. 21). The theoretical composition as
demanded by the formula FeS, is iron (Fe) 63.64; sulphur (S) 36.36.
Actual analyses nearly always show traces of nickel and sometimes
copper. The mineral was named in honor of Domenico Troili, one
of the early enthusiastic defenders of the possibility of meteorite
falls. Meunier and some others have been inclined to regard the
mineral as identical with pyrrhotite. Rose suggested the possi-
bility that the sulphide in stony meteorites might be in the form of

32 Mem. Nat. Acad. Sci., vol. 19, 1922, p. 6, footnote.
33 Sitz. Akad. Wiss. Wien, vol. 47, 1863, p. 283.

COMPOSITION AND STRUCTURE OF METEORITES 15

pyrrhotite and in the metallic as troilite. The form assumed is
somewhat variable. In the irons it most commonly occurs in glob-
ular, spherical masses in sizes up to two or three centimeters or
more in diameter; sometimes in greatly elongated, conical forms, as
in the Santa Rosa iron. These may be wholly of sulphide, or of
sulphide admixed with amorphous carbon. Such nodules are often
surrounded by a shell of schreibersite. In the stony meteorites the
sulphide occurs in much smaller masses, rarely over two or three
millimeters in diameter, scattered through the ground or closely
associated with the metallic particles. In the stone of Holbrook,
Ariz., the sulphide occurs in such size and form as to be separable
from the matrix. Material thus obtained was found to be non-
magnetic and yielded on analysis: Iron, 63.62 per cent; sulphur,
36.50 per cent; with no nickel, cobalt, nor copper. The mineral in
this case is, therefore, evidently the monosulphide troilite. Ramsay
and Borgstrém arrived at similar conclusions in their investigation
of sulphide in the Bjurbole meteorite. The idea advanced by
Allen ** to the effect that the mineral is not to be considered a true
species but rather as the end member of the pyrrhotite series of iron-
sulphur compounds is doubtless correct. The name is, however,
retained as a convenient term to distinguish the monosulphide so
characteristic of meteorites.*®

CHEMICAL COMPOSITION OF METEORITES

Since meteorites are aggregates in varying proportions of the
minerals described it must necessarily follow that they are some-
what variable in ultimate chemical composition. The average of a
selected number of analyses of wholly metallic forms has been given
in the table on page 9.

Whether or no the all iron meteorites are to be considered inde-
pendent of the stony forms, and perhaps even from different sources,*®
is an open question, though the presence of transitional forms like the
mesosiderites and pallasites, especially those of the Brenham and
Estherville type (pls. 13, 14, and 15) suggests that they may be but
residual segregations in large masses from which all of the siliceous
portions have been eliminated. The size of some of these metallic
masses, as that of Cape York or Bacubirito, to be sure, puts a con-
siderable strain upon one’s imagination, but there is nothing im-
possible, or improbable about it, and those who argue in favor of a
metallic nucleus for our own earth should certainly find no difficulty
in accepting the idea.

84 Amer. Jour. Sci., vol. 33, 1912.

86 Until recently this form of iron sulphide was regarded as of purely meteoric origin. It has of late,
however, been discovered in considerable quantity in a copper mine in Del Norte County, Calif. See
Amer. Mineralogist, May 1922, p. 77.

% Pickering, it will be recalled (Popular Astronomy, No. 165), regarded the cometary origin of iron
meteorites as plausible, but not so the stones which he felt were of terrestrial origin.

16 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

As may be readily understood, the determination of the chemical
mass composition of a stony iron is a matter of difficulty. To over-
come this so far as relates to the pallasites, Professor Tscherwinsky *7
resorted to detailed measurements to determine the relative pro-
portions of the two essential minerals (olivine and metal), and
from the known composition of each, has calculated the average
bulk composition of the bodies of which they form the essential part.
The percentage by weight of the olivines in the pallasites examined
he gives as follows: Admire, 49.69; Ahumada, 46.22; Brahin, 37.18;
Brenham, 38.71; Eagle Station, 48.29; Finmarken, 61.22; Ilimaes,
43.68; Imilac, 49.25; Krasnojarsk, 52.19; Little Miami, 49.07;
Lipovsky, 46.49; Marjahlati, 49.88; Molong, 58.16; Mount Dyrring,
75.11; Mount Vernon, 58.94; Pavlodar, 62.49; South Bend, 41.40.
The average chemical composition of the pallasites as a whole cal-
culated as stated is:

Per cent

RSTIRCE CSL, este eee er an See tenet, MRD NU ras RARITY ty AURALY Sb SOE Ravana ne ae 20. 08
Herrous Omi Ge? GBS O) RAE pe Bo te MN Nye ay 20 bl UO Bol Ba A Pay eS 7. 42
Meanesing(iMp@) iis ech oN. duvey es Lone pees sue Vy poe Been tial bak ED 23. 41
Reo) 5! Si ete eA Wie Lu RN es Li Ba a a ee eee 43. 33
INGO ie la CNG) Raa ee re atk ace ae te islet oily a 2 carte Sy ger pg 7 easy ee a 4, 91
RE OREL O oeee eeer pre t nn PANY ROS h oe Ai 6 Le Oey aes or ea v2T
BU ple ge ke ey se eh TA OE A Se 99. 42

The stony meteorites are likewise variable, particularly in their
metal content. They are of a very basic nature as a rule—that is,
are low in silica. Those highest in this constituent are the achon-
dritic varieties consisting largely of pyroxenes and feldspars like that
of Cumberland Falls, Ky. The more basic are among the kugel
chondrites like those of Felix, Ala., and Jerome, Kans.

Below are given two analyses illustrative of extremes in basicity .
and acidity and in the table following an average of 63 analyses of
the highest grade obtainable.**

Cumber- | Cumber-
Felix, land Felix, land
Ala., Falls, Ala., Fails,
extreme Ky., extreme Ky.,
basic extreme | basic extreme
type acidic type acidic
type type
SiO ge ewes ee Sa eS 8 33. 57 BBs Hees tec Oe 2 eee ee ee a 0. 62 0. 157
Aid Ogee ey SERA 2 Ole <3 3. 24 ORR Hee Oe oe. D2 ee Le we 150
Gri@gy eee tee See Sr 80 AOGZT LS retece sea et Cee aR me eee ato 784
Ge Pe A A ED TS OED, 5. 61 J888e]] PAR = Lise AA ee EER oO 28 034
Vira sa Se ss ae | ee Ue Cane ate ea ee aes | Be ered 028
INPRO ER ORARS eee | Sue ak Be 36 059\)||R@ Sia LLY Ade eae tet aa 164
Come oo. eee ele ee eee 08 Q049 | | Went ses as ee ee ee . 16 167
OTE Ss ie a pal Cay Se Chey 01 003
TINT CG) earn ve tec veneer eee 1.01 ~1238 99. 78 101. 5380
BO. Sa doa as ase S 26. 22 2591641 Less: Os foriC1S,and Be 24) ee ee . 569
ROO) ee tae ee ee 5, 45 1. 586
MgO-_--__- hips aca ee ener 19. 74 3B: 7848 [oe ero Nae er Se Al. ener om 100. 961
Tiry CAE op OA SU as . 68 £12

37 Bull. Inst. Polytech. Don. Novocherkassk, 1918.
38 For methods of analysis, see p. 57.

COMPOSITION AND STRUCTURE OF METEORITES 17

Average composition of stony meteorites

Constituent: Per cent
Sian (SiO2) Lee eme ery Oe Mie ENA, oN, APR RR ay LE Oe Bla PERS ey ARNE 38. 41
Pear CHG RUG OY (Oy) eee Tea aN a eS CUR Dae Cae ES RRP. CE LN . 16

AP ORL ORONO s) ek Reels SNE a) el i ede eel et hak Deyn edit ty
IBCOMMMMO IG, (Ar Os) ie st ee ee ee Fea Lee RB

PAtmmmnrie (CAS Oa) Sten Sean Seg nne Mag oe MERU MON a el Si saan oN 2. 86
ETTI CHO RI GE's Alea O)g) aes aeiama ed ie mnire ns alee mea tt glia Bons Oulu eau cetera a2
Chromirctoxide? (Cry Ose eet ee ee ee ee CRAIN ARE ELAS AS EPS . 40
Mamata sOxLdeRGV.2Os) x eke ns. LUN CETL ode oP vag S oyie
Metallicntron: (He)! a Ace trp itan wens a) eee opt pk: 2 Goh a 2 deve ye 12Ss5
Metallic nickel. (Ni).22.222--4_-_-4 Bem 22 gS ATTA ah 1. 09
Metallvercobalt (Co) te taercse. tote ApS SL OUR AAD aae ere 2S Uy . 10
RETEROUSHO XTC ON MEO) paseo a aeae at cs Cita nhe aed ste A Er) heey ana petal, NAC eA 13. 60
INT CKEWOxTIges (NTO) S SA ee eee PAPER PO Spl MG TE ohare . 40
Cobalivoxides(CoO) Mak! Merten. 15. Navona ar pend OL eh . 06
MOEN EAO)s oly a ett vel ig Eee iagwege Reedy ep he han hued one haope 1. 88
DATUIIMLOXIGe (0) eee pun wine ee Cyc ete ee debe owe ae
[ag CSTE SST ES AAR (Sc 2 a a ON A a SS ae 23. 66
ManMeanous oxide (Min) aoa. O2e ee yas BAe SIS SE Se . 23
Stron vim” OXIGE (SO) skeet a weMMAee gM OF NL Eat og SAREE OE
Sodan(NasO) agree yes 297 ay ap ad pny PS ee I de LG Te ee . 82
Potash KO) gol. th. act. Sse ae a epi ewes) vetted cog ol ol he ler epee . 16
Nepuiiitor ste (GNGia sO) es Re hee Ls Sar oC ae ee ye eR akg dee SO i ee
Mem GLO TN Celts O))) Sent ee SR hana lain Nees 2 Os DAES Wey . 47
Phosphoric acid iUhs@; Ae ces aie aia ee re eer vee eee ee . o4
Bs ER PDERUAN SU pate 520 ec), BE ak ERI Ma A Ss RS AMD ited A PL 2 Be EPP 1. 89
Topper (Cup ie ye eat Lie See Aryeh week) . beepber Be iy. set. eke . 01
OEE oo oC) Se a a EN A a? Lo ie a Dok PE . 16
Winlormeg(C eek eK eee Ne eee Ns EEL cts Soe gitar eeoe eS) eee 50:
100. 00

EXTERNAL AND INTERNAL FEATURES OF METEORITES

There are certain external features characteristic of nearly all
meteorites, both stones and irons, which may well be considered
before entering upon a detailed discussion of internal structures.

Apparently there is no necessary limit to the size of meteorites,
though the largest actually known is the giant brought by Commander
Peary from Cape York, m Greenland, which weighed 37% tons.
The smallest stone, constituting all that is known of the fall is that
of Mulau, Austria, which weighed but 5 grams. It is now pretty
generally conceded that meteorites, of whatever type or size, are
fragments at the time they enter our atmosphere, and that as a rule
a further fragmentation and reduction in size takes place owing to
the atmospheric pressure induced by the enormous speed at which
the body may be traveling. Iron being tougher and more resistant
than stone, it would naturally follow that among the metallic forms
giants, if such there are, would prevail. This idea is fully borne out
in fact, and attention need but be called to the frequent occurrence
of stone ‘“‘showers”’ and their rarity among irons. The form and

18 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

size of meteorites are therefore both largely controlled by the speed
of travel. In either case sufficient heat is generated for the fusion
and burning away of the immediate surface, the fused material
being stripped off nearly as fast as formed, and but a thin coating of
material quickly cooled during the last moment of the flight remains,
producing the black crust so characteristic of stones or the thin layer
of oxide on the irons. The burning does not always take place evenly
over the entire surface, the foremost point of the body naturally
being most affected, the result being a smooth rounded nose, or
“‘brustseite,” from which radiate in all directions furrows or flutings
(piezoglyphs) sometimes developed to a remarkable degree of per-
fection, as in the stone of Bath Furnace, Ky. (pl. 1). Inequalities
in composition, as the presence of troilite nodules or other causes
are often productive of deep flutings in the irons, or holes which may
extend entirely through the mass, as in the Tucson iron or several
of those of Canon Diablo, Ariz. In this connection it should be said
that the flattened, sharply edged, and irregular form of the smaller
individuals of the Canon Diablo fall are not original, but due to a
weathering, which has been productive of the ‘‘shale balls” of Bar-
ringer and other writers. That this is correct is shown not merely
by their shape—edges unrounded by fusion—but by the occasional
finding of a mass with a still unaltered nucleus of metal lying like
an oyster in its shell. Once broken, such usually undergo a further
rapid oxidation and fall to pieces.

The black crust coating the surface of freshly fallen meteoric stones
is, as noted, due to the final hasty cooling of the fused material of
the meteorite. This is, on the immediate surface, a more or less
perfect glass, which is interspersed below the immediate suriace with
unfused silicate and metallic particles. It is rarely more than a few
millimeters thick, as in the case of the Allegan stone shown in Figure
1, Plate 1, and Figure 1, Plate 28. Should the stone hold the same
relative position in the air for any appreciable distance, the fused
material stripped from the nose may accumulate to a greater thick-
ness at the rear; it is, however, but a few millimeters thick at best.

Many meteorites are found traversed by a series of black thread-
like veins, which are probably but lines of fracture produced by the
shock of impact on entering the atmosphere, as noted later (pl. 29).

The meteorite differs from terrestrial rocks not more in external
characteristics and chemical and mineralogical composition than in
the manner in which its various constituents are arranged in relation
one to another. While there are certain features, as the presence
of metal, that are sufficiently pronounced to enable one at all expe-
rienced in such matters to decide almost at a glance upon the meteoric
nature of any object, even though not seen to fall, it is nevertheless
on a study of the thin sections and surfaces both polished and

COMPOSITION AND STRUCTURE OF METEORITES 19

etched that chief reliance must be placed. As with terrestrial rocks,
it would be very difficult to give by written description alone a clear
impression of some of the curious features as revealed by this method,
and the reader is advised to consult carefully the numerous illustra-
tions which are reproductions from photographs made in part through

the microscope.
1, ALL-METAL METEORITES: SIDERITES

The structural peculiarities of the all-metal meteorites have been
well described and beautifully illustrated photographically by Cohen
and Brezina in their work Die Struktur und Zusammensetzung der
Meteoreisen (Stuttgart, 1906). As, however, this work is quite inac-
cessible to the majority of students, space may well be given here to
a description of a few of the more typical forms.

Metallic meteorites, as has been said, are composed almost wholly
of iron with small and variable percentages of nickel and cobalt.
In the main, these metals are combined to form the two alloys named,
respectively, kamacite and taenite as already described. Each of
these alloys in the large majority of cases occurs in the form of thin
plates with intervening areas of a third alloy or properly eutectic
called plessite which fills the interstices. A characteristic feature
of the kamacite and taenite is a tendency to arrange themselves
in the form of thin plates lying parallel to the faces of a possible
octahedron. To reveal this structure clearly, as shown in Figure 1,
Plate 5 and Plate 7, it is necessary to polish a flat surface and etch
it with dilute acid.*® Owing to the differential solubility of the three
alloys mentioned, they will be acted upon unequally and stand out
each with its own relief. Such markings are called Widmanstitten
figures, after their discoverer. In thickness the plates vary from
the fraction of one to several millimeters, which fact forms the basis
of separation into fine octahedrites (Of), medium octahedrites (Om),
and coarse octahedrites (Og), etc. The kamacite presents several
varietal forms, dependent upon position and internal peculiarities
brought out by magnification. In the octahedral irons the bands are
often swollen in the middle and constricted at the ends, as shown in
Plate 7. In the pallasites a band of white kamacite a few millimeters
in diameter often incloses the silicates and is known as swathing or
‘““wickel’”? kamacite. (See upper figure of pl. 7.) A thin band of
taenite may or may not lie parallel with this and between it and the

39 The all-metal meteorites furnish a problem for the metallurgist which needs scarcely be touched upon
here, and the reader is referred to standard treatises on the subject. See especially Osmond and Cartaud
in the Metallurgist, vol. 4, 1891, and also the Comptes Rendus, vol. 137, 1903, p. 1057; Stahl u. Meteoreisen
by F. Berwerth, Metallurgie, vol. 4, 1907, p. 722, reprinted under the title Steel and Meteoric Iron in the
Iron and Steel Institute, September, 1907, Ein Naturlisches System der Eisen Meteoriten, Sitz d. Kais.
Akad. der Wiss, vol. 123, Abt. 1914, p. 1047; and finally Borgstrém’s paper: On the Composition of the Nickel
Iron Alloys and on Magnetic Lines on Sections of Meteoric Irons, Fennia, Helsingfors, vol. 45, No. 2,
1925, pp. 1-18. ®

40 Details of the etching process are given by Farrington in his Meteorites, pp. 127-130.

20 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

accompanying metal. At times the kamacite plates assume broad and
irregular forms as in the iron of Ainsworth and New Baltimore (lower
figure, pl. 10; upper figure, pl. 29), predominating over all other con-
stituents; in such octahedral structure is wholly undiscernible except
on large surfaces.*! In some instances taenite and plessite are almost
wholly lacking, the entire mass of the iron being composed of the
coarse kamacite granules. An interesting varietal phase of octahedral
structure is shown in Plate 6 from an iron found in Dungannon,
Va.,” some years ago. It will be noticed that this has undergone the
partial granulation described though traces of the original octahedral
structure are still discernible. The dark areas are of graphite with
metallic inclosures.

Not all irons are octahedral. In some the metal occurs in the
form of granules so fine as to escape easy notice, and thus to appear of
a noncrystalline structure or amorphous. These irons will often
show on etching certain faint parallel lines traversing the etched
surface, which are due, according to Neuman of Vienna, after whom
they are named, to the union of crystals in definite opposed relations
technically known as twinning, and in this case parallel with the faces
of acube. (Upper figure, pl. 8.) Still other irons are distinctly gran-
ular throughout, a structure which as shown later, may be secondary
and due to the action of heat (lower figure of the same plate). The
systematic regularity of arrangement of the taenite and kamacite
plates which form the chief constituents of an octahedral iron is often
interrupted by the presence in minor quantities of various accessory
minerals as cohenite, schreibersite, and troilite or carbon nodules
which last are as a rule distributed without order or, it may be, lying
parallel with the kamacite bands.

Nearly all irons carry varying, though minute quantities of olivine
and other silicates, quartz, and occasionally diamonds in quantities
so small as to be detected only in the residues left when the metal is
dissolved away by dilute acid. Graphite when present is left as a
black amorphous mud. The ubiquitous lawrencite makes its pres-
ence known through exudation of greenish drops on a polished surface
which quickly absorbs moisture and oxidizes to a rusty red color.
It is this mineral which brings about the rapid destruction of many
an iron meteorite and which perhaps explains the fact that no meteor-
ites are found in any but the most recent of formations.

Closely related to the wholly metallic forms are those of a somewhat
limited group represented to advantage in Plate 9 by a cross section
of a mass found near the bounding corners of Arizona, Colorado,
New Mexico, and Utah in the United States and hence known as the

41 Farrington’s tabulations of analyses seems to show that the texture varies with the nickel content, the
finest crystallization being found in irons richest in nickel. The ratio is, however, by no means constant.
Field Museum publication, No. 120, 1907.

42 Proc. U. S. Nat. Mus., vol. 62, 1928, art. 18.

COMPOSITION AND STRUCTURE OF METEORITES pil

Four Corners meteorite. This is described * as ‘“‘a granular mass
of octahedral iron with silicate inclosures”’ which in this case are
mainly pyroxenic. It is to be noted that each granule of the metal
has its own crystallographic orientation, the structure as a whole
resembling very much that of a coarsely crystalline aggregate of
calcite or feldspar. The pyroxenic portions (dark in the figure) are
finely granular and sometimes in the condition of a fine sand.. The
closest analogue to this peculiar iron is thought to be that of Copiapo
(Dheesa), Chile, as described by various authorities.

A second type which might well be mentioned here is that of Per-
simmon Creek in Cherokee County, N.C. (Lower figure, pl.9.) This
peculiar mass is described * as ‘‘a granular octahedrite”’ but might
better be designated as an agglomerate of masses of metal and troilite
(A and B in the plate), and occasional dark masses consisting of a
dense aggregate of graphite, troilite and olivine (C in the plate).
The metallic portions are composed of granules each with its own
crystallographic orientation and an octahedral structure as in the iron
of Four Corners.

2. STONY-IRON METEORITES: SIDEROLITES

The stony-iron meteorites are classified as (1) Lodranites, crystalline
granular aggregates of olivine and bronzite in a fine network of metal.
But one of this type is known—that of Lodran, India, which fell in
1868 and of which but 970 grams are known to exist. (2) Mesosider-
ites, or grahamites, aggregates of olivine, bronzite, plagioclase, and
augite, sometimes chondritic or with a crystalline structure in a con-
tinuous net of metal. (3) Siderophyres consisting of bronzite and
nickel-iron with accessory asmanite (tridymite). But asingle meteor-
ite of this type also is known—that of Steinbach (Breitenbach) Sax-
ony, which was not seen to fall but was found in 1751. (4) Pallasites
consisting of olivine in a continuous network or sponge of metal.
With the stony irons are also included by some authorities breccia-
like masses of nickel-iron with crystalline chondrites like that of
Copiapo and the octahedral iron with crystalline chondrites of
Netschievo.

Meteorites of the mesosiderite or grahamite group are well rep-
resented by the finds of Crab Orchard (Rockwood), Morristown,
Hainholz, and Vaca Muerta. The fall of Estherville though com-
monly here classed is really of a somewhat different type. The com-
mon structure, as shown in Plate 12, isthatof a dense net or sponge of
metal the interstices of which are filled by silicate minerals in the form
of small, single and angular particles and aggregates it may be two or
more centimeters in diameter. The metal is rarely segregated in
blebs a centimeter in diameter which yield Widmanstitten figures

43 Proc. Nat. Acad. of Sciences, vol. 10, 1924, p. 312.
4 Proc. U. S. Nat. Mus., vol. 27, 1904, p. 955.

22 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

when etched. The silicates in the Crab Orchard meteorite were ad-
judged on chemical grounds to be enstatite and anorthite. Vaca
Muerta and Hainholz are in so close agreement as to need no further
notice here.

Thin sections of the stony portions of the Morristown mesosiderite
show it to be holocrystalline granular, sometimes strongly cataclastic.
The latter structure is particularly conspicuous in those portions rich
in metallic iron, where the feldspars are often enclosed in the form of
sharply angular fragments in the iron or in its numerous embay-
ments. The appearance is not, however, that of a clastic rock, but
rather that of a crystalline mass which has been subjected to dynamic
agencies. The structure as a whole is quite irregular, and sometimes
porphyritic through the presence of large pyroxenic masses which may
be 5 to 8 mm. in diameter.

The groundmass of the stone is composed mainly of granules of
pyroxenes and plagioclase of such size as to render their determination
a matter of considerable ease, but which are interspersed with in-
numerable rounded and irregular granular forms so minute and so
lacking in crystal outlines as to obscure their true mineralogical nature.

The feldspars are in angular fragments showiag polysynthetic
twinning and numerous cavities and enclosures. Partial analyses on
a minute quantity indicate it to be anorthite. A very small amount
of olivine is present.

The remarkable meteorite of Estherville, Iowa, is, however, as noted
above, of a different type, though how far this difference is original,
and how far due to metamorphism remains yet to be shown. The
stone has been the subject of much discussion, a general summary of
which, up to 1915, is given by Farrington. As shown in Plate 13 it
consists of disconnected and irregular blebs of metal distributed
throughout the silicates and often with irregular cavities intervening.
The silicate components are enstatite, diallage, olivine and anorthite.
The enstatite occurs in two forms, a green and highly lustrous variety
and a yellow-brown opalescent filled with minute glass cavities. It is
to the last that Smith gave the name ‘“‘peckhamite.”’ Subsequent
studies 7 have shown this to be but an altered phase of the green
mineral, a change evidently brought about through the agency of heat.
Both the pyroxenes and olivine occur at times in globular pebblelike
forms. The groundmass is holocrystalline anorthite, often showing
signs of incipient fusions and other indications of metamorphism.
(See p. 40).

Pallasites Meteorites of this group—with the exception of
Brezina’s rékickites—differ in that the prevailing silicate (olivine)
occurs in such forms as to seemingly show its crystal development

45 Amer. Journ. Sci., vol. 34, 1887, p. 387.
4 Mem. Nat. Acad. Sciences, vol. 13.
47 Proc. U. S. Nat. Mus., vol. 58, pp. 363-370.

COMPOSITION AND STRUCTURE OF METEORITES a

out of a metallic magma; a condition difficult to realize. Both
Rose and Kokscharow have measured and determined crystal facets
on the olivines in the Krasnojarsk pallasite.“* That, however, the
olivines did not crystallize in all cases in the position they now occupy
is shown in the lower figure, Plate 13, where the silicate is in sharply
fragmental form, a condition thought by Brezina*® to be brought
about by movement in the plastic metal in which the olivines are
embedded.

Plate 14 is from a pallasite found some years ago at a locality
known as Brenham, Kans. The light, net-like portion is composed
of nickel-iron alloys identical in composition, so far as now ascer-
tained, with those of the all-metal meteorites. The dark areas are
silicate minerals—in this case olivine (peridot). The structure has
been compared, not inaptly, to that of a sponge in which the original
sponge material is metal, the silicates filing the meshes. Meteorites
of this type are somewhat rare, only about 20 now being known.
It is to be noted that the metal, wherever surfaces of sufficient size
are exposed, shows a tripartite structure and is never granular.
Further, that the kamacite bands often surround the olivines in a
form known as “‘swathing”’ kamacite (pl. 7) or white iron, on account
of the color and brilliant reflection. Between the kamacite and
plessite is often a thin band of taenite as in the all-metal forms.

3. STONY METEORITES, OR AEROLITES 5

The structure of many stony meteorites is of so confused and
heterogeneous a character as to be at times almost indescribable in
words, and one must refer to the illustrations. In their study the
same devices are employed as are commonly used for terrestrial
rocks. The figures shown in the accompanying plates are from pho-
tomicrographs made from the thin sections prepared in the customary
manner.

The one great difficulty in the determination and description of
meteoric minerals and structures lies in the imperfect crystal develop-
ment of the individual constituents, their shattered condition and
discoloration caused by oxidation of the lawrencite. This is par-
ticularly the case in the chondritic varieties which often present in
the section but a confused aggregate of polarizing points so charged
with secondary iron oxides as to render indeterminable any but the
two or three prevailing constituents, and uncertain the true nature

48 Tt is still a question if these faces may not be due to compression by interference in process of crystalli-
zation of a granular olivine aggregate before the introduction of the metal. See, Concerning the Origin
of the Metal in Meteorites, Proc. U. S. Nat. Mus., vol. 73, 1928, no. 2742, pp. 1-7. Also, Calcite Oolites
with Pentagonal and dodecahedral form, by E. V. Shannon, Journ. Washington Acad. of Sci., October 4,
1927.

42 Die Meteoriten Sammlungen, 1895. See also Merrill, Concerning the Origin of the Metal in Meteorites,
Proc. U. S. Nat. Mus., vol. 73, art. 21, 1928, p. 4.

50 A most excellent series of photomicrographs showing structures of meteoric stones, with descriptive
matter, is given in Tschermak’s Die Mikroskopische Beschaffenheit der Meteoriten, Stuttgart, 1885.
Unfortunately this work is not generally available.

24 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

of the original structure, whether crystalline, glassy, or fragmental.
This feature has caused a great diversion of opinion among students.
As will be observed, the present writer bases his conclusions as to
the original clastic (fragmental) nature of the chondritic varieties not
on structure alone, but on the presence in close association of one
and the same mineral under varietal forms of crystal development
such as are seemingly impossible products of direct cooling from a
molten magma.

Few meteoric stones, probably not over a score of those now known,
show the crystalline structure characteristic of terrestrial igneous
rocks, either basalts or peridotites. In the prevailing system of —
classification they are divided into two general groups. I, Calcium-
aluminum-rich stones nearly free of nickel iron and without chon-
drules, and II, magnesium-rich stones likewise nearly free of nickel
iron and nearly or completely free of chondrules. Group I is again
subdivided into (1) the angrites—of which but a single representative
is known, which consists mainly of a dark brownish augite and a
little olivine iron sulphide and with a crystalline granular structure;
(2) the eukrites, which consist essentially of augite and anorthite
with iron sulphide and also a crystalline granular structure like many
dolerites; (3) the shergottites of which there is also but a single
representative known, which consists of augite and the isotropic
feldspar maskelynite and a little magnetite, with likewise a crystal-
line granular structure; and (4) the howardites, consisting of augite,
anorthite, bronzite, and olivine in a tuffaceous ground with some-
times eukritic segregations.

Group IT is likewise subdivided on mineralogical and structural
grounds into (1) the bustites, of which but a single representative is
known, which consists essentially of diopside and bronzite with
smaller quantities of oldhamite, plagioclase, nickel iron, and osbornite,
with a nearly crystalline structure; (2) the chassignites consisting of
an iron-rich olivine and small quantities of chromite and with a
crystalline granular structure, of which type but a single stone is
known; (3) the chladnites, consisting of a crystalline granular aggre-
gate of a rhombic pyroxene; and (4) the amphoterites consisting of
olivine and bronzite with smaller quantities of sulphide and nickel
iron. The structure is sometimes granular and sometimes chondritic.

The microstructure of the eukrites and howardites has been
described in detail by various writers, including Tschermak, Wahl,
Berwerth, and others, all of which have been the subject of review
by Lacroix *! to whose work the reader is referred for details. The
Bereba eukrite is described as a breccia of doleritic fragments cemented
by recrystallized finely pulverulent material of the same mineral
nature. The essential constituents are pyroxene and anorthite with

51 Arch. du Museum D’Historie Naturelle, vol. 1, ser. 6, 1926.

COMPOSITION AND STRUCTURE OF METEORITES 2

secondary magnetite, pyrrhotite, and quartz in minor quantities.
The pyroxene of the unaltered fragments is of a brownish color and
filled with inclosures; that of the fine granular recrystallized ground
is of a yellowish color and free of inclosures. The howardites, of
which that of Teilleul is considered typical, is described as consti-
tuted almost exclusively of angular fragments of bronzite with fer-
ruginous inclusions; a ‘‘diopside-bronzite prive d’inclusions ferrugi-
nous’? and anorthite, with some chromite. The group would seem
really to consist of the same materials as the cataclastic, pulverulent
interstitial portions of the eukrites.

The stone of Juvinas consisting of a colorless anorthite and gray to
brownish augite shows under the microscope a holocrystalline struc-
ture not unlike that of many basic terrestrial rocks and like them
containing minute geode-like cavities. That of Shergotty (fig. 2,
pl. 16) differs in showing broad plates of brown augite with inter-
spaces occupied by a clear, colorless, and transparent feldspathic
mineral which is optically quite isotropic and to which the name
maskelynite has been applied. Stones of the Nakhla type (fig. 1,
pl. 16), of which but one example is known, consist of an even, granular
aggregate of green pyroxene (diopside) with olivine and occasionally
a plagioclase feldspar and a little magnetite. It might well pass for
a terrestrial pyroxenite. The chassignites, as represented by the
single occurrence in Chassigny, are fine crystalline granular aggregates
of olivine not greatly dissimilar to some terrestrial dunites. The
chladnites as represented by the Bishopville stone consist of a rather
coarse crystalline granular aggregate of nearly white enstatite with
small amounts of a plagioclase feldspar and occasional troilite granules.
The stone is remarkable for its poverty in metal, analyses showing less
than 1 per cent of this constituent.”

In many meteorites (both chondritic and otherwise) a brecciated
structure is plainly evident even to the unaided eye. This may be
due to a commingling of rock fragments from diverse sources, or from
crushing in mass, or perhaps from both, as shown in the meteoric
stone which fell in Kentucky in 1919 (pl. 17). Here are plainly
commingled two types of stone which have been compressed suffi-
ciently to produce incipient twinning in the enstatite particles after
a manner well known to petrologists. The manner in which the
metal is disseminated throughout this stone is of interest and will be
referred to later.

Other stones, like those of Supuhee, India, or Rose City, Mich..,
are plainly agglomerates of pebble-like bodies embedded in finer
material of the same mineral nature. In the St. Michel, Finland,

52 It is well to remark here that metal in the stony meteorites is present in appreciable quantities only in
those the fragmental and tuffaceous origin of which is readily apparent and is almost completely absent
in the crystalline or achondritic varieties.

26 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

stone described by Borgstrom * the fragments are more angular, the
stone partaking of the nature of a breccia (pl. 17).

Fully 90 per cent of the stony meteorites are characterized by the
presence of small spherical bodies embedded in a fragmental or crys-
talline ground of the same mineral nature. Mineralogically speaking,
the bodies are of pyroxene or olivine, rarely feldspar, though some-
times glassy and without the development of determinable mineral
species. In sizes they vary from too small to be visible to the unaided
eye to rarely a centimeter in diameter. These are called chondrules
from a Greek word meaning a grain and are of exceptional interest on
account of their unique form and probable origin. A discussion of
these features must, however, be left until later.

The chondritic meteorites, or chondrites, as they are called, form a
group quite variable in types of structure as shown in the several
plates here devoted to the subject. They are at times so friable as
to crumble easily in the thumb and fingers, or again are very hard
and tough, this feature being imparted by metamorphism. In the
first instance the chondrules may fall away entire. In the second they
may be so firmly embedded as to break with the matrix. All inter-
mediate stages occur. A description of the internal structure—the
manner in which the various minerals are disposed relative to one
another in a chondritic meteorite—is a matter of no small difficulty.
As a whole, the group may be said to consist of a heterogeneous aggre-
gate of minerals largely olivine and pyroxenes, mainly in a fragmental
condition, with varying amounts of interstitial metal and metallic
sulphides throughout which the chondrules are scattered in varying
proportions. It is only when we consider the crystalline chondrites
that a satisfactory verbal description can be given.

The chondrites are described by Wiilfing (Meteoriten in Samm-
lungen, pp. 449-454) as magnesia-rich stones consisting essentially of
olivine, bronzite, nickel iron, and iron sulphide, and with the excep-
tion of Novo-Urei plainly chondritic and tuffaceous. They are
divided into:

1. Howarditic chondrites, comprising transition members from the
howardites into the true chondrites.

2. White chondrites, yellowish white tuffaceous stones with chon-
drules for the mest part of the same color.

3. Intermediate chondrites, transition forms into the gray chondrites.

4. Gray chondrites, yellowish to blue gray tuffaceous stones with
variously colored chondrules firmly embedded in the ground mass.

5. Black chondrites, firm, dark gray to black stones in which the
color is due in part to carbonaceous matter and in part to pyrrhotite.
The chondrules are mostly of a lighter color. Meunier has shown
that in some instances the blackening has been produced by heat.

*8 Bull. Com. Geologique de Finland, No. 34, 1912.

COMPOSITION AND: STRUCTURE OF METEORITES 2d

6. Kugelchen chondrites. This large group consists of very numer-
ous well-developed chondrules embedded in (1) a tuffaceous ground
which is almost wholly of chondrules and (2) one in which the ground
is so loose and friable that the chondrules are readily broken away.
There are transitional forms into the next.

7. Crystalline chondrites, consisting of a crystalline ground in which
the hard chondrules are firmly embedded.

8. Carbonaceous chondrites, consisting of black ground due to car-
bonaceous matter, and carrying but little metal; hence noted for lack
of density. The Orvinites differ in showing a fluidal structure, the
crystal Tadjerites are part glassy ground and the Ureilite crystals are
black, sometimes chondritic, s)metimes granular masses consisting
mainly of olivine and showing transitional forms into the group
classed under nickel-iron with silicates.

Obviously such a system of classification permits of no sharp dis-
criminations. That of Prior, given below, has much to recommend
it in this regard. In this the chondrites are separated according to
the true character of their prevailing pyroxenic constituent as follows:

(a) Enstatite-chondrites.

(6) Bronzite-chondrites.

(c) Hypersthene-chondrites.

To the members of each of these groups are applied the qualifica-
tions according to color (white, intermediate, gray, black); structure
(crystalline, spherical, brecciated, veined); and composition (car-
bonaceous, etc.), used in the Tschermak-Brezina classification.

As illustrative of a pronounced type of a spherulitic or kugelchen
chondrite reference may be made to the stones of Allegan, Michigan;
Selma, Alabama; and Bjurbole, Finland, Figures 1 and 2, Plate 18,
also Plate 25. These stones are very friable, so friable indeed that
thin sections can be prepared only with difficulty. Under the
microscope they show in a marked degree the tuffaceous structure
characteristic of their class. Three or more types of chondrules are
not infrequently present in the same stone, (1) the ordinary enstatite
chondrulJe showing in a section a fan-shaped radiating structure; (2)
others composed of olivines, sometimes quite idiomorphic developed
in a black glass; and (3) dense structureless forms consisting evidently
of enstatite. These are all sharply differentiated from the ground and
break away from it readily. The groundmass is itself a confused
admixture of olivine and enstatite particles with interspersed metal,
metallic sulphide and chromic iron, the silicates being almost univer-
sally of a fragmental nature. Much of the interstitial material is so
fine and dustlike that it is practically impossible to determine its
nature in the section, but when isolated it is found to consist of
fresh and sharply angular splinters of olivine and other silicates.

59587—30——3

28 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

Many of the enstatite chondrules are beautifully perfect spheres
and others oval and elongated. They occur also in all stages of frag-
mentation as described in the numerous publications. Irregular
porphyritic forms occur; such the author regards as fragmental forms
due to trituration and designated as ‘“‘chondroids.’’ Many stones,
like those of Mezé Madaras, Russia; Selma, Alabama; and Cedar,
Texas (pl. 18), are made up in almost their entirety of chondrules and
chondritic fragments, the ‘“chondrulites’’ of Chamberlin.* (See fur-
ther under Chondrule, its Nature and Origin, p. 29.)

From stones of this type there is a constant gradation to the
crystalline chondrites of which the stones of Bluff, Estacado, or Hen-
dersonville may serve for purposes of illustration. These are compact,
dense stones showing a polished surface thickly studded with gray
chondrules of a few millimeters in diameter which are sharply differ-
entiated from the ground, sometimes breaking with it and sometimes
falling away leaving the surface studded with little saucer-shaped
pits. Some of the chondrules are of the radiating enstatite type;
others barred and porphyritic. The groundmass consists of a closely
intergrown aggregate of olivines and pyroxenes interspersed with
metallic particles and granules of iron sulphide. Under as high a
power as the thickness of the section will permit the use, the inter-
stitial matter polarizes faintly and shows a granular to fibrous struc-
ture. As a whole the structure is not that of minerals crystallizing
freely from a molten magma but is suggestive of a partial recrystalli-
zation of fine detrital material as seen in many metamorphic schists
(pl. 19).

Continual variations of this are found in the same and other types
of chondritic stones, but it is to be noted that as the stones partake
more and more of the nature of crystalline rocks the included chon-
drules grow less and less perfect, merging finally into the groundmass
until they quite disappear.

The dark color of the so-called black chondrites as shown by
Meunier * and subsequently by the present writer * is due mainly
to heating though it may be in some cases in part to the presence of
carbonaceous matter. The coloring material in some cases is inter-
stitial, or again where the heating has been prolonged penetrating the
cleavage and fracture lines of the silicates. In the case of the stone
of Sevrukof, Russia, the heat has been sufficient to produce a partial
fusion now manifested by the presence of a little interstitial brownish
glass.” Daubree, as quoted by Eberhard, thought this heating

‘4 The Two Solar Families.

55 Comptes Rendus, vol. 6, p. 178.

56 Proc. Nat. Acad. Sci., vol. 4, 1918, pp. 178-180.

8? The writer will here say that he has never found in any meteorite what he considered an original, re-
sidual glass such as is characteristic of many terrestrial igneous rocks. Such glass as may exist is secondary,
as in the case above. This, of course, does not apply to other chondrules, which are often more or less
vitreous.

88 Arch. f. d. Naturkunde Liv. Est. u. Kurlands, ser. 1, vol. 9, 1882.

COMPOSITION AND STRUCTURE OF METEORITES 29

to have taken place during the passage of the stone through the
atmosphere. ;

The stone of Indarch, Russia (fig. 1, pl. 20), is microscopically
of a dark greenish gray color, firm and compact, admitting of a polish,
and on the polished surface thickly studded with small, dark, almost
black chondrules and nodular masses of metal and troilite, the largest
of which are rarely over 1 millimeter in diameter. Under a pocket
lens the chondrules are mostly of a greenish color, though some are
nearly black. They break with the matrix in which they are em-
bedded. In thin sections and under the microscope the structure is
quite obscure. Owing to the prevalence of graphite, with which it is
everywhere impregnated it presents a dense black irresolvable ground
throughout which are scattered the iron and iron sulphide, together
with abundant sharp splinters of pyroxene and numerous more or less
fragmentary chondrules of the same mineral in both porphyritic and
radiating forms. All of the well crystallized forms, both in isolated
particles and in the chondrules, belong to the polysynthetically
twinned clinoenstatite type. Calcium sulphide, oldhamite, occurs
in this stone in the form of irregular areas, sometimes interstitial
and sometimes inclosed in the enstatite. It is of a yellow brown color,
sometimes greenish, completely isotropic, and with well developed
cubic cleavage.

The metal in stones of the chondritic class occurs as small irregular
particles, sometimes almost chondrule-like, as does also the troilite.
More commonly the two minerals are closely associated, the sul-
phide in irregular form being completely surrounded by a border of
metal (upper figure, pl. 21) which penetrates into the interstices of the
silicates enacting the part of a binding constituent. Or again, the
metal may occur simply capping the sulphide, or as a collar completely
surrounding a silicate particle as in Figure 2, Plate 20 (Cullison).
In many instances it occurs in the form of thin filaments traversing
the interstices and completely enfolding the silicates and penetrating
into fracture crevices as in the stone of Cumberland Falls, Ky., all of
which points to its origin as a secondary product from some pre-
existing form, perhaps a chloride. (Lower figure, pl. 21.) As else-
where noted the metal is most abundant in stones of a pronounced
chondritic type.

CHONDRULE: ITS NATURE AND ORIGIN ®

The term ‘‘chondrit,’’ from the Greek xovdpos, a grain, was first
used, so far as I am aware, by Gustav Rose to designate a class of
stony meteorites characterized by the occurrence of small granules
or “‘kugeln.”’

59 See Concerning the Origin of the Metal in Meteorites, by this author, Proc. U. 8S. Nat. Mus., vol. 73,
art. 21, pp. 1-7, 1928.

60 See On Chondrules and Chondritie Structures in Meteorites, by George P. Merrill, Proc. Nat. Acad.
of Sciences, vol. 6, no. 8, 1920, pp. 449-472.

30 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

... “Sie ist durch kleine Kugeln ausgezeichnet die aus einem
noch nicht bestimmten Magnesia-Silicate bestehen, und in einem fein
kornigen Gemenge eingemengt sind,” etc. The word, with the
addition of the terminal e, as Chondrite, has been very generally
adopted, with its original meaning, by English and American writers.
Unfortunately, as it would seem, a further modification of the word
as chondros, chondrule, chondrus, or chondrum has been introduced, at
first apparently synonymous in meaning with ‘‘kugel”’ as used by
Rose though it is to be noted that he did not define the word quite
as clearly as might be desired. He wrote: ““* * * in Bruche
erscheinen sie theils wneben theils fasrig, im letzern Fall jedoch stets
nur sehr feinfasrig, indessen doch immer bestimmt erkennbar fasrig,
besonders unter der Lupe * * * nie radial, sondern immer
excentrisch fasrig. . . .

No further reference is made to those of “‘uneben Bruche”’ and one
is left only to surmise that they may have been of a granular or
porphyritic rather than fibrous structure. The fact that Rose’s
work was written before the day of thin sections doubtless accounts
for the undetermined character of the magnesian silicate.

Tschermak in his Mikroskopische Beschaffenheit (1885) was little
more explicit in his use of terms than was Rose. He wrote: “ Kugel-
chen und wberhaupt rundliche Kérper, welche bald aus einem einzigen
krystallindividuum, bald aus mehreren bestehen, 6fters auch aus
verschiedenen Gemengtheilen zusammengestzt sind, bilden das
Gestein fast allen (Borkut) oder sie lagern unverletzt, d6fters auch
zersplittert in einer lockeren bis festen Tuffmasse.’’ Elsewhere he
includes all the rounded forms under the term ‘“‘chondren,”’ though
in his plate legends and descriptions he designates botu as kugelchen,
thus using the two terms synonymously.

A perusal of the literature shows that by English and American
writers, the terms “chondrule,” ‘‘chondrus,” “‘chondrum”’ or “chon-
dros”? are now and have for some years been applied to the rounded
and oval granules presenting a considerable range in mineral com-
position and still wider range in internal structure, thus making the
terms synonymous with kugel or kugelchen as used by Tschermak
above. Of later years and as illustrated in the generally adopted
scheme of classification ™ there has seemed a disposition to use the
term kugel in a descriptive adjective sense, as ‘‘kugelchen chondrit,”’
under which name are included stones containing chondrules (or
chondri) having a radiate structure—the spherulitic ® chondrites of
American writers. There has thus apparently arisen in the minds of
many a confusion which, as it seems to the writer, has been in part
at least responsible for the diverse views expressed concerning the

61 See Farrington’s Meteorites, p. 200.
® Or “‘globular,”’ see Proc. Amer. Philos. Soc., vol. 43, 1904 (p. 238).

COMPOSITION AND STRUCTURE OF METEORITES 31

origin of these peculiar bodies. In other words, there has been a
failure to recognize or discriminate between the kugelchen with
radiate structure and the often polysomatic forms with the irregular
fracture.

The following pages represent an attempt on the part of the author
to make this discrimination and to show how far proposed theories
may apply to the various forms presented.

At the outset and for the purpose of making clear what is to follow,
it will be well to figure and describe a few characteristic forms of the
individual chondrules. This notwithstanding the previous most excel-
lent and comprehensive work of Tschermak and Cohen.® (See pls.
18, 22-25.)

Mineralogically, the chondrules, using the word in its broadest
and most comprehensive sense, in nearly all meteorites are composed
chiefly of the minerals olivine or pyroxene, the latter in either ortho-
rhombic or monoclinic forms, or both. Some are largely of an
undifferentiated glass. Feldspars occur but rarely except in the form
known as maskelynite. In addition are occasional inclosures of metal
or metallic sulphides, chromite or other minor constituents. The
metallic iron sometimes occurs in rounded chondritelike blebs,
though it is doubtful if this should be referred to under that name.
Structurally, the chondrules in the same meteorite may vary from
densely cryptocrystalline, almost amorphous, to those that are part
glassy and porphyritic or even holocrystalline.

1. Glassy, eryptocrystalline, and radiated forms.—In Figures 1 and 2,
Plate 22, are shown examples of cryptocrystalline forms from the
stones of Barratta, Australia, and Cullison, Kans. That of Figure 1
is of a peculiar brownish translucency and very dense, resembling
the ‘‘felsitic” structure of the early petrologist. In the Cullison
stone, Figure 2, the chondrules, also of a brownish color, are not com-
pletely isotropic but between crossed nicols break up into several ill
defined areas over which the dark cloud sweeps faintly and irregularly
as the stage revolves. The material seems to be a partially devitrified
glass in a condition of optical stress as from sudden cooling. Chon-
drules of this type and those next to be described more nearly resemble
the spherulites of the terrestrial rocks than any others which have
come under the writer’s notice. Their outlines are at times as sharply
demarked from the matrix in which they are embedded as are the
spherulites in the rhyolitic obsidians of the Yellowstone National
Park. .

Chondrules of the radiating type are shown in Figures 3, 4, and 5,
from the meteorites of Elm Creek, Hessle, and Parnallee. The
mineral in all cases is enstatite ** and the outline of the spherule as

63 See particularly, Die Mikroscophische Beschaffenheit der Meteoritenkunde, respectively.
64 No attempt in these pages has been made to distinguish between enstatite and the ferruginous varieties
bronzite and hypersthene.

32 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

sharp and clean as though it had been turned on a lathe. In the
Elm Creek example crystallization evidently began at one point on
the surface of the spherule and extended inward throughout, but the
cooling proceeded too rapidly for the production of an optically
perfect crystal. In the Hessle stone (fig. 4), there were evidently
several initial points of crystallization. Forms like these grade
imperceptibly into such as are shown in Figure 5, in which the radi-
ating bars have unmistakably the crystallographic properties of
enstatite.

2. Half glassy, barred and porphyritic forms.—Porphyritic forms are
characteristics of both olivine and enstatite chondrules, while the
barred forms, such as are shown in the upper figures of Plate 23, are
limited mainly, if not wholly, to monosomatic forms composed of
olivine. In Figure 1, from the Beaver Creek stone, the white
portions are olivine which extinguish practically as a single unit;
the black portions are glass. It is to be noted that the outlines of
the chondrule though sharp are not smooth as in those described
above, but have projecting particles extending out into the ground;
also that this border portion often contains enclosures. In Figure
2 from the Cullison stone, the bars are bent and curved and do
not all extinguish simultaneously, as the stage is revolved, the dark
cloud sweeping over it irregularly, indicating a condition of stress.
Here, as in the last, the border is not sharply demarked from the
ground and it is often impossible to say if a certain crystal particle
belongs to one portion or the other. It should be noted that this
stone is a crystalline spherulitic chondrite. According to Tschermak,
in chrondrules of this nature the olivine bars are sometimes inter-
laminated with plagioclase (as for example, in the Dhurmsala stone).
In the porphyritic form shown in Figure 2, Plate 24, from the Tennasilm
stone, the granular ground abuts sharply against the blackglass of the
chondrule with only on one side a manifested tendency to penetrate
into and beyond the border. It is to be noted that the enstatite
phenocrysts within the chondrule and near the border are often cut
off sharply as though the sphere, originally much larger, had been
uniformly reduced by abrasion. This will be referred to later.

Holocrystalline chondrules—As would naturally be expected, these
porphyritic forms, through a reduction of the proportional amount
of glass, pass gradually into those which are almost or quite holo-
crystalline and polysomatic as shown in Figures 1 and 3, Piate 22, from
the Barratta and Elm Creekstones, respectively. Of peculiar interest
are those of the polysynthetically twinned pyroxene (fig. 3). For
some unexplained reason, these rarely grade into the half glassy
porphyritic forms, the entire chondrule consisting of the closely
crowded pyroxenes with comparatively little, if any, interstitial glass.

COMPOSITION AND STRUCTURE OF METEORITES 33

In Figure 1, Plate 24, from the Parnallee stone, it will be noted that
the crystals are in some instances slightly curved, their vertical axes
lying approximately parallel with the circumference of the circle
which forms the border of the section. The appearance is as if the
chondrules had been molded by external forces after the crystals had
formed but while yet in a more or less plastic condition. Again, the
pyroxene crystals abut sharply against the border and are cut off at
the margin as in the half glassy, porphyritic forms mentioned above,
and as shown in the figures. Occasional forms are met with which
have all the appearance of fragments, slightly rounded, of holo-
crystalline granular rocks, which as noted later, they are believed
to be.

Secondary borders about chondrules—A not uncommon feature of
the chondrule is the narrow border or rind about the circumference.
These borders as a rule, are of lighter color than the interior, of a
clear, more pellucid nature, though it may be including portions of the
minerals characteristic of the matrix in which they are embedded.
This is well shown in the olivine chondrule, Figure 1, Plate 23. This
border has an appearance at once suggestive of the secondary inter-
growth or enlargement often seen in feldspars and other minerals of
terrestrial rocks. The later portions sometimes, though not always,
have the same optical orientation as the interior.*° In some instances
the chondrules are surrounded by an irregular border of metal or
metallic sulphide.

Double or compound chondrules —Occasional forms are met with in
which a large crystal of olivine or pyroxene is inclosed by a border of
finer crystals of the same mineral but suggestive of a later generation.
Of greater interest is the occasional occurrence of a chondrule within
the mass of a second or larger form, as figured by Tschermak, on
Figure 1, Plate 8, of his Beschaffenheit.

Theories of origin.—In this review it will perhaps not be necessary
to go back much beyond the time of the introduction of the micro-
scope and thin sections into the study of rock structures since obvi-
ously little that was accurate could be told of them by the naked eye
alone.

A brief glance at the literature is sufficient to suggest that many
of the opinions expressed have been based upon examinations of
but a limited number of occurrences which quite failed to yield
the information necessary for building satisfactory hypotheses or
conclusions.

Reichenbach, as early as 1860 wrote: ‘Aus allerdem wird es
klar, das die Einschlusse in dem Meteoriten, als die Triimmer und

65 ] am not certain if this border is of a like nature to that described by Tschermak about some of the

chondrules of the Grosnaja stone and which he considered of secondary origin.
66 Pogg. Ann., vol. 3, 1860, p. 384. See also Chamberlin’s The Two Solar Families.

34 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

die geshicbartigen Knollen und Kugeln darin, keine einfach nachen
Bestantheile, sondern nichte anderes sind als auch wie die Meteoriten.
Meteoriten nur von anderer Anordung ein und derselben naturn
Bestandtheile.”’ And again: ‘‘Es sind also die Einschlusse theils
kleine meteoriten, theils Triimmer von meteoriten von hohenen Alter
als diejenigen meteoriten es sind, in welche sie eingeschlossen vorkom-
men; es sind altere kleinere meteoriten in jungerern grossern mete-
oriten.”’ In brief, and in plain English, he believed each particle as
now found to represent a minute but independent meteorite derived
from the breaking up of some older preexisting stone and now in-
cluded as a constituent part of one new formed.

In discussing the microscopic structure of meteoric stones, H. C.
Sorby, in 1864 © wrote:

It would, therefore, appear that after the material of the meteorites was
melted, a considerable portion was broken up into small fragments subsequently
collected together, and more or less consolidated by mechanical and chemical
action. * * * Apparently this breaking up occurred in some cases when the
melted matter had become crystalline, but in others the form of the particles
lead me to conclude that it was broken up into detached globules while still
melted. This seems to have been the origin of some of the round grains met with
in meteorites; for they occasionally still contain a considerable amount of glass,
and the crystals which have been found in it are arranged in groups radiating
from one or more points on the external surface in such a manner as to indicate
that they were developed after the fragments had acquired their present spheroidal
shape.

In continuation of this same idea in 1877, Sorby wrote:

As is well known, glassy particles are sometimes given off from terrestrial
voleanoes, but on entering the atmosphere they are immediately solidified and
remain as mere fibers, like Pele’s hair, or as more or less irregular laminae, like
pumice dust. The nearest approach to the globules in meteorites is met with in
some artificial products. By directing a strong blast of hot air or steam into
melted glassy furnace slag, it is blown into a spray, and usually gives rise to
pear-shaped globules, each having a long, hair-like tail, which is formed because
the surrounding air is too cold to retain the slag in a state of perfect fluidity.
Very often the fibers are of the chief product. I have never observed any such
fibers in meteorites. The formation of such alone could not apparently occur
unless the spray were blown into an atmosphere heated up to near the point of
fusion, so that the glass might remain fluid until collected into globules. The
retention of a true vitreous condition in such fused stony material would depend
on both the chemical composition and the rate of cooling, and its permanent
retention would in any case be impossible if the original glassy globule were
afterwards kept for a long time at a temperature somewhat under that of fusion.
The combination of all these conditions may very well be looked upon as unusual,
and we may thus explain why grains containing the glass are comparatively
very rare; but though rare they point out what was the origin of many others.
In by far the greater number of cases the general basis has been completely
devitrified, and the larger crystals are surrounded by a fine-grained stony mass.

87 Proce. Royal Soc., June, 1864.
68 Nature, London, April 5, 1877, p. 296.

COMPOSITION AND STRUCTURE OF METEORITES 35

Other grains occur with a fan-shaped arrangement of crystalline needles, which an
uncautious, nonmicroscopical observer might counfound with simple concretions.
They have, however, a structure entirely different from any concretions met with
in terrestrial rocks, as for example, that of oolitic grains. In them we often see a
well-marked nucleus, on which radiating crystals have been deposited equally
on all sides, and the external form is manifestly due to the growth of these crys-
tals. On the contrary the grains in meteorites now under consideration have an
external form independent of the crystals which do not radiate from the center,
but from one or more places on the surface. They have, indeed, a structure
absolutely identical with that of some artificial blowpipe beads which become
crystalline on cooling. With a little care these can be made to crystallize from
one point, and then the crystals shoot out from that point in a fan-shaped bundle,
until the whole bead is altered.

In this case we clearly see that the form of the bead was due to fusion, and
existed prior to the formation of the crystals. The general structure of both of
these and the previously described spherical grains also show that their rounded
shape was not due to mechanical wearing. Moreover, melted globules with
well defined outline could not be formed in a mass of rock pressing them on all
sides, and I, therefore, argue that some at least of the constituent particles of
meteorites were originally detached glassy globules, like drops of fiery rain.®

In this Sorby would appear to have had reference only to “kugels”
with radiate, internal structure.

Tschermak, who together with Haidinger, was one of the first to
pronounce on the tuff-like character of the chondritic meteorites,
announced in 18747° the opinion that the individual chondrules
(kugelchen) were but rock particles which became simply rounded
under conditions similar to such as might exist in the throat of a
terrestrial volcano.

Ich wiederhole hier nur das Eine, dass Ich die Chondrite fur Zerreibungs-Tufe,
und die Kugelchen derselben fur solche Gesteinspartikelchen halte, welche
wegen ihrer Zaihigkeit bei dem Zerreiben des Gesteines nicht in Splitter aufgelost,
sondern, abgerundet wurden.

And again in 1875:7

Man kann sich allenfalls vorstellen dass die Steinmassen, welche der Zerrei-
bung ausgesetzt waren, ziemlich weich gewesen seien und wiirde sich dadurch der
Vorstellung Daubrees naihern, welcher an ein Gestein denkt, welches in einer
Gasmasse wirbelnd erstarrte; doch ist es sicher, dass die Kugelchen das Resultat
einer Zerreibung sind.

69 Sorby’s idea, as it seems to the author, may be visualized and made probable by a consideration of a
meteorological phenomenon common during the winter months in cold latitudes. It not infrequently
happens that on the occasion of a cold winter storm, owing to meteorological conditions, the snow (con-
densed water vapor) falls not in the customary form of flakes, but in that of hard, icy pellets of approximate
pin-head size, comparable with a chondrule. In a cold, dry atmosphere and with a high, intermittent
wind, these particles are driven at a great speed in all directions through the air, constantly colliding with
one another and ultimately reaching the ground, where, under the same influences they are drifted over the
surface, still colliding with one another and the various obstructions, until an ever variable proportion
of them are shattered or disintegrated—finally coming to rest to be compacted in a more or less solid mass
comparable with that of the more friable chondritic meteorite stones. The theory, it will be observed,
thus accounts for the origin of the chondrule as well as that of the stone itself.

70 Die Triimmerstructur d. meteoriten, etc., Sitz. k. Acad. Wiss., Wien., vol. 70, 1874, (4).

1 Idem, vol. 71, 1875.

36 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

In both of these quoted expressions Tschermak seems to have had
in mind only the granular and porphyritic polysomatic forms, and the
fragmental ‘‘kugelchen.”’

Three years later” after a consideration of the depressions and
excrescences occurring in and on the chondrules of the Tieschitz
meteorite, he came to a partial agreement with Sorby conceiving
that

Die Kiigelchen sind nach wie vor wegen der tuffartigen Beschaffenheit der
Meteorsteine als Resultate vulcanischer Eruptionen und Explosionen anzusehen,
aber ihre Form diirfte doch eher von einem plastischen Zustande, als von der
Zerreibung starrer Partikel abzuleiten sein.

And again, after another four years ™ he announced:

Ich hatte * * * gu der Ansicht gefuhrt wurde, dass die Kiigelchen der
Chondrite als erstarrte Tropfen anzusehen sind, wahrend die aus Splittern
bestehende Grundmasse nach wie vor als vulkanischer Detritus zu betrachten
waren.

Daubrée ™ seems also to have held the opinion that the majority
of chondrules were simply débris particles rounded by attrition. He
wrote:

J’ai montré que la structure globulaire telle qu’elle se présente dans certains
types * * * a été imitée artificiellement et s’explique par une sorte de
granulation operée au moment ou la substance se solidifie. Mais le plus souvent
les globules des météorites paraissent étre des simples débris arrondis par
frottement.

F. Rinne,” by means of a simple electric device, was able to fuse
the silicate constituents of meteorites and by abrupt alterations of
the strength of the current produce a “‘spratzen”’ of the melt resulting
in the projection from the crucible of small drops which quickly
cooled in the form of “‘kugels.’”’ To some such’ action he would
ascribe the formation of meteoric chondrules. Later, by the aid
of an oxygen blast and a Linnemann burner he was able to produce
enstatite beads evidently in every way comparable with meteoric
chondrules.” These, it will be observed, are really synthetic demon-
strations of the possible correctness of Sorby’s views. Berwerth in
1901 ” announced his conviction that the chondritic stones were
tuffs more or less completely metamorphosed by heat, and seemed
to regard the individual chondrules as portions of the melt that
cooled in globular form. Borgstrom % in his description of the
Hvittis stone (1903) (a crystalline chondrite), says that the chon-
drules are always so firmly intergrown with the ground that it is
often impossible to determine where the one leaves off and the other

72 Denk. Math. Natur. Classe kaiser. Akad. Wiss., vol. 39, 1878.
73 Sitz. k. k. Akad. Wiss., Wien, vol. 95, 1882, p. 205.

™ Géologie Expérimentale, 1879, p. 530.

75 Neues Jahrbuch Min. Pet., vol. 2, 1895, pp. 229-246.

76 Idem, 1897, pp. 259-261.

7 Centralblatt Min., etc., No. 21, pp. 641-647, 1901.

78 Die Meteoriten von Hvittis u. Marjalatti, Helsingfors, 1903.

COMPOSITION AND STRUCTURE OF METEORITES ot

begins. In many instances, the enstatites of a chondrule extend out
into the ground mass with which they are intergrown. As noted, the
Hvittis stone is a crystalline chondrite; this might suggest either a
crystallization of the chondrule 7 situ or a case of secondary enlarge-
ment. In writing of the Shelburne stone, a gray chondrite, however,
he says,”

Each individual chondrule represents a structure of cooling and crystallization
from a molten state, and as their structure shows an intimate relation to the
boundary of the chondrule it must be supposed that each, at the time of its
solidification, was a separate unit. Because chondrules of the same chemical
composition have a different structure, they must have been formed under differ-
ent physical conditions. Since such a variety of conditions can not have existed
in the narrow space in which the different structures now are met with, the
chondrules must have accumulated after solidification.

Such a condition is well shown in Figure 2, Plate 18, from the stone
of Cedar, Tex. Meunier ® basing an opinion apparently upon the
theoretic work of M. Faye, suggests the probability of the chondrules
resulting from the sudden condensation of a cyclonic vapor.

Il parait difficile de ne pas admettre que les chondres sont aux roches de pre-
cipitation gazeuse ce que les dragées de Carlsbad et le fer en grains sont aux
roches de précipitant aqueus * * * Conformement 4 la terminologie dont
font usage les paleontologistes 4 propos du vent fossile, du soleil fossile, de la
pluie fossile, on serait tenté de les qualifier de cyclones photosphériques fosstles.

This is conceivable, to the present writer, only in the case of
radiate enstatite or monosomatic forms.

Brezina, to whom is so largely due the building up of the magnifi-
cent collection at Vienna, concludes a review * of the subject with the
statement:

Durch die vorangefuhrten Beobachtungen kénnen wohl die alteren An-
schauungsweisen als beseitigt betrachtet werden, und wir kénnen wohl mit
Bestimmtheit die Meteoriten als gestore iiber hastete Krystallbildungen in
einem einzigen gemengten Magma bezeichen.

This apparently includes both the ground mass and its chqndrules.

Hussak, basing an opinion on experimental work by himself and
Dolter ” suggests that chondritic meteorites, like that of Uberaba,
Brazil (a crystalline chondrite), originate through the long continued
immersion of meteoric stones in a nickel-iron magma, and are to be
regarded as true volcanic ejectmenta,

Ich méchte demnach die Meteorsteine durch ultrabasische Eruptivgesteine

vergleichen und die Bildung der Chondren wie der Triimmerstructur und der
schwarzen Adern als eine magmatische Eirnwirkung vor der Ejerktion ansehen.

7 Trans. Royal Astr. Soc., Canada, 1904.

80 C. R. Paris Acad. Sci., vol. 96, 1883, p. 868.

§1 Die Meteoritensammlung, etc., 1885.

82 Neues Jahrb. fur Min., etc., vol. 1, 1884, pp. 18-43. They immersed fragments of an “‘olivinefels’”’
for many hours in a slowly cooling melt of nephelin basalt. The stone was strongly attacked and the
outer portions, in close contact with the melt, shattered and corroded, the olivine granules becoming filled
with embayments and enclosures of a secondary colorless glass, all strongly suggestive of meteoric
chondrules.

38 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

Daher die vollstandigen Ubergange in Siderite und die Deutlichen Korrosion-
serscheinungen an den grossen Olivinkristallen der Pallasite.

C. Klein, in 1906 ®* evidently basing an opinion largely on figures
of chondrules in the works of Hahn and Tschermak, affirmed that
there occur many ideally perfect forms that lack the eccentric radiat-
ing structure, but are ‘‘radial strahlig’” from a center, equally in all
directions and are true spherulites. Those not having this perfection
of structure are considered fragments. It may be well to note before
going further, that Klein apparently stands alone in holding these
views though they may be correct for certain forms.

Wahl * would explain the formation of the chondrule as due to
the cooling of a silicate melt in a heated atmosphere, the resultant
drop crystallizing from the surface inward.

Die Entstehung der Chondren liasst sich also ganz allemein als durch Zer-
stiubung von Silikatschmelz fluss innerhalb einer heissen Atmosphaére und
Kristallization der hierdurch enstandenen Tropfen von aussen nach innen zu
erklaren.

This again would seem to refer only to the cryptocrystalline,
radiating enstatite, and the barred and monosomatic olivine chon-
drules.

Finally in 1913, Fermor * of the India Survey, suggested that the
chondrules are remelted garnets.

The views of the present writer have been set forth elsewhere * and
need not be repeated here in their entirety. It suffices to say that
struck by the discordant character of the views expressed he studied
the forms not merely as shown in the thin sections, but in their
complete forms as freed from the matrix of some of the more friable
stones. (See pl. 25.) It was found (1) that the most perfectly spherical
and oval forms occurred in those stones the tuffaceous nature of
which was beyond question. These show a cryptocrystalline or
radiating internal structure and are mineralogically of pyroxene. (Fig.
5, pl. 22, and fig. 1, pl. 25.) They often show excrescences or saucer-
shaped ‘depressions, as through shrinkage or interference during
solidification. (2) Other forms, more irregular in shape (fig. 2 of
plate 25) show a rougher surface, and interiorly are of a polysomatic
nature—composed of phenocrysts of olivine or pyroxene in a more
or less glassy base or of an almost holocrystalline aggregate of one or
more minerals. His conclusions were then to the effect that:

1. Only the chondrules of glass and cryptocrystalline or radiating
enstatites (kugelchen) present the rounded or oval form with smooth
rindlike crust and surfaces, with often one or more saucerlike depres-
sions or excrescences such as are consistent with a theory of origin
as fused drops of ‘‘fiery rain.”’ (Sorby.)

83 Studien uber Meteoriten, p. 35.

& Zeitschrift Anorg. Chem., vol. 69, 1910, pp. 52-96.
85 Records Geol. Survey India, vol. 43, 1918, p. 45.
86 Proc. Nat. Acad. Sci., vol. 6, 1920, p. 449.

COMPOSITION AND STRUCTURE OF METEORITES 39

2. Chondrules of a compound, holocrystalline nature, and those
porphyritic through the development of olivine or pyroxene pheno-
crysts in a more or less glassy base are lacking in smooth exteriors and
though often quite spherical in outline, are as a rule more or less
irregular and in many instances show unmistakable evidences of an
origin of form through mechanical attrition. These last should be
designated chondroidal forms, rather than true chondrules. These
distinctions are well shown in Figures 3 and 5, Plate 22, and in the
general view from a thin section of the stone of Cedar, Tex. (fig. 2,
pl. 18).

ORIGIN OF OTHER TYPES OF STRUCTURE IN METEORIC STONES

The question of the origin of the various types of texture and
internal structure of stony meteorites has been, as has that of the
chondrules themselves, a much disputed one, as already noted. By
many, including such authorities as Brezina, Link, Renard, and the
American Wadsworth, the obscure and confused structures shown by
stones of the chondritic group are due merely to hasty crystallization
succeeded in some cases by crushing. To others, including Tscher-
mak, Sorby, Berwerth, Wahl, and the writer, they are for the most
part due to a tuffaceous origin, accompanied in many instances by
metamorphism. Thatis, they arecomparable with more or less com-
pacted and altered masses of volcanic ash or tuff. Certain stones,
like those of El Nakhla, Juvinas, and Shergotty, are apparently
products of direct cooling from a molten magma and their clastic
structure, when present, due to mechanical causes. Others, as those
of Allegan, Hessle, and Quenggouk are so plainly tuffaceous as to
seemingly be beyond argument. There yet remain certain abundant
types, however, belonging to what are classed as the crystalline,
erystalline-chondrite, and white, gray, and intermediate chondrite
groups, the structures of which are obscure and which, though
commonly regarded as metamorphosed tuffs, yet furnish grounds for
reasonable doubt as to their origin.

The writer, however, in a recent summary, regards the tuffaceous
nature of the stones classed as spherulitic chondrites as no longer open
to argument. The crystalline types mentioned he considers products
of metamorphism, in this agreeing with the other workers quoted.
The grounds for this belief are summarized as below. *

The most perfect chondrules and chondroidal forms are found in
those stones the fragmental nature of which is most pronounced, and
become less perfect, more highly altered, often merging imperceptibly
into the groundmass as the stones pass from fragmental into crystal-
line forms (pls. 19 and 20) as those of Estacado and Bluff and Indarch.

87 On Metamorphism in Meteorites, Bull. Geol. Soc. Amer., vol. 32, 1921, pp. 395-416.

40 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

It is therefore suggested, though not insisted upon, that the mere
presence of a chondrule in a meteorite, whatever its condition, is
indicative of a tuffaceous origin. The clear, limpid interstitial ‘‘glass”’
sometimes quite isotropic and sometimes doubly refracting, known as
maskelynite, is shown to have been, together with the phosphate
merrillite among the last of the constituents to solidify, and probably
a product of a reheating and cooling too abrupt for crystallization.
The dark, glassy interstitial material sometimes surrounding a chon-
drule (fig. 1, pl. 27) and the pellucid borders presented by some of
the feldspars in the Estherville stony iron, are considered of like origin.
It is shown further that the occasional crushing of the individual
constituent, while productive of a cataclase structure, is a very minor
feature and without necessary bearing on the question of the original
nature of the stone (figs. 3, 4, 5, pl. 27). Lacroix also holds to this view.
This condition, it is variously conceived, may have been produced by
compression within the mass, by the shock of a collision, or too abrupt
and extreme changes in temperature, as when a comet approaches
the sun and then flies off again into the cold of space. It is possible
that all three may have operated at various times and under various
conditions. It is also conceivable that it may be in part due to
impact with the earth’s atmosphere. (See further under Metamor-
phism).
METAMORPHISM IN METEORITES

Not all of the phenomena of structure and composition noted can
be considered original. Some are unquestionably of secondary
origin—a result of accident or changed conditions, and may in part
be designated metamorphic, as with terrestrial rocks.

With the possible exception of the chondrules, no meteoric con-
stituent, it may be stated, presents a more interesting puzzle than
the metallic portion. As has been noted, this is not a simple but a
compound body consisting of three more or less definite alloys of a
composition, and an arrangement among themselves not found in
terrestrial irons, artificial or otherwise.’ Indeed the octahedral ar-
rangement of the plates, when such exists, is considered sufficiently
characteristic to insure the meteoric nature of any iron in which it
may be found, whether or not seen to fall. The conditions under
which such an arrangement could take place are not as yet quite
understood and need not be considered here. The feature that now
concerns us is its lack of stability, its susceptibility to change, under
changed conditions. It has been shown that if an octahedral iron
like that of Toluca, Mexico, an etched slice of which is shown in
Figure 1, Plate 26, be heated for a few hours at a temperature below

88 Artificial alloys comparable with the meteoric irons have been formed. (See Benedict, Neues Jahr. ,
1912, vol. 1, p. 44). While, however, these show an octahedral structure the clearly marked separation into
alternating plates of kamacite and taenite is lacking.

COMPOSITION AND STRUCTURE OF METEORITES 41

redness, it gradually assumes the granular structure shown in Figure 2.
If the heat be continued for a sufficient period, the octahedral structure
will entirely disappear and the iron show only the granular structure.
This feature has has been noted and described by Berwerth * of the
Vienna Museum, who applied to it the name metabolism, and to
irons in which the change had taken place that of metabolites.°°
Some irons, like that of Roeburne, Australia, which, so far as known,
have not been heated since reaching the earth, show a combination
of the two structures, and others are wholly granular. Are these
octahedral irons which, in their wanderings have become highly heated,
perhaps by too close proximity to the sun, and become thus meta-
morphosed? Whoshallsay? Itis at least something worth thinking
about.

The ultimate source of the metal has likewise been a matter of
speculation. By Daubree and some others it has been thought to
have been derived, by a process of reduction, from some iron-rich
silicate such as olivine. This is, however, extremely improbable since
nowhere are there evident signs of the process in its incomplete stage.
Moreover, the olivine fragments in such iron-rich forms as that of
Admire, Kans., are all perfectly fresh and sharp as so many frag-
ments of broken glass. That the iron was never in a molten condition
is shown not only by the uncorroded condition of the silicates, but
by the physical condition of the metal itself, which is not that of a
metal cooling from fusion, like ordinary ‘‘cast iron,’ but is rather
that of the soft, malleable material commonly known as ‘wrought
iron”? which may be smelted from its ores at a comparatively low
temperature. It seems, therefore, altogether probable that the metal
results from reduction from the chloride form and that the small
amount of this material now found as lawrencite is but a residue,
as was suggested by Meunier several years ago.°*!

That it plainly was not a portion of the molten magma from which
the other constituents crystallized out is shown further by the position
it often occupies relative to the silicate constituents as shown in
Plates 20 and 21, where it is found as a mere film enwrapping chon-
drules and crystal fragments indicating an extreme degree of fluidity.
It is possible to conceive of this having been brought about through
the percolation into the interstices of the porous tuff of a gaseous or
liquid chloride, afterwards to be reduced.” Such a reduction, as

89 Sitz. der Kais. Akad. der Wiss., vol. 114, May 1908.

°0 That the granular structure might be secondary was first suggested by Sorby in 1887. He did not prove
definitely that it might be produced artificially by heating.

%1 See Concerning the Origin of the Metal in Meteorites, by George P. Merrill, Proc. U. S. Nat. Mus.,
vol. 73, Art. 21, pp. 1-7, pls. 1-3, 1928.

% The Estherville meteorite offers a strong argument in favor of the origin of the metal through chloride
reduction. This meteorite is a metamorphosed conglomerate, quite slaglike in portions, and the metal
often occurs only partially filling the cavities as would naturally be the case did not the supply of material
continue throughout the reducing process, the chloride consisting of but 44.1 per cent of iron and 55.9 per
cent chlorine.

<
42 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

noted by Nordenskiold “ and others, must have taken place outside
of our atmosphere and in one deficient in oxygen. Perhaps attention
need be called to the fact that metal occurs in quantity only in stones
which are plainly of a fragmental nature. In achondritic types of an
original crystalline structure it is almost entirely lacking. It may
be noted, incidentally, that the average amount of metallic iron in
stony meteorites is 11.98 per cent, which is equivalent to 16.55 per
cent of magnetite or 27.16 per cent of purely ferrous lawrencite.

The sharply angular, uncorroded condition of the silicates in pal-
lasites of the réckiky group has been noted as indicative of low temper-
ature reduction of the metal. The question may well arise, however,
could not this structure be produced by a shearing pressure on a
pallasite of normal structure in the same manner as foliated and schis-
tose rocks are derived from massive terrestrial forms? It seems at
least possible.

Since very early in the study of meteorites, there have been held
radically different opinions among students as to the causes of the
clastic structure so pronounced a feature of many stonesand particu-
larly those of the chondritic types. This has been referred to else-
where (p. 39) but it will be well to enlarge upon the matter here.
Attention has been called to the fact that the most perfect chondrules
like that shown in Figure 3, Plate 22, occur only in those rocks con-
cerning the tuffaceous nature of which there could be little doubt.
And further, that in those stones which are approximately crystalline,
the chondrules, where such have existed, are more or less distorted and
sometimes obliterated as in Plate 23. Further than this there are
often evident signs of compression in the mass such as has led to
fragmentation of certain constituents, as shown in Plate 27. It is
true that a portion of this fragmentation may be due, as has been
contended, to abrupt changes in temperature as when a meteorite
approaches the sun and then rushes off once more into the cold of
space, or to the shock of a collision; but in any case they are secondary
and have little to do with the original structure of the stone.

Evidences of metamorphism in which heat is the primary factor
are afforded by the Bereba eukrite described by Lacroix “ and in the
chondritic breccia of St. Michel described by Borgstrém % who says
distinctly :

Die Grundmasse is kein Verfestigungsprodukt, sonder das Resultat einer
unvolstaindigen Metamorphose eines Triimmergesteins, das aus Kornchen und

Splitterchen derselben Mineralien die auch jetzt den Stein aufbauen bestanden
hat.

% Zeits. d geol. Gesell., 1881, p. 25.

% Merrill, George P.: On Metamorphism in Meteorites, Bull. Geol. Soc. Amer., vol. 32, 1921, pp. 395-416.
% Archiv. du Mus. d’hist. Nat., ser. 6, vol. 1, 1926, p. 35.

% Bull. Com. Geol. de Finland, No. 34, 1912, p. 36.

COMPOSITION AND STRUCTURE OF METEORITES 43

In like manner the groundmass of the Hendersonville stone is
described as not at all that of minerals crystallizing freely from a
molten magma, but suggestive, rather, of a partial recrystallization of
fine detrital material as seen in metamorphic schists. Instances
in which the direct action of heat alone is more evident is afforded by
Figure 1, Plate 24, which is that of a chondrule in the stone of Parnallee,
India. That the chondrule is foreign to the ground in which it is
embedded is obvious. The present interest in it lies in the dark,
glassy border by which it is surrounded and which is considered due
to the action of heat on the fine, dust-like material in which it was
embedded. That it is not an original residual glass should be evident
to any petrographer.

An equally instructive illustration of heat action is shown in the
transformation of a normal plagioclase feldspar into the mineral
maskelynite, as first noted by Tschermak and since verified by others.
This is considered by the writers as indicative of a reelevation of
temperature sufficient to change its character even if not completely
fuse it, and, on sudden cooling, leave it in the form of a feldspathic
glass. As stated elsewhere, the mineral is not always isotropic but
shows frequent transition stages to the normal mineral. In the Mocs
meteorite the feldspar occurs, according to Tschermak, as plagioclase
in the mass of the stone and as maskelynite in the crust.

The possibility of a refusion and crystallization of the feldspar
without the formation of maskelynite has been shown by the writer
in the case of the Estherville meteorite which is regarded as a meta-
morphosed agglomerate, the finer portions of which (fig. 2, pl. 2)
show a structure not unlike that of some partially altered crystalline
schists in which the feldspars fluxed without altering the fragmental
structure of the pyroxenes. A striking feature of this meteorite which
has not before been mentioned is the presence in it of boulderlike
masses of a different structural nature than the mass of the material
(pl. 13). It is to be noted that while around the border the silicate
material seems to merge into the general ground of the main mass,
the interior while evidently of the same mineral nature is of a more
regular texture and the metal is in very fine threads which at times
completely surround, or enfold the silicate particles as is often the
case in ordinary chondritic stones. The percentage amount of metal,
it should be said, is plainly much less than in the main mass. The
question arises, Is the nodule a pebble or a portion of the original
ground of the agglomerate which has escaped the metamorphic changes
of the rest of the mass? Jn either case the metamorphic nature of the
meteorite seems fully substantiated. Lacroix ** has noted the recrys-

% Merrill: Proc. U.S. Nat. Mus., vol. 32, 1907, p. 80.
88 Archiv. du mus. d’hist. Nat., ser. 6, vol. 1, 1926, p. 35.

59587—30 4

44 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

tallization of the finely granular interstitial silicates in stones of the
eukrite and howardite groups.

Many of the stony meteorites are traversed by small, black, thread-
like veins, at most but a few millimeters in diameter (fig. 2, pl. 29)
which are plainly due to a fracturing of the stone, though whether or
not prior to entrance into the earth’s atmosphere is a question. A
greatly enlarged section of one of these from the Bluff, Fayette County,
Tex., meteorite is shown in Figure 2, Plate 28. The filling material
of the vein is of a nearly coal black color, opaque, and of an undeter-
mined nature, inclosing white and gray particles of the minerals
composing the body of the stone. Occasionally a slight movement
between the walls of these veins has developed a structure known
as slickensides in terrestrial rocks. In the illustration shown, no
such movement has taken place, and it will be noted that the coloring
material penetrates into the walls in the form of small veinlets on
either hand. Much discussion has taken place concerning the origin
of these veins and a great divergence of opinion manifested, a part
of which is evidently due td mistaken ideas regarding the nature of
the filling material, often referred to as “metallic.” As a matter of
fact, the material is metal in comparatively few cases, but is apparently
identical both in composition and origin with that forming the base
of the so-called black chondrites, and can best be accounted for through
a slight modification of the idea expressed by Wahl, the shock from a
collision producing the fracture, along which is immediately prop-
agated a heat wave sufficient to produce the result. In this way the
minute ramifications (spirts) of the vein matter into either wall, as
shown in the figure, would be readily accounted for. The occasional
presence of metal, or metallic sulphide, as an apparent filling con-
stituent, can be best explained as has Farrington * on the assumption
that either constituent occurred in the form of more or less connected
filaments. Fracturing would naturally take place along these lines
rather than across them. That heat could have melted the metal
without affecting the silicates is impossible, and that the filling matter
is not of the same nature as the crust (that is, a glass) is almost
self-evident.

That collisions among meteorites are not limited to the stony forms
is strikingly shown in Figure 1, Plate 29, from a polished slice of an
iron meteorite found a few years ago in Somerset County, Pa. It
will be noted that the iron is traversed by a fine, thread-like fissure
along which has taken place a movement for a distance of something
like a centimeter. In short, it is like a fault in terrestrial rocks. As
the iron is soft and malleable, we are apparently justified in the

99 Amer. Journ. Sci., vol. 11, 1901, p. 59.

COMPOSITION AND STRUCTURE OF METEORITES 45

assumption that the shock producing the fracture took place some-
where in space where the metal had become so cold as to be brittle.

The black coating on the surface of the stony meteorites is, as
already noted, a more or less perfect glass, due to the fusion of the
various constituents from the heat generated during the passage of
the stone through the atmosphere. This, as shown in thin section
(see pl. 28, fig. 1) is never of more than a few millimeters in thickness.
It consists, in the case of the Allegan stone figured, of a black glass
interspersed with numerous residuary particles of metal and unfused
silicates, which passes gradually into the unaltered granular stone.
Sections of the thick, blebby glass from the lower surface show air
vesicles and numerous crystallites imperfectly secreted from the
glassy base, and too small to be seen in the figure, together with the
residuary unfused particles of the original minerals.

The so-called ‘‘black chondrites’”’ are considered by Meunier and
others as chondrites of the ordinary white, gray, or kugeln type which
have been heated to a temperature considerably short of that of fusion,
as already noted.

METEORITES COMPARED WITH TERRESTRIAL ROCKS

From what has been written it must be evident that, though com-
posed of the same elemental matter, meteorites differ in some very
marked respects from terrestrial rocks. Nevertheless there are
resemblances, some of which, when one considers the problematic
source of these bodies, are of peculiar interest.

All known meteorites are composed of volcanic materials, and none
has shown any traces of animal or vegetable life, unless the carbona-
ceous matter is to be so considered. This, however, is a wholly
unnecessary and, indeed, unwarranted assumption. It is true that
the German, Otto Hahn, when the possibilities of the microscopic
study of rocks were first becoming realized, described as organic
(corals and crinoids) what are now known to be but incipient crystal-
lizations of silicate minerals. Nothing in the nature of a terrestrial
sedimentary rock, a sandstone, shale, or limestone, or a metamorphic
like a schist or gneiss, has yet, so far as known, come to us from space,
nothing of a pumiceous nature, and nothing in content of silica,
alumina, lime, or alkalies corresponding to the granites (the tektites,
if meteoric, would most nearly correspond to this type of terrestrial
rock) and nothing of the nature of a true vein rock. Further, and
this seems the more singular when theories of earth history are
considered, nothing that can with certainty be ascribed to a meteoric
origin has been found in terrestrial beds of any geological horizon

46 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

but the most recent.!. If such have fallen during earlier periods, they
must have been a quite different type, or what is more probable,-
become so thoroughly decomposed or otherwise altered as to be
unrecognizable.

Owing to the presence of abundant oxygen in our atmosphere, the
iron in the terrestrial rocks is nearly always in an oxidized condition;
its presence as metal is exceptional.

Other minerals noted as found in meteorites in smaller quantities
and lacking in terrestrial forms are cohenite,? lawrencite, merrillite,
osbornite, and schreibersite, elsewhere described. It will be seen,
therefore, that the chief chemical and mineralogical differences lie in
the presence of unoxidized combinations in the meteorites, and not
in elemental composition. Rarely occur forms of crystallization like
that shown in Figure 2, Plate 29, which are more nearly allied to the
basalts and pyroxenites. The nearest terrestrial approach to this
meteorite are the iron-bearing basalts of Greenland and Ober Cassel,
in Germany, which shows native metal dispersed throughout a ground
of silicate minerals.?

The nearest terrestrial equivalent to the stony meteorites as a
whole is a comparatively insignificant group of intrusive igneous rocks
to which the name peridotite is given. These, like the meteorites, are
composed essentially of the minerals olivine and pyroxene, and so
close is their analogy that an element found in one may be predicated
for the other, though not necessarily in the same form of combination.
The most pronounced difference is in the commonly fragmental nature
of the meteorite and presence of iron in the metallic state. We know
of no volcanic (tuffaceous) equivalent of our peridotites unless the
diamond-bearing breccias of Arkansas and South Africa be so con-
sidered. A singularly striking similarity les in the presence of
diamonds and platinum in certain members of both groups, though
in meteorites only in minute quantities.

In the following table is given in I the average results of the analy-
ses of 63 stony meteorites; in II that of 8 peridotites, the group of
terrestrial rocks most nearly allied to meteorites; and in III the
average composition of the rocks of all kinds composing the earth’s
crust.*

1 This fact was noted by Olbers nearly 90 years ago. Ward’s statement as to the Pliocene age of
the Lujan mesosiderite seems contradicted by its having been found in ‘‘an undisturbed Quaternary
formation.”

? Cohenite has been reported by O. Sjostrém in the native iron of Greenland.

8 Chemical analyses show this iron to contain only about 1.90 per cent of nickel. In both cases the metal
is secondary and in the case of Ober Cassel derived by reduction from the sulphide pycrhotite. “/ ith

4 Clarke, F. W., Data of Geochemistry. Bull. 491, U S. Geol. Survey, 1911, p. 27. The 63 analyses of
stony meteorites include all varieties of which complete and satisfactory analyses were available.

COMPOSITION AND STRUCTURE OF METEORITES 47

Constituent I II Til
Poul eas CRS Og) eee SI a Pl 2 LN ht 38. 41 37. 78 59. 93
AHANICIOXICO) CLIO) cece on cores. cee wn ua ol ee eee LY 716 . 58 . 74
MPR ra eR CLE) (ISTH O)a) sees RL ea I hs eae LE eb On ee Nines) | noo Je 2 |. eee
PALCORINTHORICE) (21 Og) homens meen ne ne ie etn ae gee ene ce eae INONGA |b nen oes . 03
UNL YaTTIA TAS CANO g pee es lens OMe eal co OY El da te hs Da 2. 86 3.11 14, 97
DAMN OPORTO CG2 Olay nomen a eae te a ee Ce I ee See . 92 2. 41 2. 58
hhromic. oxidei(CrsOy ee toe eee IN eo Da Ie SN - 40 .19 .05
PAIAG TT ORICON Vas )eansese ses se a ae ree ne ee Eee ono “ETACG. |e soa ee bee . 02
Ivelablicyinont ie) <8. 2402 = Lk ooo ae ot Ble eo Se ee go ek 2530) |anee esas — = bee eee

VISTA CG THCKOL \(CNIL) = = eee ss See eee Ce oe Ra a = ete ee 00 pis cae eee
Metallic cobalt (Co) LO Ee 2a Sate See
Ferrous oxide (FeO) 5 18. 36 3. 42
Nickel oxide (NiO)-_------------- c -18 . 03
Cobalt oxide (CoO)_ 0672. 2 eee
Mime; (CaO) ee 2%. -- 3 3. 06 4,78
Barium oxide (BaO)- fe fy ANIOTIGS, cpa oer ees oe
INE Si ay (VLE ©) ae sy: Cag RE ys Wahab ee Oe oe ee ee ee ee . 66 28. 38 3.85
NVR ATIOUSIORIO ONC Win ©) 5 seas Speen We ene nL ee Se Se ole ao Soil .10
Sirontinrmoxide: (SPO) spe a ee ee Ee eee Es lege INO1Gs 42 9te seer . 04
SURI SUN (EN RTO) eas ena eee Noe Seeks eae ie Oe Re ee oe eR oe See ee . 82 . 68 3. 40
[RUSTE en) ee Se) a TB RO Nan oo Dg SP yeah SEE aE oe Su ee .16 on 2. 99
SRT Bt sun Gloia Oe eee ne oe ee VR ee a PS 2 PE eS seen eae a REACH ae eee 01
priory CEG@)) el eek a a eI re se ee .47 3. 79 1.94
FIR STMOPICVACIE (bse Os) scene eee eee ease ete ne eI Oe SE Re te ene . 34 .10 . 26
SUPE ie oe at eek 8 SS ade ee ue we a Ee ee ks Seg TSSOP | ie ee oli
SO GDTOTH UCU) nee oe Se ee are eat ial Hoe eee ce OER ee | SOG oe ee eee eee
GS PSareE ea (OC) a a i a ie oe ed ate SG pee ee ES Se
COULD NDA ITE S(O eS a I a SSR Sa erie aL CT eR Osh ee eea ore . 06
Rr OKHIC ACIMU(C Ws) ese pele eke el eee SC Wl sh et ate (2) 75 48
PAE HINO) (ieee oe eee a oat oO 50 ee aa ee eae Gio)ee eee Seems .10

| 160.00 100. 00 100. 00

The most important of the differences brought out by the analyses
are (1) the excess of silica (SiO2) and alumina (AI,O;) in the terrestrial
rocks; (2) the presence of a considerable amount of free iron and
proportionately large quantities of ferrous oxide (FeO) and mag-
nesia (MgO) in the meteorites. The presence of many of the rarer
elements tabulated as constituents of the terrestrial igneous rocks
has not yet been fully established in those of meteoric origin. As
noted, however, many of them have been found in amounts too small
to estimate.> Here indeed is a striking thought—that throughout
all space so far as yet made known to us, there exists so great a uni-
formity of material yet so individualized that one conversant there-
with can tell almost at a glance whether celestial or terrestrial in
origin.

TEKTITES

Of late years there has been much discussion relative to the possible
meteoric nature of certain glass pebbles of a green or black color
with peculiar markings, found on the immediate surface or slightly
embedded in Tertiary and Quaternary gravels in Australia, Moldavia,
islands of the Malayan Peninsula, and a few other localities. The
Australian forms, variously known as “bombs,” “obsidian buttons,”
‘“‘australities”, etc., are of a dense black glass, rounded, irregularly

5 Merrill: The composition of stony meteorites compared with that of terrestrial igneous rocks, and
considered with reference to their efficacy in world making; Amer. Journ. Sci., ser. 4, vol. 27, 1909, p. 469.
Also: Report on researches on the chemical and mineralogical composition of meteorites, with special
reference to their minor constituents. Mem. Nat. Acad. Sci., vol. 14, memoir 1, 1916.

48 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

oval forms without distinct surface markings. Those of Moldavia
are of a green color, of extremely irregular forms, and gashed and
impressed in a manner to suggest they may, in plastic condition, have
been subject to mastication like a mass of chewing gum. The forms
from the island of Billiton are the most singular and unaccountable of
all. Like those of Australia, they are of a dense, opaque glass, their
distinguishing character being as in the last case, the peculiar surface
markings. Chemically, as shown in the accompanying table, they
average a trifle higher in silica, calcium, and ferric oxide than the
ordinary obsidian, and are poorer in alkalies. Their resemblance to
known meteorities in composition is remote. Each locality yields
its own peculiar forms, though all are grouped under the general
name of “tektite.’”’ In no instance have there as yet been dis-
covered facts relative to their occurrence such as can give a clue to
their origin. By some they are firmly believed to be glassy meteor-
ites. How far the difficulty of accounting for them otherwise may
have had influence in the formation of such an opinion the writer is
not prepared to say.
Analyses of tektites

Constituent I 10 Ill IV

Bi@pze SEAM ESS CUES 1 SEK oe ESOP Ae), SES EE TUL Ee Paes tee 75. 87 69. 80 76. 25 77. 96.
TAUTS © ope ee pe Se ag Ae ee ee SS a an eee 14. 35 15. 02 11.30 12. 20
Merah ss SLD > AE UO ME Ssh REET. Cae . 22 | . 40 35 14
MeO Beek eis ee STL TENSE AROS Te a Ce ri ee ere OL A) eee ease 4. 65 3. 88 3.36
Ss Ue O)S Me INE Tg AY oy Ng a Ce ate EP Te ee Ek Ae a 2 29 2.47 1. 48 1. 48
CRO Ae es BEES STE RUMEN RL LEE TONE Te Sy Eh YR | eel 3. 20 2. 60 1. 94
Nias OPE 2 SEE SESS: SA eee. SY Eee PETS 3. 96 1,29 1. 23 61
Key ©) eae DS ee a OR AU Rc ae oie we is bo RAR ONa | ee Ee, 4. 65 2. 65 1.82 2. 70
FRO FOO ALPES 2. S12 Es MEE PAY aS Ed TT ee ey Boulet sone 2) 22s eae
TG © TOORE ES EM AGT EE Lia SSE RRR YO ENG Dee | ae de (ei fe GAP eee ee
AT Ope Ree ey ee NSPE ST BR ADE REE REE PERT ie eee Ree Trace . 80 65) | 2 ae ae
VETO) SRE io TA ek Dee! AEE alee ae FLA LN Sk AS Bee .18 06 10
SOEECE ee Eee COT EERE Ee Pe CENA TEES METER eee i ee 23) |. Fase SLL es ree Be eer

NO tale FO EES ANE EEE EIS EES eee Ee ACE ERE! 99. 90 100. 46 99. 96 100. 49

I. Obsidian pebble. Colombia. Analyst, J. E. Whitfield.

Il. Obsidianite. Upper Weld, Tasmania. Analyst, W. F. Hillebrand.
III. Obsidianite. Near Hamilton, Victoria. Analyst, @. Ampl.
IV. Moldavite. Tribitsch, Bohemia.

COMBUSTIBLE METEORITES

The fact that there is so complete a lack of coordination between
the periodic meteoric showers and the fall of meteorites has more
than once suggested the possibility that the materials of the first
mentioned might be of an easily combustible nature, so that they were
consumed in their passage through the atmosphere, while only those
which were composed chiefly of metal or silicate materials survive,
and this only in part. The fact, too, that there is known a group of
carbonaceous meteorites (Orgueil, etc.) containing in certain cases as
high as 14 per cent of volatile matter, gives the suggestion a certain
degree of probability. In this connection, then, it may be well to

COMPOSITION AND STRUCTURE OF METEORITES 49

consider the often-reported occurrences of the fall of combustible
matter, even thoughin several of the instances there seems good reason
for doubting their absolute authenticity.

The earliest account to be mentioned is taken from the second
American edition of the New Edinburgh Encyclopedia, where it is
stated that a meteorite fell near Roa, in Borgus, Spain, in 1438 and
was reported to have been of very light material, resembling con-
densed sea froth. It is also said in the same publication that after
the fall of a fire ball in Lusatia in 1795 there was found on the ground
a viscous substance having the consistency, color, and odor of brown
varnish. This was examined by Chladni, who thought it to be com-
posed mainly of sulphur and carbon. One of the best authenticated
reports of this nature is that in connection with the fall of stones
near Hessle in 1869. This was accompanied by a quantity of carbo-
naceous matter of a coffee color, in the form both of powder and in
masses as large as the hand. When freed from the metallic substance,
it was found to consist of: Carbon, 51.6 per cent; hydrogen, 3.8 per
cent; oxygen, 15.7 per cent; silica, 16.7 per cent; ferrous oxide, 8.4
per cent; magnesia, 1.5 per cent; lime, 0.8 per cent; and soda with
traces of lithia, 1.5 per cent.

M. Meunier,® under the title ‘‘Substance singuliére recueillie a
la suite d’un météore rapporté 4 la foudre,”’ described a peculiar
resinous substance which was represented to him as having been
deposited over the suface of various objects during a violent thunder-
shower. The material is largely organic, burning with a resinous
odor, and differs from ordinary fulgurite in that it is of the same
nature regardless of the substance over which it is deposited, show-
ing at once that it is not derived by direct fusion. Meunier ex-
pressed a doubt as to whether or not the material is the effect of a
thunderbolt.

In the same volume,’ A. Tracul takes up the matter and claims
that the material is really a product of meteoric conditions, and says
that during a shower on the 25th of August, 1880, he saw issuing
from a black cloud a luminous body, very brilliant, slightly yellow-
ish but almost white, of an elongated form with the two tips in the
form of briefly attenuated cones. This body remained visible dur-
ing the brief time and then it disappeared, reentering the cloud, but
as it disappeared, there separated from it a small quantity of material
which fell vertically downward, as a heavy body under the influence
of gravity. It left behind it a luminous train, the edges of which
were manifestly reddish, sparkling globules. These, in the latter
part of their course, fell nearly vertically. Although none of this

6 Comptes Rendus, vol. 103, 1886, p. 837.
7 Idem, p. 848.

50 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

material was found, the writer (Tracul) regarded it as perhaps
identical with that described by Meunier.

The fall of a meteor or fire-ball, witnessed by Professor Dewey,
at Amherst, Mass., in August, 1819, is described by a Mr. Rufus
Graves.* In this case the object was described as being as large as a
man’s hand, and the conditions of observation were such that there
could seemingly be no doubt where it fell. This was in the evening.
Early on the ensuing morning there was found at this point ‘‘a sub-
stance unlike anything before observed by anyone who saw it,” and
which was regarded as beyond reasonable doubt the residuum of the
meteoric body. It was described as of circular form, resembling a
sauce or salad dish, bottom upwards, about 8 inches in diameter,
something more than an inch in thickness, and of a bright buff color.
Interiorly it was of a pulpy consistency like soft soap, with a suffocating
and very offensive odor. After brief exposure it changed into a livid
color resembling veinous blood. It shortly began to liquefy, and in
the course of a few days evaporated, leaving a small, dark colored
residuum which, when rubbed between the fingers, produced a fine,
ash-colored powder without taste or smell. Nitric and muriatic acid
seemed to have no chemical effect upon it, while with sulphuric acid a
violent effervescence ensued and nearly the whole substance dissolved.

This would certainly indicate that it was of organic and probably
fungoidal nature, and not meteoric.

Again, in the same journal (vol. 16, 1829), is given the translation
of an account of a like gelatinous material found in a wet meadow in
Germany and under such conditions that it was supposed to be
meteoric and was distinguished by the name “‘sterne-gallerte”’ (star
jelly). Counselor Doctor Schultes considered it as a ‘‘tremella
nostoc.’”? Buchner thought otherwise because he could discover no
organic structure, and maintained that it could be neither plant nor
animal, as a whole, but might be a product, like gum or mucus. The
writer of the article, Doctor Brandes, regarded the masses as either
animal excretions or gelatinous meteors, but he did not think it
probable that they were like the manna of the Israelites, as had been
suggested. A reference is made to the observations of a Spanish
soldier who, while standing on sentinel duty, during cool nights had
frequently observed shooting stars and in the morning, in wet places,
in spots where he thought the stars had fallen, he would find white,
gelatinous masses which soon dissolved. He quotes also the work of
a Mr. Schwabe, an apothecary of Dessau, who examined a gelatinous
mass found in a wet meadow, and who decided that it was the real
“nostoc commune” of Vauch. The writer enters into a somewhat
elaborate discussion of the chemical nature and general appearance of

8 Amer. Journ. Sci., vol. 2, 1820.

COMPOSITION AND STRUCTURE OF METEORITES ol

these bodies and suggests that they may have been of the nature of
snail spawn jelly, at any rate of organic and terrestrial nature. Ina
footnote to the article, reference is made to the observations of a
Mr. John Treat, who, while with the army of General Washington
during the Revolution, saw a shooting star fall within a few yards
of him. He immediately went to the spot and found there a gelati-
nous mass which ‘if we recollect right was still sparkling.”” Other
similar observations are given. ‘There would seem no question, how-
ever, but that these were in all cases of an organic (fungoidal) nature,
and of terrestrial origin.

A large share of these reported occurrences are doubtless due to the
observers having been mistaken in their identification of the fallen
material.

Within but a few months the writer was interviewed by a person
who brought a sample of meteoric material ‘‘seen to fall’? which
proved to be but a flat, widespreading fungoidal growth. No amount
of argument could convince the finder that he was in error.

Meteorites are not themselves magnetic, but nearly all meteoric
irons will acquire strong and permanent magnetism, though this
property may be in part, not wholly, destroyed by heat. It may be
remarked here, as noted elsewhere ® that in process of oxidation,
meteoric irons assume first a magnetic oxide stage before passing over
into the normal nonmagnetic sesquioxide.

CHONDRITIC STRUCTURES IN TERRESTRIAL ROCKS

The question is likely to arise, ‘‘Do these chondrules and chon-
droidal forms have any exact counterpart in terrestrial rocks?’”’ The
answer, with the information available to-day, is ‘‘No.” Neverthe-
less it will be well to consider a few cases which on first glance at
least, so closely resemble the meteoric chondrule as to merit attention.
Before entering upon this discussion it will be well, however, to refer
back, perhaps repeat in part, the writer’s dictum to the effect that
only those chondrules with smooth surfaces, often indented and show-
ing internally amorphous, microcrystalline, radiate, or barred struc-
tures, may have the origin ascribed to them by Sorby—are cooled
drops of molten matter having the chemical composition of enstatite
or of olivine. Those of a rough exterior and internally porphyritic
or holocyrstalline are products of the mechanical attrition of pre-
existing cooled rock masses.'°

A most suggestive example of chondroidai structure in a terrestrial
rock is afforded by a “kugelgriinstein’”’ found at Stefanschacht,
Schemnitz, Hungary. The stone is quite massive and of a light

§ Proc. U. S. Nat. Mus., vol. 24, 1902, p. 910.
10 Proc. Nat. Acad. Sci., vol. 6, no. 8, 1920 pp. 449-472.

o2 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

green-gray color, consisting of a dense ground carrying numerous
rounded chondrule-like bodies of a more compact texture and intern-
ally of a darker color, but which break away readily, often with por-
tions of the ground adhering. (Upper figure, pl. 30.) The ground
itself is so fine in texture as to render its mineralogical determination
by the unaided eye a matter of difficulty, though studded with small
whitish specks suggestive of a feldspar. The ‘‘kugels”’ vary in size
from but 2-3 mm. to rarely 20 mm. in diameter. When cut across
they show a darker ground than the matrix in which they are em-
bedded and are distinctly porphyritic even to the unaided eye. (PI.
30, lower figure.) In the thin section they show a normal, though
somewhat altered andesitic structure of feldspar and hornblende
phenocrysts in a microlitic ground. The kugels separate readily from
the matrix but are plainly not inclusions of foreign matter. Even by
the naked eye they may be seen on a polished surface to grade into
it without sharp lines of demarcation. From an examination of hand
specimens only, not having studied the occurrence in the field, one is
inclined to accept the conclusions of Szterinyi™ to the effect that they
are magnetic segregations liberated through the propylitic form of
decomposition which the stone has undergone. Whatever be their
origin their chondritic nature is wholly simulated.

Messrs. Cushing and Weinschenk, in their description of the phono-
lites of the Hegaus,’ mention an interesting tuff which in addition to
other constituents carries abundant kugel forms which are easily
distinguished through their dark color, hardness, and lustrous fracture.
These the microscope shows to be of melilite basalt, and plainly are
not rolled pebbles.

Sondern kleine Answiirflinge, wie sie itiberall einen integrirenden Bestandtheil

der Basalttuffe bilden und durch hiufig zu betrachtende centrische Structur
verrathen, dass ihre Form durchaus primar ist.

In their mode of occurrence, form, and structure they are described
as comparable with the chondrules of meteorites.

In response to a letter, the late Professor Cushing, then at the Case
School of Applied Science in Cleveland, kindly sent one of his original
specimens, with permission to use as much as might be necessary for
the investigation. This yielded the material from which the accom-
panying illustrations were prepared. In Figure 1 of Plate 31 a nucleal
augite is surrounded by a zone of elongated melilites which in a general
way are arrayed with their longer axes lying in the circumference of
a circle having the nucleus as a center. This is assumed to be the
‘“‘centrische”’ structure of the authors quoted. Be this as it may, the
centric structure while suggestive is not quite that of the majority

11 Foldtani Kozlony, Budapest, Nos. 12-13, 1882-83, pp. 207-222.
12 Tschermak’s Min. u. Nat. Mitheil., vol. 13, 1892, p. 36.

COMPOSITION AND STRUCTURE OF METEORITES 53

of chondrules. It is to be noted, however, that they break away from
the matrix leaving a smooth concavity simulating the chondrules in
the tuffaceous meteorites—spherulitic chondrites.

Among a series of specimens brought from the Hawaiian Islands
in 1920 by Dr. H. S. Washington were some loose aggregates of
fine volcanic ash labelled ‘‘fossil rain’? from the Kilauea eruption
of 1790. These are often pisolitic, strongly suggestive of the chon-
dritic structure so pronounced in meteorites of the Bjurbole type.
The entire mass, however, pisolites and all, quickly falls to pieces
when wet, and shows itself to consist of finely comminuted glass and
the silicate minerals characteristic of the lavas of this flow. The
chondroidal forms are entirely fragmental and the particles show no
order in their arrangement. Apparently they have originated in
place and are due merely to a haphazard aggregating of the finer
particles in the ash through the influence of water falling in the
form of fine drops such as would result through the condensation
of steam. Such forms are readily imitated in this same dry ash by
gently dropping into it small, isolated drops of water, and hence
the expression ‘‘fossil rain’? which, though on many accounts objec-
tionable, is expressive. The pisolites described above are apparently
of the same nature as those occurring in ash from the Krakatoa
craters and figured by Friedlander. Like forms are to be found in
ash from Pompeii. In these last the spherules are somewhat harder
and effervesce for a time when treated with a dilute acid, after which
they are readily reduced to a mud by crushing between the fingers.
Prof. J. A. Udden has described “ the formation of small pellicles
of a somewhat similar nature occurring in a volcanic ash in McPherson
County, Kansas. These he ascribes to wave action. However, their
resemblance to the meteoric chondrules is purely superficial.

A beautiful illustration of apparent chondritic structure is furnished
by the basaltic tuff from the “Anterior Lava Sheet’’ of Connecticut
described by W. G. Foye.'® The stone is fine grained, dark gray,
somewhat laminated, and shows scattered throughout its mass
abundant black spherules in varying sizes up to five millimeters.
These are a trifle rough on exterior surfaces, but break away easily
from the matrix leaving smooth cavities often lined with a portion
of the exterior shell of the spherule. In thin section they are plainly
fragmental, and show a thin outer zone or border of fine, dark material
enclosing the coarser, clastic silicate particles (see fig. 2, pl. 31) which
form the general groundmass of the stone. The structure on the
whole so closely simulates that of the ‘fossil rain drops’? noted
elsewhere, as to suggest a like origin for both. Be this as it may,

13 Zeit. fur Vulkanologie, vol. 1, Heft 1, 1914.
14 American Geologist, vol. 11, 1893, pp. 269-271.
15 Bull. Geol. Soc. Amer., vol. 15, no. 2, 1924, p. 335.

54 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

the chondritic character so evident on casual inspection, is not borne
out by close study.

A suggested resemblance to chondritic structure is found in the
peculiar drift bowlders found some years ago at Thetford, Vermont.
These were described by Dr. E. O. Hovey” as resembling conglom-
erate the most conspicuous feature of which is the numerous rounded
masses of granular olivine scattered through it. In addition are
numerous rounded pebble-like grayish green pyroxenes sometimes
reaching dimensions of several inches. The pebble form of the last
is however wholly assumed, the microscope and thin section showing
them to be crystalline secretions partially resorbed. The olivine
ageregates are, however, true inclusions in a crystalline ground of
augite and feldspars like those in the basalts of the Eifel and Rum-
berg, Bohemia; or the meteorite of Parnallee, India. On a gigantic
scale the Bohemian examples do bear a resemblance to the chon-
droidal forms found in some meteorites as that of Bjurbole. As,
however, the groundmass of the basalts is crystallime and glassy
there is no real connection.

Another very suggestive example of an imitattve form is shown in
Plate 32, Figure 2. The rock is a periodtite from a dike 2 miles east
of Raton, N. Mex., collected many years ago by Orestes St. John,
and with other collections recently turned over to the National
Museum by Dr. Frank Springer. Megascopically the rock shows a
dense dark greenish gray, nearly black ground thickly studded with
dark phenocryst and occasional large spherulitic forms and sizes up
to a centimeter in diameter. These are of olivine which sometimes
break away in form and manner strikingly simulating the meteoric
chondrule.

In thin section the rock is found highly altered but consisting
mainly of olivine and a rhombic pyroxene with abundant small
octahedra of pleonast and numerous colorless needles with the
characteristics of apatite, though their optical properties are wholly
obscured by decomposition. An abundant phosphomolybdate reac-
tion is produced when a drop of acid ammonium-molybdate is placed
upon the slide. The one time presence of a glass base is indicated,
but here too decomposition has gone too far for satisfactory deter-
mination. It can only be said that numerous interstitial areas are
nearly colorless and wholly isotropic.

The chondritic structure is found to be wholly simulated and is
produced by large oval and spherical olivines which have undergone
a partial serpentinous and chloritic alteration as shown in the figure.
This alteration begins with the formation of a thin coating (border
in the section) around the outside and all stages to complete altera-
tion of the granule. Such alteration is common, but rarely as in

16 Trans. New York Acad. Sci., vol. 13, 1893-94, p. 161.

COMPOSITION AND STRUCTURE OF METEORITES do

this case does the resultant forms so simulate those of chondritic
meteorites.

Other descriptive matter suggestive of chondritic structure is
found in the literature but the writer has been unable to secure
examples of the material upon which the descriptions are based.

Rinne ” describes chondrule-like forms occurring in a nephelin-
basalt tuff from the Hussenberges in Westphalia. He says (p. 243):

Es liegen eine grosse Anszahl rundlicher Auswiirflinge schon in der Ebene
eines Diinnschliffes bei einander. Sie zeigen in ihrer braunlichgelben Glasmasse
scharfe Einsprenglinge von Olivin und auch monoklinen Augit. Solche Bildun-
gen sind immerhin in ihrer allgemeinen Erscheinung porphyrischen Chondren
vergleichbar. Ein Unterschied liegt zwischen beiden darin, dass bei den
vorliegenden, iridischen Bildungen das Glas mit vielen Blasenriumen versehen
ist. Es fehlt ferner z. B. die strahlige Chondrenstructur die man wohl nur bei
auch chemisch und mineralogisch den Chondriten dihnlichen Gesteinen erwarten
kann.

The accompanying figure (fig. 1, pl. 32) from Rinne’s paper is so
similar in structure to the porphyritic kugels (chondroids) in many
meteoric stones as to strongly suggest an identity in composition and
origin.

Von F. Schalch, in the 19th lieferung of the Beitrige zur Geolo-
gischen Karte der Schweiz, pages 103-104, describes what is like-
wise apparently a chondritic phonolite tuff. He says:

Die sonst gleichartige Tuffmasse fihrt an zahlreichen Stellen * * *
runde Kugelchen von Erbsen—bis Haselnussgrésse, selben von noch betracht-
licheren Dimensionen, die aus einem festeren und auch 6fters etwas verschieden
gefairbten Substrate bestehen und dem Gestein eine ausgeprigte Pisolithstructur
verleihen.

And again on page 105:

Man findet dieselben von ganz verschiedener Grésse; gewohnlich sind sie
indess nicht viel iber nussgross, seltener faustgross oder von noch bedeuten-
deren Dimensionen. Meist besitzen sie ziemlich scharfe Kanten und Ecken,
bisweilen sind sie aber auch mehr oder weniger abgerundet die grésseren Stiicke
hie und da zersprungen und deren Theilfragmente wieder durch Tuffmasse mit
einander verkittet. Zuweilen findet man sie nur vereinzelt, an anderen Stellen
nehmen sie derart an Menge zu, dass sie neben den oben genannten grésseren
Krystallfragmenten und Pisolithkugeln geradezu die Hauptmasse des Tuffes
bilden.

Gumbel, in his Geognostische Beschreibung des Kdénigreichs
Bayern (Abt. ITI, p. 226), described a ‘‘schalstein”’ which apparently
is of like kugel form in structure. He says:

Bei diesen Abanderungen besteht das Gestein bald mehr, bald weniger vor-
herrschend aus kleinen und kleinsten Brocken von verschiedenartigem oder doch
in verschiedenen Stadien der Umbildung begriffenem Diabas mit vorherrschend

abgerundeten Umrissen, welche durch eine Zwischenmasse nach Art des ge-
wohnlichen Schalsteins, vorherrschend durch die chloripitische Substanz ver-

17 Neues Jahrbuch fur Min., etc., vol. 2, 1895, pp. 229-246,

506 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

kittet sind. Die Diabasbréckchen, welche sichtlich durch abrollung ihre abge-
rundete Form erhalten haben, sind meist stark zersetzt, lassen jedoch noch
deutlich die Texturverhiltnisse des Krystallinisch kornigen Gesteins in der
Verschiedenartigkeit des Diabases wahrnehmen. * * * Dazu gesellen sich
zuweilen Brocken oder knollenartig rundliche Stiickchen mehr oder weniger
verinderten Nebengesteins, namentlich Fornfelsihnliche Fragmente und kugelige
Ausscheidungen, welche beim Durchschlagen die Zusammensetzung der als
Perldiabas beschrieben Varietiten besitzen. Diese enthalten haufig im Centrum
grossere Putzen von verindertem Gestein, um welche sich mit nach aussen
abnehmender Haufigkeit einzelne kleine Perlen oder Knéllchen der verinderten
Substanz in zonen- oder schalenahnlicher Anordnung anlegen. (Berneck.)

The spherulites of the acid volcanic glasses like those of the Yellow-
stone National Park, in appearance and mode of occurrence often so
closely simulate the kugels of meteorites as to demand attention here.
Indeed Klein '® affirms that those kugels with radiate structure:
‘“‘Kehte spharolithe darstellen.”’ Berwerth, too, it may be recalled
compared the kugels in meteorites to the spherules in artificial
glasses, an error to which I have elsewhere called attention.”

The term spherulite, as now used by the leading petrologists, it
may be said, refers to the complex growths of minerals of one or more
species, principally miscroscopic quartz and orthoclase with minor
quantities of tridymite, colloidal silica, and minute forms of other
accessories characteristic of highly siliceous igneous rocks. ‘The
essential character of spherulitic growth is the crystallization of
minerals from one or more points with a radiating or divergent ar-
rangement.” (Iddings.) In external form they are often beauti-
fully spheroidal, though not necessarily so, and break from the matrix,
with which they are practically identical in composition, as freely as
do the chondrules from a meteorite of the spherulitic chondrite type.
In thin sections between crossed nicols they show a black cross which
is due to the elongated form and radial arrangement of the principal
constituents. That they originate in place and their formation is
but a peculiar phase of magmatic crystallization is unmistakable.
The question of their origin throws no light upon that of the granular
and porphyritic chondrules, and seemingly none upon those of the
eccentric radiating type.

Brief reference should be made to other occurrences of kugellike
forms in terrestrial rocks.

The kugels of the spheroidal granites of Donigal, England, West-
moreland, Sweden, Quonochontogue Beach, R. I., and those of
Finland, recently described by Sederholm *” all show a marked con-
centric structure, rarely if at all simulated by the meteoric chondrule.
The nearest approach to these forms that has thus far come under

18 Studien tiber Meteoriten, 1906, p. 35.
19 Bull. Geol. Soe. Amer., vol. 32, 1921, p. 409.
2 Bull. Com. Geologique de Finland, No. 83, 1928.

COMPOSITION AND STRUCTURE OF METEORITES 57

my observation are those described * in the stones of Parnallee and
Tennasilm. Even here, however, the resemblance is slight, consisting
in the case of the latter in an outer zone of crystals surrounding a
granular interior. The kugel gabbros as those of Slittmossa, Norway,
and Dehesa, Calif., show a variable granular nucleus surrounded
by more or less radiating and elongated minerals of the same nature
as those in the main rock mass. This holds true also of the kugel
diorites of Corsica and Davie County, N. C. The structure and
origin of these is so obviously different from those of the true chondrule
as to render discussion unnecessary.

A further striking and decisive distinction between the meteoric
and terrestrial forms lies in the fact that in the last named the
spherules in any rock mass are invariably of a single type only. In
meteorites, two or more types, varying in mineral composition and
degrees of crystallization usually occur in the same stone and often
in close juxtaposition.

Finally, and so far as relates to the kugelchen, or spherulitic
chondrites there is a difference in that the meteoric chondrules are
in all cases foreign to their hosts. Those of the terrestrial rocks,
with the exception of the Thetford conglomerate, are formed in
place.

METHODS OF ANALYSES OF STONY METEORITES

The making of satisfactory chemical analyses of stony meteorites is
attended with difficulties little appreciated by one accustomed only
to ordinary silicate work. This is due to several causes, the chief of
which is the presence of iron in its several forms (as metal alloyed with
nickel and cobalt, as phosphide, sulphide, and carbide and also in
ferrous and ferric combinations in the silicates) and the desirability
of determining each of these in its individual percentage amount.
On first thought the magnetic separation of the iron might seem in
all cases most feasible, but long experience has shown the impossi-
bility of complete separation and many analysts have resorted to the
expedient of solvents, as digestion in mercuric chloride. The follow-
ing actual analysis illustrates the method adopted by Museum chem-
ists in the analysis of the Florence, Tex., stone.” It is in substantial
agreement with that used by Prior of the British Museum and J. E.
Whitfield of Philadelphia.

A piece of the stone weighing 19.08 grams was crushed to pass
80 mesh. The larger particles of metal which would not pass the
sieve were pounded flat to free them from silicate as far as possible
and reserved. The powdered material was then worked over thor-

21 On Chondrules and Chondritic Structure. Proc. Nat. Acad. Sci., vol. 6, 1920.
22J. T. Lonsdale, American Mineralogist, Vol. 12, no. 11, November, 1927.

58 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

oughly with a hand magnet and all attracted material was removed
and added to the metal. This separation gave:

Grams
Attracted: portion...) 4-225 252 etek ee eee Ae eee 3. 8425
Unatéracted. portion.2t 42 345 a5 Sie ats eee Meee ee ieee ec ee 15. 2362
19. 0787

The attracted portion was all digested in aqua regia, evaporated
to dryness, taken up in hydrochloric acid and filtered. The undis-
solved portion was washed thoroughly in hot water, then digested
for one hour in hot 10 per cent sodium carbonate solution, filtered,
and the then insoluble residue washed thoroughly with hot water,
then with dilute hydrochloric acid and again with hot water. The
residue was weighed as insoluble silicate. This gave:

Grams

DISSOI VEG ts lid aee Wa uke ep A Ne oe lB ne eo. St ee ee ee 3. 5501
TMTN SC Lita ess Sk Rs . 2924
3. 8425

The alkaline and acid filtrates were combined, evaporated to dry-
ness, taken up in hydrochloric acid, and the silica separated and
determined in the usual manner. The solution was then made up
to 500 cc. in a calibrated flask at 38° C. and divided into aliquot
portions. Portions of 50 cc. (equivalent to 0.35512 gm.) were used
for determination of MnO, P, and $8; 100 ce. (0.71024 gm.) for the
portion used for Fe, FeO, Ni, CO, CaO, Al,O;, and MgO, and the
remaining 250 ce. (1.7756 gm.) was used for determination of Cu.

The portion for sulphur was evaporated to approximate dryness
on the steam bath, an excess of hydrochloric acid and of potassium
chloride added and twice again evaporated to approximate dryness
with concentrated hydrochloric acid to expel nitrates. The potas-
sium chloride unites with the ferric chloride to form a crystalline
double salt, thus facilitating evaporation. The material was then
taken up in hydrochloric acid, precipitated with barium chloride and
weighed as barium sulphate on a small Gooch crucible.

The portion for manganese was evaporated to dryness several
times with strong nitric acid, taken up in nitric acid, diluted and
boiled with bromine, precipitate with ammonia. The precipitate
was dissolved in nitric acid, thus freeing from chlorides, oxidized with
silver nitrate solution and ammonium persulphate, and determined
colorimetrically.

The portion for phosphorous was likewise evaporated with nitric
acid several times to free from chlorides, precipitated with ammonia,
dissolved in nitric acid and precipitated with ammonium molybdate
reagent. It was filtered on a small Gooch crucible and weighed
as phosphomolybdic anhydride, 24MoQ3.P,0s.

COMPOSITION AND STRUCTURE OF METEORITES 59

The portion for the main analysis was precipitated twice with
ammonia to separate iron and alumina from the other constituents.
The nickel was then precipitated in the ammoniacal filtrate with
dimethylglyoxime. The filtrate from the dimethylglyoxime precip-
itate was evaporated to dryness on the steam bath and treated
with strong nitric acid to expel ammonium salts and destroy the
excess of the dimethylglyoxime reagent. This was necessary before
separating cobalt, as the element can not be precipitated as sulphide
in the presence of dimethylglyoxime, although this compound does
not interfere with the determination of lime as oxalate or magnesia
as phosphate in the usual manner. The dry residue from the evap-
oration was evaporated once more with hydrochloric acid, taken
up in dilute hydrochloric acid, made ammoniacal, saturated with
hydrogen sulphide and filtered. The filtrate after freeing from H.S
was used for the determination of lime and magnesia. The precip-
itate of cobalt sulphide, somewhat contaminated with iron and some
other impurities was ignited, was digested in acid and precipitated
with ammonia, the cobalt passing into the filtrate in very small
volume of solution from which it was recovered and weighed as
CoSOQx.

The remaining acid solution was evaporated to free from nitrates
and precipitated with H,S. The precipitated impure copper sul-
phide was digested in a porcelain crucible, the solution made am-
moniacal and the impurities filtered out. The filtrate was again
acidified with hydrochloric acid and again precipitated with H,S,
ignited in a small tared porcelain crucible, digested in the crucible
with a few drops of nitric acid, evaporated to dryness, and weighed
as CuO. No platinum could be detected in the amount of material
used. The results of analysis of this portion were as follows:

Soluble silicate: Grams
SLO) pers OR AME ee Be PF Se ba ta ae 0. 0404
Al,O3 See SS eee Se Seat See eee a Se ee Trace.
Ue T CO) seated EOE yah Na nn mad HOR el see Trace.
Ie) A ap oe ae ER A LOE IR Tr SED A PDN ear romeo sie 28 A . 0187
CW SO ayes ae kB Et RL Ge PS Ey Sg ait Trace.
ITs) ath a ple n hTe y eipa A Ne ateeah d . 0435
IN Gh a SA cy er 1 A a Tg SO re A el 2 0. 1026
Troilite:
A eee ste Sry, 5S NN oY Be RI ee RR EL coe ees at . 0151
SS ee ORL re 1 A at SRM ged gL Ue Fe ok 0087
SER Ercan memene y DAN NE Ve Riel pe Ea EN I By oy by Bes yoy cee ae 0238

59587—30——_5

60 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

Metal: Grams

Ts hg ee ah yea EP NON Ds SE a NNER Fk SO Oud Sin ey Reale eae 7 eR 3. 0680

ING 2 Go a ema oe SOR ne A a . 2780

Cee EUAN AN EAD UE RE, AEA GRY A ee Se . 0143

CS UE a Pr ED A SE OPT EEO Se ED CEN . 0003

TR ea aks SB a ol Sek ia a Al A ays ikea ly a an . 0006
PROG. 2 Rath A A os ORDA ERR LUMENS TR2 R OG Ra ee Re Sroole
Grand Otel ae, See A econ ia Re alt Sate na ANS, eS 3. 4876

The sulphur is all stated as S and enough Fe deducted to form
troilite with all of it. The phosphorous is stated as P and the iron
sufficient to form orthosilicate with the excess of Si over MgO is
calculated to FeO. The metal then is found to make up 17.62 per
cent of the whole meteorite and to have the following composition:

Net composition of metallic portion

Per cent

Trome(He) i. ee ei tege fot a tena ugh ee Aas lap pags Daly hel De Ee ahs 91. 277
ING LN) iene es et eI Re OR OIL Lion Neg aS An Oat ee 8. 270
WobalGe(Co) he Seis She eos eae ee Ie cata Need er Ree phe erie wes oS . 426
(COPPCD ICC) io: 2k ic See eps ee rah ne 2a Pat Thc eee ee es ae . 009
PHOSPHOROUS: (Eee ote eee eae Sit ae eee nc Le, can, Sr Re a oa ea, . 018
100. 000

THE SOLUBLE SILICATE

For the analysis of the soluble silicate portion, 2.0000 grams of the
unattracted powder were taken and treated in the same manner as
the preceding by digestion in aqua regia followed by hydrochloric
acid and sodium carbonate solution. The results were:

Grams
Dissolved fui tee (sabe wy pcb fy eR bie) le cag tn A as ane Os aa tl epee Bea a 0. 9137
Mira Scola es Sh ie Ne MN Ne ile Se Ae a eae Ue aU 1. 0863

After separation of the silica the acid solution was made up to 250 ce.
in a calibrated flask at 38° C. Portions of 50 cc. were taken for
Mn, P.O;, and S. The remaining 100 cc. was used for FeO, Al,O;,
CaO, NiO, MgO, etc. The results obtained were as follows:

Composition of soluble silicate portion

Grams Grams
RSL) ya seh pate ge pas De UL Re OM ZS BERS O git ole oes nea ROE Rn rere . 0065
DIST) papeheretr Serene, Lacuna Geen ieee ae racers in Cae oS aire ey eee ei aU . 0784
SANTA 0) pte ate i ly gl ya NE BHO OS Oa PINT Gee ee, Sane ere ane ce eer te . 0035
Bie Ge eae One ape eee Bin 2 NGOS! [PS eee ae ee oats ne ac a . 0450
Ca Oe ere SA hr a Rs 0150
Wee eae alle ha Resw ig hs _ 3290 - 9218
Dry NO) yee tea eye Ce 0135

Per cent, 100.83.

Except for enough to combine with the sulphur to form troilite
the iron is calculated as FeO. The Ni is all determined as NiO in
this portion and the phosphorus as P.O.

COMPOSITION AND STRUCTURE OF METEORITES 61
THE INSOLUBLE SILICATE

From the last treatment there remained slightly over 1 gram of
clean acid washed insoluble silicate. One-half gram (0.5000) of this
was weighed out, fused with sodium carbonate and analyzed by
standard methods to give the composition of the insoluble portion.
The chromite was undecomposed by the sodium carbonate fusion
and was weighed with the silica. It remained in the residue when
the silica was treated with hydrofluoric and sulphuric acids. After
weighing this residue was again treated in the platinum crucible
with hydrofluoric and sulphuric acids, evaporated to sulphuric acid
fumes, diluted with water and the chromite filtered out and weighed.
It was not in amount sufficient for analysis and, in the summation
below it is assumed to have the composition FeO.Cr,0;. The analy-
sis of the insoluble silicate gave the following results:

Analysis of insoluble silicate portion

Per cent Per cent
VINO) escape 0 ak 2 ie teh Sea ei c POLE LO ES Sp Aa HINT GO eae na en en ey ae Nee eda None.
PU AO) eet ee ee Ci ee ae I ASSGaih OHTOMIGe is ee ee, See 0. 60
MICE) yhoo Acne ee OER ay a SRA ABTACOS a IN Onstae n Soe tee Fam ove ret 87
SR ere Ete eae eee ak Bs 2 SSO 8 5 hig OMe As See Ore a Tees 1. 28
ROO ae dee: Pei A 4. 56
WA Ce) renee 2 oS pip eis Na ee DOES 100. 37

No trace of nickel could be detected by the dimethylglyoxime
method. Alkalies were separatively determined by the J. Lawrence
Smith method and all assigned to the insoluble silicate. The direct
determination of alkalies in the remnant of the insoluble silicate was
considered impractical, since the material had been digested in sodium
chloride and might have retained a minute amount of soda. This
remainder was therefore used for microscopic examination.

The composition of the portion of the meteorite ground for analysis
is then:

Weight Weight

(grams) Per cent (grams) Per cent
pO ney saree 2 os BS 6. 8903 SERA Co ase wee ee 0. 01438 0. 075
eTeR CO) sot ot Trace. er Aceh a @ sees 2s eres . 0003 . 002
1 IES @ cies fella . 5258 DRS Oi es eee eases are ae . d015 1. 842
Be @ se = 2 = oak 2. 0755 MOM SOE ie kere ees ates ee va ay . 0006 . 003
Ma Oa a ee . 5308 DEES ail Ope . 0495 . 259
IVS OEE week 82 ka 4. 4594 BESS TAU)N | Pe Gas 0 Ps ee ee . 0350 . 183
ein) th Ss ae . 1028 BS Oba MeN CO) ereenn eae es sya . 1094 . 513
Ce ee . 0267 LA OM |PRINias (0) eee ete” Se . 0742 . 388
HE Pah ae eer Ee 3. 6807 19. 290 ae
Nesey tomes atl ee ees . 2780 1. 492 19. 2048 100. 689

The unused acid and alkali extracted portion of the insoluble
silicate was screened free from fines and the portion between 80 and
200 mesh was passed through a methylene-iodide-bromoform heavy
solution in an attempt to isolate grains of feldspars for optical exam-

»

62 BULLETIN 149, UNITED STATES NATIONAL MUSEUM

ination. Only an exceedingly small light portion was obtained. This
consisted of grains of feldspar containing numerous inclusions of the
silicate of higher index, probably pyroxene for which the feldspar
forms a matrix. Practically none of the feldspar grains showed twin-
ning and most of them had a refractive index, beta, of about 1.550,
varying about 0.002.

The unused portion of the unattracted powder was likewise screened
free from fines and passed through heavy solution. The lightest
product separated from this likewise consisted of only a few mixed
grains consisting of a feldspar with very abundant rounded inclusions
of pyroxene and olivine. The feldspar showed no twinning and is
optically positive with 2V medium, beta index from 1.550 to 1.553.
The feldspar thus appears to be a fairly sodic plagioclase approaching
andesine in composition. Occasional grains of the feldspar-bearing
material are heavily pigmented with a black substance which is
probably largely carbon.

The heavier concentrate from the insoluble silicate consists of a
few clear grains which exhibit moderately high bi-refringence. <A
majority of these were aggregate in structure, greatly dusted with
inclusions and gave very little optical data. The clearer grains were
biaxial and positive with a beta index of about 1.680. No cleavage
was observed and only one or two grains showed recognizable poly-
synthetic twinning. The optical properties are those of diopside.

O

U. S’ NATIONAL MUSEUM BULLETIN 149, PLATE 1

EXTERNAL FORM OF ALLEGAN AND BATH FURNACE STONES;
AND OF MAZAPIL IRON

FOR EXPLANATION OF PLATE SEE PAGE 18

U. S. NATIONAL MUSEUM BULLETIN 149, PLATE 2

(1), CARBON NODULE IN CANON DIABLO METEORIC IRON; (2), RE-
FUSED FELDSPAR IN ESTHERVILLE METEORIC STONY IRON

FOR EXPLANATION OF PLATE SEE PAGES 3, 4, AND 43

BULLETIN 149, PLATE 3

U. S. NATIONAL MUSEUM

titan

P>
a

8 =
mY "
art id

oy

MERRILLITE IN

(2),

MASKELYNITE IN TROUP METEORIC STONE;

(1).

NEW CONCORD METEORIC STONE

FOR EXPLANATION OF PLATE SEE PAGES 5 AND 7

U. S. NATIONAL MUSEUM BULLETIN 149, PLATE 4

(1), FRAGMENTAL TWINNED PYROXENES IN JOHNSTOWN METEORITE;
(2), FRAGMENTAL PYROXENES IN ESTHERVILLE METEORITE

FOR EXPLANATION OF PLATE SEE PAGE 11

U. S. NATIONAL MUSEUM BULLETIN 149, PLATE 5

(1), WIDMANSTATTEN FIGURES, ENLARGED; (2), SCHREIBERSITE IN ARISPE
AND (LOWER) IN TOMBIGBEE IRONS

FOR EXPLANATION OF PLATE SEE PAGES 8, 13, AND 19

U. S. NATIONAL MUSEUM BULLETIN 149, PLATE 6

(1), METEORIC IRON FROM DUNGANNON, VA., SHOWING PARTIAL
GRANULATION; (2), THE SAME GREATLY ENLARGED

FOR EXPLANATION OF PLATE SEE PAGE 20

U. S. NATIONAL MUSEUM BULLETIN 149, PLATE 7

(UPPER), SWATHING KAMACITE IN ADMIRE PALLASITE; (LOWER), OCTA-
HEDRAL STRUCTURE AND REICHENBACHIAN LINES IN CARLETON IRON

FOR EXPLANATION OF PLATE SEE PAGES 19 AND 23

U. S. NATIONAL MUSEUM BULLETIN 149, PLATE 8

(UPPER). NEUMAN LINES IN BRAUNAU IRON; (LOWER), GRANULAR STRUC-
TURE IN MEJILLONES IRON

FOR EXPLANATION OF PLATE SEE PAGE 20

U. S. NATIONAL MUSEUM BULLETIN 149, PLATE 9

(UPPER), OCTAHEDRAL IRON WITH SILICATE INCLOSURES, FOUR CORNERS,
N. MEX.; (LOWER), PERSIMMON CREEK, N. C., IRON

FOR EXPLANATION OF PLATE SEE PAGES 12, 20, AND 21

U. S. NATIONAL MUSEUM BUREN MAS Ie ArinE e110

(UPPER), BRECCIATED HEXAHEDRITE, KENDALL COUNTY, TEX.; (LOWER),
COARSEST OR KAMACITE OCTAHEDRITE, AINSWORTH, NEBR.

FOR EXPLANATION OF PLATE SEE PAGE 20

U. S. NATIONAL MUSEUM BULLETIN 149, PLATE 11

(1), SWOLLEN KAMACITE IN MESA VERDE IRON; (2), STRUCTURE OF FOUR
CORNERS IRON, ENLARGED; (3), QUARTZ CRYSTALS IN ST. MARKS STONE;
(LOWEST), CHONDRULES IN SHARPS, VA., STONE

FOR EXPLANATION OF PLATE SEE PAGE 14

BULLETIN 149, PLATE 12

U. S. NATIONAL MUSEUM

GRAHAMITES FROM MORRISTOWN AND CRAB ORCHARD STONY IRONS

FOR EXFLANATICN CF FLATE SEE FAGE 21

U. S. NATIONAL MUSEUM BULLETIN 149, PLATE 13

(UPPER), ESTHERVILLE STONY-IRON; (LOWER), SAME WITH INCLOSURE

FOR EXPLANATION OF PLATE SEE PAGES 15, 22, 23, AND 43

BULLETIN 149, PLATE 14

U. S. NATIONAL MUSEUM

PALLASITE FROM BRENHAM, KANS.

FOR EXPLANATION OF PLATE SEE PAGES 15 AND 23

U. S. NATIONAL MUSEUM BULLETIN 149, PLATE 15

(UPPER), SECTION CF THE ADMIRE PALLASITE; (LOWER), AN IRON-
RICH PORTION OF BRENHAM PALLASITE

FOR EXPLANATION OF PLATE SEE PAGE 15

U. S. NATIONAL MUSEUM BULLETIN 149, PLATE 16

MICROSTRUCTURE OF, (1), EL NAKHLA STONE; AND (2), OF THE
SHERGOTTY STONE

FOR EXPLANATION OF PLATE SEE PAGES 5, AND 25

149, PLATE 17

BULLETIN

U. S. NATIONAL MUSEUM

BRECCIATED STRUCTURE OF THE, CUMBERLAND FALLS, AND (LOWER),

(UPPER).

ST. MICHEL STONES

FOR EXPLANATION OF PLATE SEE PAGES 25 AND 26

U. S. NATIONAL MUSEUM BULLETIN 149, PLATE 18

CHONDRULES AND CHONDRITIC STRUCTURE OF SELMA, ALA., AND CEDAR,
TEX., STONES

FOR EXPLANATION OF PLATE SEE PAGES 27, 28, 29, 31, 37, AND 39

U. S. NATIONAL MUSEUM BULLETIN 149, PLATE 19

(UPPER), MICROSTRUCTURE OF THE BLUFF, AND (LOWER), ESTA-
CADO STONES

FOR EXPLANATION OF PLATE SEE PAGES 28 AND 39

U. S. NATIONAL MUSEUM BULLETIN 149, PLATE 20

MICROSTRUCTURE OF, (1), INDARCH STONE, AND (2), OF CULLISON
STONE, SHOWING A COLLAR OF METAL ABOUT A FRAGMENTAL
CHONDRULE

FOR EXPLANATION OF PLATE SEE PAGES 29, 39, AND 41

U. S. NATIONAL MUSEUM BULLETIN 149, PLATE 21

(UPPER), SLICE OF ANTHONY METEORITE SHOWING TROILITE WITH BORDER
OF METAL AND (LOWER). SECTION OF DARK INCLOSURE OF CUMBERLAND
FALLS STONE SHOWING: DISTRIBUTION OF METAL

FOR EXPLANATION OF PLATE SEE PAGES 14, 29, AND 41

U. S. NATIONAL MUSEUM BULLETIN 149, PLATE 22

CHONDRULES IN BARRATTA, CULLISON, ELM CREEK, HESSLE, AND PARNALLEE
STONES

FOR EXPLANATION OF PLATE SEE PAGES 10, 11, 31, 32, 38, 39, AND 42

U. S. NATIONAL MUSEUM BULLETIN 149, PLATE 23

(UPPER), OLIVINE CHONDRULES IN BEAVER CREEK AND CULLISON STONES;
(LOWER LEFT), ENSTATITE CHONDRULE IN CULLISON: (LOWER RIGHT),
TWINNED ENSTATITE CHONDRULES IN PARNALLEE STONE

FOR EXPLANATION OF PLATE SEE PAGES 10, 31, 32, 33, AND 42

BULLETIN 149, PLATE 24

U. S. NATIONAL MUSEUM

2) TENNASILM: AND (3) ELM

CREEK STONES
FOR EXPLANATION OF PLATE SEE PAGES 31, 32, 33, AND 43

(

CHONDRULES IN (1) PARNALLEE;

U. S. NATIONAL MUSEUM BULLETIN 149, PLATE 25

(1 AND 2), CHONDRULES AND CHONDROIDAL FORMS FROM THE BJURBOLE
STONE, (3), BROKEN SURFACE OF BJURBOLE STONE, ABOUT NATURAL SIZE

FOR EXPLANATION OF PLATE SEE PAGES 27, 31, AND 38

PLATE 26

BULLETIN 149,

U. S. NATIONAL MUSEUM

AFTER ROASTING

)

2

(

BEFORE AND,

es

FOR EXPLANATION OF PLATE SEE PAGES 40 AND 41

(1

’

IRON

TOLUCA

U. S. NATIONAL MUSEUM BULLETIN 149, PLATE 27

(1), CHONDRULE IN PARNALLEE STONE WITH SECONDARY GLASS BORDER;
(2), SAME SHOWING EFFECTS OF CRUSHING; (3, 4, AND 5), CHONDRULES FROM
HENDERSONVILLE, TROUP, AND ENSISHEIM STONES SHOWING EFFECTS OF
CRUSHING

FOR EXPLANATION OF PLATE SEE PAGES 40 AND 42

U. S. NATIONAL MUSEUM BULLETIN 149, PLATE 28

(1), SECTION THROUGH CRUST OF ALLEGAN STONE; (2), SECTION THROUGH
BLACK VEIN IN BLUFF STONE

FOR EXPLANATION OF PLATE SEE PAGES 18, 44, AND 45

149, PLATE 29

BULLETIN

U. S. NATIONAL MUSEUM

(LOWER), SHAT-

TERED STONE FROM CEDAR, FAYETTE COUNTY, TEX.

(UPPER), FAULTED METEORIC IRON. NEW BALTIMORE, 'PA.;

FOR EXPLANATION OF PLATE SEE PAGES 18, 20, 44, AND 46

U. S. NATIONAL MUSEUM BULLETIN 149, PLATE 30

(UPPER), BROKEN SURFACE OF ‘‘KUGELNGRUNSTEIN,”’
SCHEMNITZ, HUNGARY, SHOWING CHONDROIDAL
FORMS; (LOWER), CUT AND POLISHED SURFACE OF
THE SAME

FOR EXPLANATION OF PLATE SEE PAGE 52

U. S. NATIONAL MUSEUM BULLETIN 149, PLATE 31

(1), MICROSECTION OF PHONOLITE TUFF, HEGAU, GERMANY, (AFTER
CUSHING); (2), CHONDROIDAL FORM IN BASALTIC TUFF, (AFTER
FOYE)

FOR EXPLANATION OF PLATE SEE PAGES 52 AND 53

U. S. NATIONAL MUSEUM BULLETIN 149, PLATE 32

(1), CHONDROIDAL FORMS IN NEPHELINE BASALT, HUSSENBERGES, WEST-
PHALIA, (AFTER RINNE); (2). PSEUDOCHONDRITIC FORMS OF OLIVINE IN
PERIDOTITE, RATON, N. MEX.

FOR EXPLANATION OF PLATE SEE PAGES 54 AND 55

HSONIAN INSTITUTI

iii
2286

9088 01421

‘wl