Submitted in Partial Fulfillment of the Requirements For

Home , 409 Aspasia, Alais meteorite, Angrite, Beardsley meteorite, Brianite, Chaoite

A SYSTEMATIC STUDY OF THE SPECTRAL REFLECTIVITY

CHARACTERISTICS OF THE METEORITE CLASSES

WITH APPLICATIONS TO THE INTERPRETATION OF ASTEROID SPECTRA

FOR MINERALOGICAL AND PETROLOGICAL INFORMATION

MICHAEL J. GAFFEY

B.A., University of Iowa (1968)

M.S., University of Iowa (1970)

Submitted in

Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

in the

Department of Earth and Planetary Sciences

Massachusetts Institute of Technology

February, 1974

Signature of Author

Department of Earth and Planetary Sciences, January 7, 1974

Certified by ,~,

Thesis Supervisor

Accepted by

on A SYSTEMATIC STUDY OF THE SPECTRAL REFLECTIVITY CHARACTERISTICS OF THE IETEORITE CLASSES WITH APPLICATIONS TO THE INTERPRETATION

OF ASTEROID SPECTRA FOR MINERALOGICAL AND PETROLOGICAL INFORMATION by

Michael J. Gaffey

B.A., University of Iowa (1968) M.S., University of Iowa (1970) Submitted in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy in the Department of Earth and Planetary Sciences Massachusetts Institute of Technology

January 7 , 1973

ABSTRACT

The purpose of this thesis is to study the spectral reflectivity of a natural, cosmically occurring series of solid materials (meteorites) and to discuss the applicability of this information to the interpretation of asteroid spectra for the mineralogical and petrological compositions of the asteroid surfaces.

A detailed study of the transmission spectra of oriented olivine crystals shows that the interaction of a photon with the groundstate orbital of a transition metal ion is a simple function of the shape of the orbital (distribution of electron probability) and the vibration vector of the photon. The features in the reflection spectrum of a silicate material can be correlated with crystal field effects in the mineral phases. The compound features in the reflection spectra can be calculated using the centers and halfwidths of the absorptions due to each transition in the individual mineral phases. Physical mixtures of minerals are spectrally dominated by the dielectric phases with the highest optical density. Particle size variations strongly affect the albedo of a particulate surface but only weakly affect the relative reflectivity curve over a wide range of particle sizes.

Asteroidal surfaces should be fragmental mixtures of the underlying material. The particle size distribution and glass content of the surface should have minimal effect on the relative reflection spectra, such that variations between the spectral reflectivity curves of asteroids are primarily functions of petrological and mineralogical variations between the surfaces of these bodies. The meteorites represent a mineralogical and petrological series which should share many of the compositional characteristics of asteroidal material.

Measurements were made of the spectral reflectivity curves of 156 meteorite specimens from essentially all meteorite classes. Examination of high quality spectra (unaltered specimens) from these data has shown that each meteorite class has a characteristic spectral reflectivity curve. Varia- tions between spectra of different materials can be understood in terms of the abundance and distribution of mineral phases, metamorphic grade and shock history of the specimens. The characteristic curves can be compared directly to narrow bandpass filter measurements of asteroid spectra to determine the mineralogy and petrology of the asteroid surfaces. A lithological interpretation of several asteroid spectra indicates a wide variety of materials are present. The relative proportions of these materials are totally unlike the proportions of meteorite types observed in terrestrial falls.

Thesis Supervisors Thomas B. McCord Acknowledgements

Scientifically, there are at least four individuals without whose active and expert cooperation, this thesis would not have been possible in anything like its present form or on the relatively short timescale of its completion. I would like to acknowledge the contribution of each of these person separately.

Dr. Thomas B. McCord was my advisor throughout this effort and supplied not only the essential scientific and financial assistance, but also provided a great deal of advise and encouragement. Dr. McCord, perhaps more importantly, provided an astronomical framework within which this work could be developed and exploited to the utmost. He also provided valuable critical evaluations of the preliminary versions of this thesis. Dr. Roger G. Burns provided not only a great deal of insight into the physics of crystal field theory and the nature of compositional data contained in transmission spectra, but also provided the wherewithall to undertake an experimental program to correlate the known physics of transmission spectra to the measured characteristics of the reflection spectra. Dr. Burns also provided invaluable critical assistance in the preparation of this manuscript, both in form and content. Dr. John B. Adams provided the equipment required to make the spectral reflectivity measurements of my specimens as well as providing a great deal of insight into the nature of the reflection spectrum of a material and the parameters which govern variations in these spectra. His previous work with the reflection spectra of rocks and minerals provide an excellent background for my work.

Dr. Edward Olsen, of the Field Museum, provided the bulk

of the major samples utilized in this study. A whole day of his

time was spent going through the museum's collection, literally from Abee to Zhoutnevyi, allowing me to chose the specimens

most needed for my work. Indeed, he supplied specimens which

I would have probably found impossible to obtain from any other source. He also provided a great deal of information into the

nature and relationships of meteoritic materials which helped

greatly in my understanding of this subject.

On a more personal basis, irreplacable and invaluable support was provided by Susan (Jenks) Gaffey, my fiance and wife during this period. The work could probably have been

done without her aid and understanding, but it might not have been worth while.

A great many other indiviuals provided scientific or technical input which greatly expedited this work. Their contributions to this thesis are many and varied. Dr. Clifford Frondel and Mr. David Cook provided a large number of meteorite specimens from the Harvard collection which provided an excellant starting selection. Dr. Carleton Moore of Arizona

State University and Dr. Karl Turekian of Yale University also provided meteorite specimens to flesh out this work.

Dr. John Lewis provided a great deal of insight into the nature of meteoritic materials and possible models for their 5 origins. Dr. Clark Chapman provided both a critical review of portions of this work and a large amount of practical information concerning asteroids and the asteroidal spectra. Dr. Robert

Huguenin assisted my understanding by sharing his knowledge about charge tranfer effects in silicates.

Ms. Carle Pieters and Mr. Michael Charette both provided a forum for discussing the work and various results which helped greatly in separating the wheat from the chaff.

Mr. George Fawcett and Mr. Lawrence Bass both provided invaluable aid in my battles with the Big Blue Box (IBM 360-65). With their aid I managed to win the war. To all of the above individuals and to any whom I might have inadvertantly omitted, I offer my very sincere gratitude

(and maybe even a trip to Cabots). Table of Contents

Abstract

Acknowledgements Table of Contents

List of Figures

List of Tables List of Appendices

Chapter Is Asteroids and Meteorites: Background a) Introduction I-1 b) Purpose of 'this study I-2 c) Asteroids as members of the Solar System I-4 d) The choice of meteorites to represent the asteroidal surface materials I-7 e) Other materials for comparison I-11 References - Chapter I 1-14

Chapter II: Previous spectral reflectivity work a) Meteorites and natural silicates - Laboratory measurements II-1 b) Asteroids - Observational work II-4 c) Mineraloeic and petrologic interpretation of asteroid spectra II-7 References - Chapter II II-8 Chapter III: Spectral reflectivity of natural silicate materials: Internal Effects a) Introduction III-1 b) Crystal Field Theory III-1 c) Crystal field effects measured in minerals III-5 d) An ular variations of photon interaction III-8 e) Angular variation in photon interaction - Experimental III-10 f) Photon interaction: Implications and discussion III-18 g) Summary and conclusions III-22 References - Chapter III III-22 Chapter IV, Spectral reflectivity of natural materialss external Effects a) Introduction IV-1 b) The nature of light reflection - dielectric materials Iv-1 c) Particle size IV-2 d) Particle packing effects IV-5 e) Illumination angle effects IV-6 f) Mixing of mineral phases IV-7 g) Metal phases IV-10 h) Conclusion IV-12 References - Chapter IV IV-12

Chapter V: Atteroid surfacess Physical and lithological characteristics a) Introduction V-1 b) Surface texture and materials: Dynamic considerations V-1 c) Surface texture and composition - Obser- vational Data V-13 d) Conclusions V 14 References - Chapter V

Chapter VI: Meteorites: Mineralogy and petrology a) Introduction VI-1 b) Meteoritic minerals VI-2 c) Meteorite types: Irons VI-5 d) Meteorite types: Stony-irons VI-8 e) Meteorite types: Stones - Chondrites VI-10 f) Meteorite types: Achondrites VI-21 g) Alternate meteorite compositions VI-27 References - Chapter VI VI-28 Chapter VI: Experimental Procedures a) Sample selection VII-1 b), Sample preoaration VII-5 c) Spectroreflectometer - Description and operation VII-7 d) Data reduction and display VII-8 References - Chapter VII VII-9 Chapter VIII: Meteorite class spectral characteristics a) Introduction VIII-1 b) Eucrites, howardites and diogenites (Basaltic or hypersthene achondrites) VIII-2 c) Anrites (augite achondrite) VIII-4 d) Nakhlites (diopside-olivine achondrites) VIII-4 e) Chassipnites (olivine achondrite) VIII-5 f) Aubrites (enstatite achondrites) VIII-5 g) Ureilites (olivine-pigeonite achondrites) VIII-7 h) Octahedrites and nickel-rich ataxites VIII-? i) Mesosiderite (pyroxene-plagioclase stony-iron) VIII-8 j) Chondritess Introduction VIII-9 k) Enstatite chondrites VIII-10 1) Ordinary chondrites VIII-11 m) Bronzite chondrites (H-type) VIII-12 n) Hypersthene chondrites (L-type) VIII-13 o) Amphoterites (LL-type) VIII-13 p) Black chondrites VIII-13 q) Carbonaceous chondrites (C-type) VIII-14 r) Alternate meteorite composition - Anorthosite VIII-16 s) Conclusions VIII-17 References - Chapter VIII VIII-18

Chapter IX: Interpretive applications a) Introduction IX-1 b) Philosophical approach IX-1 c) Validity of comparing asteroid and laboratory data formats IX-3 d) Comparison procedures IX-4 e) Conclusions and implications IX-10 References - Chapter IX IX-10 Appendices List of Figures

Fig. I-l, Distribution of asteroid semimajor axes. I-5a Fig. II-ls Color index distribution of rock types II-la

Fig. II-2s U-B, B-V Distributions on meteorites, lunar and terrestrial rocks. II-2a

Fig. II-3: Spectral reflectivity curves of several terrestrial rocks, minerals and meteorites. II-3a

Fig. II-4: Spectral reflectivity curves of several terrestrial minerals. II-3b

Fig. II-5: UBV Color index plot of the asteroids. II-6a Fig. 11-6: Spectral reflectivity curves of several asteroids. II-7a

Fig. III-1: Boundary surfaces of atomic orbitals. III-3a

Fig. III-2: 3d-orbitals in an octahedral coordination site. III-3b

Fig. III-3: a) Relative enerzy levels of d-orbitals III-3c of a transition metal ion in octahedral co-ordination b) Relative energy levels of d-orbitals of a transition metal ion in co-ordination sites of various symmetries.

Fig. III-4s a) Polarized absorption spectra of III-5a fayalite. b) The polarized absorption spectra of olivines resolved into component Gaussian-shaped bands. Fig. III-5s Compositional variations of absorption III-6 ma ima in the polarized spectra of 2 + MaZ+ - Fe olivines.

Fig. III-6: Variations in the polarized absorption III-9a spectra of fayalite with changes in orientation of each vibration axis.

Fig. III-71 Spectral reflectivity curves of olivine. II-9 Figr. III-8s Rotation axes and light paths with respect to the crystallographic axes. III-lOa Fig. III-9abcde: Variation of absorption feature III-11a-a-j in olivine spectrum during rotation of photon vibration vector. List of Fimures (continued)

Fig. III-lOabcde: Variation of olivine absorption III-15a-e feature between adjacent transmission spectra( e8' )

Fig. III-11: Energy level diaErams for olivine III-17a absorption feature.

Fig. III-12abcde: Variation of olivine absorption III-18a-e feature with rotation spectral regions.

Fig. III-13: The M(2) Cation site in olivine. III-19a Fig. IV-ls a) Variation of albedo with mean particle IV-2a diameter. b) Spectral reflectance curves representative of glasses, crystalline acidic rocks and crystalline basic- ultrabasic rocks.

Fig. IV-2: Schematic representation of variation in IV-4a the spectral reflectivity of a particulate surface with particle size.

Fig. IV-3: Spectral reflectance curves illustrating IV-5a of different packing of > 40 micron powders.

Fig. IV-4, R/B variations with incident angle fot IV-5a sifted powders of size fraction 12 ( > 40 microns).

Fig. IV-5s Spectral reflectivity curve of ordinary IV-7a chondrite compared with the component mineral phases.

Fig. IV-6 s Schematic production of meteorite spec- IV-9a trum from component mineral spectra.

Fig. IV-7: Production of artificial C-type chondrite IV-9b spectrum.

Fi,. IV-8, Normalized spectral reflectivity of four IV-lla metals.

Fig. V-ls a) Ratio of mass lost from planet to mass V-3a of projectile (basalt target). bl Ratio of mass lost from planet to mass of projectiles unbonded sand target, hypothetical velocity distribution 2. c) Ratio of mass lost from planet to mass of projectile, Unbonded sand target, hypothetical velocity distribution 3. List of Fieures (continued)

Fig. V-2: a) Fraction of mass ejected at speeds V-3b greater than V for hypothetical ejecta velocity distributions. b) Differential ejecta velocity distribution.

Fig. V-3s Grain size distributions for several V-9a lunar soils.

Fig,. V-4: Spectral reflectivity of glass made from V-10 12063 whole-rock powder; mixtures of rock powder plus 20% glass and plus 55% glass; compared with curve of Apollo 12 surface fines.

Fig. V-51 Spectral reflectivity curves of several V-14a asteroids.

Fig. VI-1 Frequency of distribution of nickel con- V-5a tent in analyses of iron meteorites. Fig. VI-2: Relationship between oxidized iron and VI-1la iron as metal and sulfide in analyses of observed falls, illustrating the separation into distinct subgroups and variation within the subgroups.

Fig. VI-3: Population of the Van Schmus and Wood VI-13 chemical-metamorphic subtypes of the chondrites. Fig. VI-4, Frequency distribution of the mole per VI-15a cent ratios Fe/(Fe+mg) in olivines for the chondrites.

Fig. VI-5, Iron contents of olivines and low- VI-15a,b calcium pyroxenes in a) four unmeta- morphised or slightly metamorphosed ordinary chondrites, b) two moderately unequilibrated ordinary chondrites and c) a nearly equilibrated chondrite (Tennasilm) and a completely equilibrated chondrite (Modoc).

Fig. VI-6s Frequency distributions of the mole per VI-15c cent ratios Fe/(Fe*Mg) in rhombic pyroxene in the chondrites.

Fig. VI-7s Histograms of mineral compositions of VI-19a 10 C2 chondrites and one C3 chondrites. Fig. VII-l1 Comparison of an unrusted chondrite, a VII-2a rusted chondrite and iron rust spectra. 12 List of Fieures (continued)

Fig. VIII-ls a) Spectral reflectivity curves of Type 1 VIII-3a eucrites. b) Spectral reflectivity curves of Type 2 VIII-3b eucrites.

Fig. VIII-2s Band-Band plot for eucrites, howardites VIII-3c and diogenites.

Fig. VIII-3s Spectral reflectivity curves of howardites .0 VIII-4a

Fig. VIII-4: Spectral reflectivity curves of diogenites . VIII-4b

Fig. VIII-5s Spectral reflectivity curves of an angrite .0 VIII-4c Fig . VIII-6, Spectral reflectivity curve of a nakhlite. VIII-5a Fig. VIII-7s Spectral reflectivity curve of a chassig- VIII-5b nite.

Fig. VIII-8 Spectral reflectivity curves of aubrites. VIII-6c Fig. VIII-9s Spectral reflectivity curve of a ureilite. VIII-7a Fig. VIII-lOs Spectral reflectivity curves of several VIII-7b iron meteorites. Fig. VIII-li Spectral reflectivity curve of a VIII-8a mesosiderite.

Fig. VIII-12: Spectral reflectivity curves of ensta- VIII-lOa,b tite chondrites: a) Type E4&E5; b) Type E6

Fig. VIII-13s Spectral reflectivity curves of bronzite VIII-12a,b,c chondritess a) H3 & H4; b) H5; c) H6.

Fig. VIII-14s Band-Band plot for ordinary chondrites. VIII-12d Fig. VIII-15: Effect of metamorphic grade on the s VIII-12e strength of the absorption feature.

Fig. VIII-16: Spectral reflectivity curves of hyper- VIII-13a,b,c,d sthene chondrites: a) Type L3; Type L4; c) Type L5; d) Type L6.

Fig. 17 Spectral reflectivity curves of ampho- VIII-e,f teritic chondrites: a) Type LL & LL5 b) Type LL6. Fig. VIII-18: Spectral reflectivity curves of VIII-13g black chondrites. List of Figures (continued) Fig. VIII-19s Spectral reflectivity curves of VIII-14abcd carbonaceous chondritess a) Type Cl; b) Type C2; c) Type C30; d) Type C3V. Fig. VIII-20: Spectral reflectivity of anorthitite VIII-16a feldspar from lunar soil.

Fig. IX-l Spectral reflectivity curves of meteorite IX-4a-1 types for comparison to asteroid spectras a) Eucrites - Type 1; b) Eucrites - Type 2; c) Howardites; d) Diogenites; e) H-type chondrites; f) LL-type chondrites; g) L4; h) L5; i) L6; j) Black chondrites; k) Cl; 1) C2; m) C30; n) C3V; o) En- statite chondrites; p) aubrites; q) Chass- ignite; 4) Nakhlite; s) Ureilite; u) mesosiderite; and v) iron meteorites. Fig. IX-2 Spectral reflectivity curves of asteroidss a) 1 Ceres IX-6a b) 2 Pallas IX-6b c) 3 Juno IX-7a d) 4 Vesta IX-7b e) 12 Victcria IX-Sa f) 16 Psyche IX-9a 14 List of Tables

Table I-l: Distribution of asteroid semimajor axes I-5a Table 1-2: Distribution of asteroid orbital inclina- I-5b tions

Table I-3s Radiant and orbital elements of three I-10 Prairie Net fireballs, the Pribram meteorite, and four high-density meteoroids

Table III-ls Intensity of absorption feature at 11-13 maximum (@=900) as a function of rotation axis (6) Table III-2: Centers and halfwidths of Gaussian bands III-14 a resolved from transmission spectra.

Table IV-ls Spectral reflectivity (0.56 ) of various IV-5 meteorites.

Table V-ls Escape velocity (meters/sec) for spherical V-5a body of a given radius and density.

Table V-2: Densities of solar system materials. V-6

Table V-3: Dynamic characteristics of nine asteroids V-7 with IT measured diameters.

Table VI-1 Meteoritic minerals (as of 1962) VI-3a Table VI-2s Meteoritic minerals discovered since 1962. VI-3b

Table VI-3s Sinmmary of characteristics of chemical- VI-12a petrologic chcndrite subtypes.

Table VII-l, Types and number of meteorites measured. VII-1 Table VIII-l, MYodal mineral data for eucrites, VIII-2a howardites and diogenites.

Table IX-la Characteristics of six asteroids used IX-5 in this study. 'W1--l~psrrrr~-- i -- rr.r^--r-~*.....------~iu;rrr~~r~~~-arc~---~

List of Appendices

Appendix I: Meteorites measured in this study. Appendix II: Spectral reflectivity curves of meteorites measured in this study.

Appendix IIIs Spectral reflectivity of meteorites measured in this study in the McCord 24 filter system and the UBVRI system. 16. I-i

I) Asteroids and Meteoritess Background a) Introduction

Ever since the discovery of Ceres in 1801 and the subsequent discoveries of several thousand other asteroids, these bodies have provided a particular fascination to astronomers. The existence of a belt of asteroids at the approximate position of a planet predicted by the Titius-Bode expression for planetary distances from the sun (Bode's Law) has lead to end- less speculation as to their origin and significance in the solar system. Are they the fragments of a planet disrupted by tidal interaction with Jupiter, by collision or (even more intriguing) by some intelligent agency run amok? Or are they fragments of condensate from the solar nebula which were never allowed by disruptive tidal forces to coalesce into a single body? Or more reasonably, are they simply the junk left over from the formation of the solar system which happens to lie in a region which is relatively stable against perturbation and disruption of a population of objects in this region? What- ever conclusion is finally reached, these objects offer a tan- tilizing glimpse into a distant past, which has been preserved nowhere else in our solar system.

Scientific interest in asteroids and meteorites lies mainly in the possibility that they are keys to the origin of the solar system. It is as if one has the last chapter of the I-2 novel "War and Peace", one knows how it came out, but one cannot be certain just what it was that came out. Terres- trial geology provides one with information on recent planetary processes. All hints of the starting conditions have long since been buried, eroded or shoved down a trench. Even the very ancient lunar rocks have been extensively remelted and altered and, as such, provide information about a planet still- born but otherwise complete. The asteroids can provide scientists with the details of planetary evolution at the moment of conception. Such concrete data will force the re- evaluation of the appropriate cosmological theories, just as the recent low measured value for the neutrino flux from the sun has forced a new look at certain 'well understood' nuclear processes going on inside the sun. b) Purpose of this study Since spacecraft missions to even a few asteroids are subjects of speculation rather than concrete planning (if for no other reason than budgetary considerations), direct physical data on these bodies will not be available in the reasonable future. In the interim, a very significant amount of information can be obtained by the utilization of remote sensing techniques. One of the most promising of these is the interpretation of asteroidal spectral reflectivity measurements to obtain mineralogical and petrological information about the surface material of these objects. McCord et al (1970) have 1-3

shown, for example, that the spectrum of the asteroid Vesta

very closely resembles that of a basaltic achondritic meteorite. Chapman et al (1973) have measured the spectra of a large number of asteroids and are involved in a continuing program

of this work. Johnson and Fanale (1973) have attempted to correlate a number of Chapman's asteroid spectra with labora-

tory spectra of carbonaceous chondritic meteorities. Hunt and Salisbury (1973) and Chapman and Salisbury (1973) have attempted

empirical correlations between meteorite and asteroid spectra for a limited range of meteorite types.

At this stage what is sorely needed is a systematic study of meteorite spectral reflectivities to produce a catalog to be used in this interpretation. Such a catalog should incor- porate a basic understanding of the physics of reflection spectra, so that the information contained in the spectra can be exploited to the utmost. Also, the reliability of such information must be determined so that it can be applied judiciously and any ambiguities or redundancies can be taken into consideration. And, finally such a catalog should be as complete as is physically possible. It should include a spectrum of every known type of meteorite and of as many specimens of each type as can be acquired. The physical condition of each specimen must be taken into consideration in order to eliminate any effects introduced terrestrially. The purpose of the present work is to compile such a catalog and to define a set of parameters for the spectral I-4

reflectivity curves of the meteorite classes which can be

used with assurance in the interpretation of asteroid spectra.

The sensitivity, limits and ambiguities of these parameters

are defined as well as the physics giving rise to the variations in the parameters. The applicability of each parameter to an asteroid surface and the conditions of that surface which may lead to uncertainties in these parameters are discussed. Finally, a preliminary interpretation of a number of asteroid

curves is made, to demonstrate the usefulness of the meteorite

spectra, and the implications of the results discussed.

c) Asteroids as Members of the Solar System The term asteroid, from the Greek 'starlike', refers specifically to the enormous number of small bodies, ranging

in diameter from a fraction of a kilometer up to a thousand

kilometers, most of which circle the sun between the orbits of Mars and Jupiter. An estimate of the number of these bodies larger than about one kilometer diameter, is on the order of a few hundred thousand. These bodies are planet-like in that they exhibit a non-transient aspect (appear as solid

bodies) as opposed to objects such as comets which exhibit very prominent transient features (tails). The asteroids are generally considered to be composed of metallic or silicate materials, although certain asteroids with orbits having

large semi-major axes may have volatile materials (solid I-Sa 20 Figure I-1 DIstributin of Asteroid Sexirajor Axes

S k - 1 :Fr. .j.-- .-- q. . - . -M :

sal .. . ."

3.1

-os 4o 4

.- Frequency distribution of semimajor axes of asteroids in intervals of 0.001 AU. Sample consists of the 1647 numbered asteroids given in the 1962 Ephemeris volume minus the 13 asteroids that are marked as "lost." The following five fall outside the diagram: 1566 Icarus, 1620 Geo-raphos, 433 Eros, 279 Thule, and 944 Hidalgo with orbital radii a of 1.077, 1.244, 1.458, 4.282, and 5.794 AU, respectively. Commensura- bility points are marked with aows, together with the ratios of periods (asteroid/ Jupiter). n Qcarin Huten, .L 1 j.)

Table I-1 Distribution of Astero id Seminajor axes

B(1, 0) < 10 Zone Number of Observed !imits, Median a, Median Number per asteroids a AU AU B(a, 0) - B(1, 0) Number unit circle of ecliptic plane

M... 15 1.077 to 1.c79 1.879 1.39 0 0 ... 150 2.153 to 2.256 2.225 2.17 3 1 .. 240 2.257 to 2.490 2.385 2.59 39 2 .. 310 2.520 to 2.705 2.618 3.13 67 3.. 200 2.708 to 2.{16 2.762 3.43 58 4. 129 2.834 to 2.956 2.887 3.68 23 5.. 122 2.964 to 3.030 3.011 3.91 17 6. 41 3.033 to 3.074 3.056 3.99 13 7. 345 3.075 to 3.263 3.151 4.16 74 8 .. 46 3.280 to 3.756 3.411 4.57 18 9 .. 20 3.887 to 3.998 3.948 5.33 8 T. 14 5.095 to 5.277 5.188 6.68 b11 L Total.. 1632 331

IVan Houten, 1971) 21 I-5b

Table I-2 Di~stributioni of Asteroid O-"ital Inclinations Observed Frequency Distributionof Inclinations in Eight Zones

[Data from class I orbits in the PLS]

i distributions for zones 0 to 8

0 1 2 3 4 5 6 7 8 1. 0 to 3 11 1 - 3 2 - 17 1 38 38 1 to 2 13 43 21 7 10 7 1 19 - 121 121 2 to 3 19 68 30 9 14 4 2 26 - 172 172 3 to 4 . 17 38 37 14 11 4 3 14 1 139 154 4 to 5 . 25 39 24 17 1 3 1 20 - 130 222 5 to 7 14 70 24 20 1 - 1 12 - 142 351 7 to 9 5 19 12 13 2 7 2 6 - 66 219 9 toll . - 5 6 8 2 13 6 12 2 54 202 11 to13 . - 5 27 - 7 3 1 5 - 48 237 13 to15 . - - 22 6 - - 1 5 - 34 196 15 to 9 . - 2 - 1 1 - 6 - 10 128 19 t23 - - 1 - - - - 5 - •6 49 23 to27 .. - - - - - 1 - 1 9 27 to35 ------

So ...... 96 298 207 94 52 44 18 148 4 961 - Sl ...... 162 532 546 212 157 113 48 319 9 - 2098 Zo.5 ..-. 0.105 0.124 0.237 0.206 0.171 0.291 0.307 0.259 - - -

Z0 .9 5 ...... 256 .351 .629 -585 .630 .611 .636 1.011 - -

(Van Houten, 1971) 22

solutions of water, methane, ammonia or sulfur dioxide) as stable components.

These bodies are collectively grouped together in a region termed the Asteroid Belt, which extends roughly between 2.0 and 4.0 AUS from the sun. However this term is a misnomer because a discontinuous distribution of objects is found throughout this region, as shown in figure I-1 and

Table I-i, with gaps (Kirkwood's Gaps) at distances corresponding to orbits with small fractional periods of that of Jupiter.

The orbits of these objects have an eccentricity range 0.0-

0.4, with most less than 0.2. The planes of the orbits have (Table I-2) inclinations to the ecliptic plane ranging up to

350- A detailed discussion of the distribution of the astroids can be found in Williams (1971) and Kiang (1971).

There are discrete groups of asteroids which can be lumped together on the basis of some shared common characteristics. These have been divided into groups which have a dynamic cause and those which have no obvious dynamic cause. Those with a dynamic cause ares a) Trojans (bodies in stable orbits at the Lagrangian points L 4 and L5 of the Sun-Jupiter system (i.e. 600 ahead and behind Jupiter in its orbit) and b) commensurability groups (groups whose members' orbits have semi-major axes such that their periods are commensurate with Jupiter, for examples Hungaria group - 2s9; Hilda group -

213t Thule group - 314). Of the dynamic groups, only the

Trojan group have any significant probability of having 23 I-6

genetically related members.

The groups which have no obvious dynamic cause ares a) Hirayama families (groups of asteroids with similar values of a, 1' and e' and may be genetically related)s b) Brouwer

groups (subgroups of the Hirayama families which have nearly constant values for the suniXp' +Xn' (longitude of the proper perihelion and longitude of proper node)) and c) jetstreams (members have nearly coincident orbits). The validity of these last two groups has been brought into question by indications that they may result from selection effects in the Palomar-Leiden Survey (PLS). The Hirayama families, however, seem to have real genetic relationships among the members. A more complete discussion of these groups can be found in Van Houten (1971). Two other groups of asteroids which are not included in the above list, the Amor and Apollo asteroids, are important in this work. These asteroids are small both in size (typically a few kilometers and hence subject to discovery only during close passage to earth) and in number (only a score are known). However, the orbits of these asteroids (the Amor group have perihelions >1 AU. the Apollo group have periheliohs

I-7

The relationship between comets and asteroids is becoming better understood. If the 'dirty iceberg' model for a cometary nucleus is accepted, it leads to the conclusion that if the non-volatile components are not carried away by the vaporizing 'ices', then in time one is left with only these non-volatile components in the nucleus. At this point the difference between this nucleus and an asteroid is purely academic. Sekanina (1971) has discussed the evolution of a cometary nucleus into an asteroidal object. Marsden (1971) has discussed several asteroidal objects (P/Arend-Rigaux and

P/Neujmin) which upon close passage to the earth were discovered to exhibit very slight but quite definite cometary activity. Some of the Apollo or Amor groups may also be decayed cometary nuclei, Icarus and Hidalgo being prime can- didates.

d) The Choice of Meteorites to Represent the Asteroidal Surface Materials The choice of the meteorites (discussed in Chapter VI) as objects for an extensive study as comparison material for the asteroids is based on several reasons. Initially, the meteorites represent the only material available on the earth's surface which is of undoubted cosmic origin and which has some indication of being connected with the asteroids. The meteorites seem to represent the cosmic debris left over from 25 I-8

the formation of the solar system. Since this is the most

widely accepted explanation for the existence of the asteroids, these two classes of objects should be similar in composition. This should not be construed to mean that it is believed that the range of meteoritic materials contained in terrestrial collections represent the entire petrological range present as asteroidal bodies. Nor does it necessarily follow that

all the petrologic types found as meteorites will be found in the form of asteroidal sized bodies. One of the goals of this line of research is to determine the interrelationship

between the meteorites and the asteroids, by testing (a) whether they are portions of the same population of objects,

differing only in mass and orbit; (b) whether they represent

two discrete populations of objects; or (c) whether the meteorites are selectively derived from certain asteroids. More-

over, a detailed study of the spectral reflectivity characteristics of all known meteorite classes will allow one to distinguish between asteroids with surfaces of meteorite-like

mineral assemblages and those with surfaces of other petrologic assemblages. This might be termed the direct approach. With the judicious application of principles of crystal field theory, information on the reflectivity of non-meteoritic materials and the effects of surface interactions, reasonable assignments (or significant eliminations) of the surface mineralogies of these objects with non-meteorite-like 26. 1-9

spectra should be possible.

It is perhaps more accurate to deal with four classes of objects rather than just two. These are meteorites,

meteors, asteroids and comets. A proper definition of a meteor is just the interaction phenomenon (usually a bright, ionized trail) of a meteoroid with the atmosphere at ex-

tremely high velocities. However recent work suggests that the meteoroids responsible for the vast majority of meteors are distinctly different than most of the materials that reach the earth's surface as meteorites. It has been known for a very long time that the majority of meteors occur in showers associated with the orbits of decayed comets. The meteoroids in question are the non-volatile materials freed by the sub- limation of a 'dirty iceberg'. Ganapathy et al (1970) concluded from a study of the contamination of the lunar soil by meteoritic material, that most of this material, derived from cometary debris, is a very primitive carbonaceous chondritic material. McCrasky and Ceplecha (1970) found that the bulk densities for most of the meteors observed by the Prairie Network was between 0.1 and 1.5 gm/cm 3 , averaging about 0.5 gm/cm 3. These densities are quite consistent with those of sponge-like carbonaceous chondritic material, especially if the material is type C01 or even more primitive.

Since the overall impression of these very primitive meteorites is that of 'mud', it is quite reasonable to suppose that less 27

I-10

consolidated material of the same basic composition is the

result of the vaporization of the 'ice' from a 'frozen mud'

or 'dirty ice'. This implies that the asteroids which are the solid nuclei of extinct comets will have compositions

similar to the most primitive material found in terrestrial meteorite collections, samples of which are used in this study. The direct connection between meteoritic material and the asteroids is based on orbital elements of meteorites determined from the measured path upon entering the atmosphere as photographed by either the Prairie Network (USA) or the All-Sky Network (Czechoslovakia and West Germany). McCrasky et al (1971) have determined the orbital elements of several meteorites and high density meteoroids (Table I-3) which indicate that the aphelions are not inconsistent with bodies

Table 1-3 (McCrasky et al, 1971)

Radiant and Orbital Elements of Three Prairie Net Fireballs, the Pfibram Meteorite, and Four High-Density Meteoroids.

True Radiant

a, a V., a, q, q', o, , i, deg deg km/sec AU e AU AU deg deg deg Lost City 315.0 39.1 14.2 1.66 0.417 0.967 2.35 161.0 283.0 12.0 40617 62.1 37.6 13.2 2.02 0.516 0.976 3.06 193.5 310.8 3.3 40503 18.0 -- 17.2 21.0 2.02 0.642 0.722 3.31 73.1 15.6 12.6 Pfibram 191.5 17.7 20.9 2.42 0.674 0.790 4.03 241.6 17.1 10.4 Iron meteor 332.0 73.2 13.4 1.03 0.117 0.928 1.17 247.5 219.1 13.3 Harvard 1242 3.0 59.2 12.2 1.33 0.262 0.984 1.68 172.1 317.1 6.9 Harvard 19516 70.5 -0.2 20.7 2.24 0.662 0.756 3.72 65.4 75.7 12.6 Harvard 7946 257.6 25.4 18.5 2.49 0.621 0.943 4.04 215.5 103.9 18.1

perturbed out of the asteroid belt. Lowrey (1971) has dis- 28

I-Il

cussed the orbital evolution of the Lost City meteorite and

has concluded that its origin cannot be definitely ascer-

tained. Thus, it is unclear whether it is an independent remnant from the formation period or whether it is a fragment

spalled off an asteroid sized body. Williams (1973) has calculated which asteroids are most likely to produce earth crossing material and finds that near certain orbital re-

sonances such production is very likely. The conclusion

reached is that these meteorites belong to the same dynamic

population as the asteroids. The extent of any genetic relationship is explored later in this work, but it is safe to say that there should be a significant degree of overlap in

'rock type' between the two groups. Anders (1971) has discussed the origins of meteorites and asteroids and the significance

of the population of meteorites arriving on the earth. He has pointed out that the present distribution of meteorites could be accounted for by a few (*10) parent or source bodies.

e) Other Materials for Comnarison

The meteorites thus present at least a very good starting material for comparison to the asteroids. However the range of material for comparison can be extended by considering the mineral phases present in the meteorites. Since certain meteorites (see Chapter VI) are specimens of essentially pure mineral phases which are present in other meteorites, it is 29

1-12 reasonable to suppose that concentrations of almost any phase present in a meteorite can occur. This does not imply that such quantities of pure mineral phases will exist as to form asteroidal sized bodies, but merely indicates that these are. reasonable comparison materials. For example, the presence in the eucrite family (pyroxene-plagioclase meteorites) of the very plagioclase rich meteorite, Serra de Mage, would make it reasonable to postulate that nearly pure anorthositic plagioclase would be a reasonable comparison material. The other major type of compositions which can occur in the Asteroid Belt, besides the very primitive carbonaceous type material, are the solid solutions ('ices') of water, methane, ammonia and hydrogen sulfide or any combination thereof. These are cosmologically abundant compounds and are the most likely 'ices' to exist, if the conditions permit. Lebofsky (1973) is completing an extensive laboratory study of the spectral reflectivity characteristics of these substances under a variety of conditions. His results will be considered more completely in the chapter on asteroid interpretation. Finally, two notes of caution should be injected. The postulate that exotic or esoteric chemical compounds are major components of these bodies should be forestalled until there is experimental evidence that the range of compositions described above is insufficient to account for the observed 30

1-13

range of spectral reflectivity curves for the asteroids. To introduce laboratory curiosities not found in natural materials thus far observed to be of cosmic origin, is to introduce a

needless degree of complexity to the problem without assurance that one is also introducing any useful expansion of the

validity of the interpretations. To carry this type of logic to its conclusion, one would eventually find oneself measuring the spectral reflectivity of rye bread (salty,

seeded or unseeded). Thus for purposes of this study and its subsequent applications, we shall limit ourselves to those

natural materials (or permutations thereof) observed in the meteoritic or cometary objects. The second note of caution, is to avoid going over-

board in the other direction and requiring that all asteroid spectral reflectivity curves must be interpretable in terms

of these limited range of materials. Such interpretations should be applied using these substances only as long as they provide reasonable matches to the asteroid curves. In most such cases, a hierarchy of petrological types, in order of probability, can be established. However if nothing fits well to the observed curve, then this should be acknowledged

and alternate materials examined for this purpose. Either overstringent or too lax restrictions on the range of comparison materials can decrease the value of this interpretation technique. 31

1-14

References

Anders, Edward (1971) Interrelations of meteorites, asteroids and comets. NASA SP-267, p. 429

Chapman, C.R.; T.B. McCord and T.V. Johnson (1973) Asteroid Spectral Reflectivities. Astron. Jour. 78 126

Chapman, C.R.I T.B. McCord and C. Pieters (1973) Minor planets and related objects, X. Spectrophotometric study of the composition of (1685) Toro. Astron. Jour. 78 502 Chapman, C.R. and J.W. Salisbury (1973) Comparisons of meteorite and asteroid spectral reflectivities. Icarus 12 507-522 Ganapathy, R.; R.R. Keayst J.G. Laul and E. Anders (1970) Trace elements in Apollo 11 lunar rocks: Implications for meteorite influx and origin of the moon. Geochim. Cosmochem. Acta Sunpl 1 1117 Hunt, G.R. and J.W. Salisbury (1973) Visible and near-infrared spectra of minerals and rockss VIII Meteorites. In preparation Johnson, T.V. and R.P. Fanale (1973) Optical properties of carbonaceous chondrites and their relationships to asteroids. J. Geophy. Res. In press Kiang, T. (1971) The distribution of asteroids in the direction perpendicular to the ecliptic plane. NASA SP-267, p. 187 Lebofsky, L.A. (1973) MIT Phd thesis in preparation. Lowrey, Barbara E. (1971) Optical evolution of Lost City meteorite. J. Georhv. Res. 76 4084

Marsden, B.G. (1971) Evolution of comets into asteroids? NASA SP-267, p. 413 McCord, T.B.; J.B. Adams and T.V. Johnson (1970) Asteroid Vesta, Spectral reflectivity and compositional implications. Science 168 1445

McCrasky, R.E. A. Posen; G. Schwartz and C.-V. Shao (1971) Lost City meteorite - Its recovery and a comparison with other fireballs. J. GeoDhv. Res. 76 4090 32

1-15

NASA SP-267 (1971) Physical studies of the minor planets. (T. Gehrels, ed.) Sekanina, Zdenek (1971) A core-mantle model for cometary nuclei and asteroids of possible cometary origin. NASA SP-267, p. 423 Van Houten, C.J. (1971) Descriptive survey of families, Tro- jans and jetstreams. NASA SP-267, p. 173 Williams, James G. (1971) Proper elements, families and belt boundaries. NASA SP-267, p. 177-

(1973) Meteorites from the Asteroid Belt? EOS (Trans. Am. Geophy. Union) 54, p. 233 i_)__i ___~_LI~_(I__YI1__II_ ~_lr___h__Pnns~___LLIY_

II) Previous Spectral Reflectivity Work a) Meteorites and natural silicates - Laboratory measurements

The first useful measurements of the reflectivity characteristics of natural materials were made after the advent of photoelectric measurement techniques. These measurements are divided into three distinct classess color indices, UBV measurements, and high resolution (multifilter) measurements, which correlate with three different astronomical observational techniques. The color index is defined as some function of the difference in reflectivity of a surface as seen in two broad regions of the visible spectrum (through two different filters). This is essentially a one-dimensional classification system, providing a single value for a given surface. A mass of literature, mostly in Soviet journals, has been compiled on the color indices of the meteorites and natural rock materials (Budnikovo, 19531 Sharonov, 1954,

1958, 1961: and Sytinskaya, 1955, 1960, 1965). Some of these results are shown on figure II-1, where the curves are drawn through points spaced at units of 0.1 in the color index. The dashed line is the distribution of the color indices for about 75 asteroids. (Chapman (1972) has pointed out that these color indices are based on erroneous data.) This one- dimensional classification system has some limited value in distinguishing between materials of different lithologies but is of almost no value in determining the actual surface petrol- Fiqure II-i Color Index Distributions of Rock Types ($vtinskaya, 1965)

t\t I I I * Il I II It

I -Bic goks I - I - g I -01 fIoo *!1 *#6I 0 0 i I - I S "OL Us OZ 0ust,8f ne"0i .i -orr 0r ece e0s.0 "OlO0"Or "0 Lianestone " :ust of iuetOrll s -0Z q4 *1 *06 0o *ad " oIi . Granitre -0si 00roks1

I t

*I ,

* i /0 \ . \ 00

2 00 '02 '04 "6 '1 '/0 '12 'i6 -0 Sandstone Meteorites

/ / I I r\ r! I I

I I \ I/ -02 00 0 ,'6 '01 /0 -02 00'0 '0 -02 !! *Of F*#4*I/6 *!# */O *Ol U10 *01 *d Metamorphic rocks Moon / I -02l O0 2 '06' 6*0 6 t0O Volcanic slag

(Dashed line indicates distribution of color indices of asteroids.) 35 1-2

ogy. The other problem inherent in such an approach, is

that such heterogeneous groups as sandstones (or metamorphic rocks or meteorites) cannot be considered as discrete groups.

While the members of each group share important characteristics, the differences between members of each group are, in retrospect, very significant. To consider all members of such groups together is a bit like adding apples and oranges and parrots. The use of a two-dimensional classification system,

specifically color indices from the UBV system, has been described by Hapke (1971) who showed the locations of a

number of materials in U-B, B-V space: terrestrial rocks -

figure II-2a, lunar rocks - figure II-2b and meteorites - figure II-2c. While these color difference diagrams are

less ambiguous than the one-dimensional diagrams shown on figure II-1, it is clear that no specific identification can be made from the UBV measurements. For example the position on the diagram U-B=0.2, B-V=O.1 could correspond to norite, any one of several different basalts or an anorthosite among the basic terrestrial rocks or with several types of ordinary chondrites, an enstatite chondrite and an enstatite achondrite among the meteorites. Thus, the UBV system cannot be used with any assurance to identify surface petrology.

High resolution reflection spectra over the visible and near infrared have been measured by a number of individuals, for a wide variety of meteorites and rock types. Adams (1968) II-2a

Figure T1-2 U-B, B-V Distributions of meteorites, lunpr and terrestrifl xocks. (Outlined regions refer to asteroid UBYV groupsl (Eapke, 1971)

a) Terrestrial Rccks .7 PERIDOTITE .6

.5- .9-

ECLOGITE 0 /-OIABASE OLIVINE BASALT V ,

ANORTHOSIT / DISCO ISLAND .2 *

I'I NORITE -J 0 .1 .2 3 .4 .5 ,6

Fe304*l. ASALT SCORIA 8-V Fe THOLEIITIC BASALT 0 *ISGAH BASALT .- FeTiO3 -. 1 1 I I I ,I -. 1 0 . I .2 .3 .4 .5 .6 B-V

cl Meteorites

0 -J-j 0 J .2 .3 .4 .5 .6 . B-V 37

II-3 showed a very significant range of reflectivity variations for silicate minerals, rocks and several meteorites in the near infrared (figure II-3). This opens up a possible method for remotely interpreting lithologic types. Hunt and Salisbury (1970) measured the reflection spectra of a number of silicate minerals (figure 11-4) but the quality of their spectra is questionable because of indications of poor control over the sample quality. The water related absorptions at 1.4 and

1.9 microns in their spectra of anhydrous minerals, such as feldspar, olivine, danburite and quartz, suggest that these are not fresh materials but rather significantly altered or weathered. Hunt and Salisbury (1973) have measured a number of meteorite spectra, several of which show strong indications of rust and hence alteration, although their final published work may exclude or segregate such spectra. Their work has also been limited in the number and types of meteorites measured. Johnson and Fanale (1973) have measured and studied the spectral reflectivity of a number of carbonaceous chondrites. Their detailed work shows no evidence of problems attributable to specimen alteration.

It should be clear that much of the previous work is either useless for interpretive purposes (one- or two-dimensional color indices) or is of limited value due either to poor quality control or to lack of sufficient sample variety.

Earlier work has also been made less convincing to the astron- II-3a 38 Figure II-3

Spectral Reflectivity Curves of Several Terrestrial Rocks, Minerals and Meteorites (Adams, 1968)

L5., Wew ' A Wa006Wq%(00 Reflectance spectra of the common iron-bearing rock-forming silicates; specimens were powders finer than 37 p. The olivines (A) have a single minimum at 1.02 to 1.05 p due to Fe - - in sixfold coordination. The orthopyroxenes (B) have a minimum at 0.90 p due to Fe - in sixfold coordination and one at 1.8 to 1.9 ; that is probably caused by Fe2- in a highly distorted octahedral site . The clinopyroxenes (C) and hornblende and biotite (not shown) have multiple minima at the shorter wavelengths, except that diopside and pigeonite have bands at 1.03 and 1.0 /, respectively. (D) Reflectance spectra of representative chondritic meteorites and basalts, showing how iron-bearing silicate minerals determine the positions of absorption bands in rocks. Orthopyroxene-rich chondrites (Holbrook. Richardton, Ladder Creek) each have a minimum at or near 0.90 p and another at 19 p. Basalts bearing olivine or clinopyroxene, or both (Little Lake, Kilauea Iki, Boulder County), each have a single minimum at or near 1.0 p. Specimens were powders finer than 37 M. II-3b

Figure II-4 Spectral Reflectivity Curves of Several Terrestrial Minerals (Hunt and Salisbury, 1970)

OLIVINE, VAR. FORSTERITE 318BCRESTMORE,CAL I F. OLIVINE VAR.FAYALITE IS JACKSON CO., N.C.

0 oo I 0 10(-- o1I00(-100- 80- 0OSFL -j ~ 0-74;&- ~774-250;& z - 50- 74-250# 0-74p

.50-1200 -"74-250, I.- 250-1200.I O 500 250-1200$ 0 0.4 05 06 07 04 05 06 0.7 i 1.0 1.5 20 2.5 WAVELENGTH IN MICRONS WAVELENGTH WAVELENGTH WAVELENGTH IN MICRONS IN MICRONS IN MICRONS

PYROXENE.VAR.HEDENBERGITE 10B SILVER STARMONT. PYROXENE VAR.BRONZITE 98 JACKSON CO., N. CAROLINA 0

> F

2 5 c 0-12OOj. 9- 74-250pz P 50- U 747 250j' hi 0-514L 6j 2-50-200$ U 0 z 04 0.5 06 0.7 1.0 15 20 2.5 20 25 9L WAVELENGTH AE1.0 1.5 WAVELENGTH IN MI!CRONS WAVELENGTH 0. WAVELENGTH IN MICRONS IN MICRONS IN MICRONS

ORTHOCLASE 138 RUGGLES MINE,N. H. PLAGIOCLASE,VAR. BYTOWNITE 1068 CRYSTAL BAY,MINN. 0 0 S100 100' O-4-744 7! 2100- I -.. _ 0-74; 0-4,, J , ...

0-51L 74- 25 ,u 74-250 i

50 -250-1200A S • . -i20 0 X " 250-1200.,

0 oj .0 e50120P 2 5 F50-1200;L.a.s 74-25050 o V I.s1.5 04 0.5 06 O7 .O 20 25 WAVELENGTH IN MICRONS WVELENGTH WAVELENGTH IN MICRONS "- ',VELENGTH IN MICRONS I'1 MICRONS DANBURITE I81B NEW YORK QUARTZ 328 0 0O 0

1 2!-,oo;-

o ,---- - 5 -10& 4 -4

-j5 74-250~ 1.0 74-250 , j 250-1200.

hi 2 5 0 - 120 0 . 0 S0.4 0 5 Oi 0.7 04 05 06 07 1.0 1.5 2.0 2.5 I-. 1.0 1.5 2- 2.5 WAVELENGTH IN MICRONS WAVELENGTH WAZVELENGTH IN MICRONS WAVELENGTH IN MICRONS IN MICRONS 40

II-4

omical community because of the apparent lack of (or failure

to communicate) an understanding,on the part of many of the workers, of the physical processes involved in the imposition

of petrologic or mineralogic information onto the reflection. spectra. One of the goals of this thesis is to establish a catalog of spectral reflectivity curves of reference materials both in sufficient variety and under high quality control and with a well documented understanding of the processes of petrologic signal imposition. b) Asteroids - Observational work An excellent review of spectral studies of asteroids can be found in Chapman et al (1971) and Chapman (1972). The reader is referred to these for more detailed information on this subject.

The first semi-quantitative work on the spectral reflectivity of asteroids was done by Bobrovnikoff (1929) who compared the photographic spectra of asteroids with those for G-type stars (similar to the sun). By use of microphotometric tracing across these spectra, he hoped to be able to remove the solar continuum and to determine the reflectivity of the asteroids themselves. His conclusions were a) no emission features were observed, b) no major absorption features in the visible portion of the spectrum were observed for Ceres or Vesta, c) the reflectivity of most asteroids was lower in the blue than in the red, d) there were observed differences 41

II-5

in the spectra, and e) several asteroids showed color differences with rotation. These conclusions have withstood the

test of time quite well. Recht (1934), Watson (1938, 1940) and Fischer (1941)

all utilized photographic photometry (relative brightness of an object on photographic plates with different spectral re-

sponse or through different filters) to obtain color indices for a number of asteroids. Recht, in a study of 34 asteroids, reached the erroneous conclusion that most asteroids are better reflectors in the blue than in the red. The large scatter and the evidence of systematic errors in Recht's data handling, have

led to strong criticism of his work. Although Watson measured color indices for seven asteroids and Fischer for 33 asteroids, the scatter in their data is still large and there is generally poor agreement with later photoelectric colors. More recently Sandakovo (1955, 1959, 1962) has utilized the photographic photometry approach for 56 asteroids, but his correlation with later photoelectric work is not good.

The advent of photoelectric photometry in the 1950's offered the first real quantitative measurements of the reflectivities of asteroids. Kitamura (1959) measured 42 asteroids in two colors at slightly longer wavelengths than the standard B and V filters. Since the 1950's Gehrels, Kuiper and their associates have published much work on the photoelectric photometry of asteroids in the standard UBV filters. II-6

These results have been summarized by Gehrels (1970). The consistency of these data is high and the estimated errors are quite small. These results show that most, if not all, asteroids reflect more strongly in the red than in the blue and there are distinct clusters of color groups of asteroids when plotted in B-V, U-B space (figure 11-5). However these filters (broad and situated only in the visible range of the spectrum) offer insufficient information (resolution and coverage) to be very useful in lithologic interpretation. This does not-imply that these UBV measurements are worthless; indeed they are quite useful in noting differences in asteroid spectra to be investigated by more appropriate interpretive techniques. Haupt in 1958 (Gehrels, 1970) measured 4 Vesta at six wavelengths and showed the presence of an absorption feature near one micron, but the significance of this feature was ignored. Gehrels et al (1970) measured three asteroids in the UBVRI filters plus several other filters in the red and the near infrared. McCord et al (1970) made the first detailed spectral reflectivity study of an asteroid, 4 Vesta, through 22 narrow bandpass filters ranging from 0.3 to 1.1 microns. Their work showed a deep absorption feature centered near

0.95 microns. Subsequent work has shown that this feature is much weaker or absent in other asteroids. Chapman (1972) and

Chapman et al (1973) have reported the measurement of asteroids 43 II-6a

Figure II-5

UBV Color Index Plot of the Asteroids

B-V

-Differences between UBV indexes of asteroids and the Moon relative to the Sun. Also see figure 1 of the paper by Chapman, Johnson, and McCord. 1

CGehrels, 1970) * 44 II-7 through these filters and have shown that a very significant range of curve shapes can be found for these objects (see figure 11-6), which implies a significant range of lithologic types on the surfaces of asteroids. c) Mineraloic and petroloic interpretation of asteroid spectra The usefulness of spectral reflectivity measurements of an asteroid in the interpretation of the surface lithology of the object is directly proportional to the wavelength range covered and the resolution of each point measured in that range. The greater the range and the higher the resolution, the more accurately the comparison can be made. (Obviously one eventually approaches a point of diminishing returns but this is not a current problem.) Photographic photometry is definitely a marginal technique to obtain even approximate color indices. These color indices have minimal wavelength coverage and poor resolution so that this technique is useful only in establishing gross differences between asteroid spectra, and is useless as an interpretive technique. The UBV measurements of meteorite and mineral reflectivities discussed in part 'a', are quite ambiguous. The multiplicity of lithologies with the same UBV color indices prevents any real petrological interpretations from these data. However, these data are useful for indicating differences in asteroid spectra to be investigated by more detailed techniques. Thus work prior to 1970 is of little value. II-7a 45 Figure II-6

Spectral Reflectivity Curves of Several Asteroids

is• I i I I I I" 0 563 SUL.CEA S , r r ii . 22,AMOTE . (Chapman et.al, .1973:

03 0.5 07 09 l.1 WWELENGTH "m' wZLEwwww c 46

II-8

in the interpretation of surface mineralogy and petrology of the asteroids. Recent high resolution spectra of both asteroids from telescopic observations and comparison rocks and minerals in the laboratory form the basis for identifying the surface materials of the asteroidal bodies. Available asteroid spectra are increasing with respect to both the number of objects observed and the number of observations on each object so as to improve the statistical validity of the data. The number and quality of the reference spectra, especially those for the meteorites, is still quite low. Only very recently (at this writing several papers are in press but none are yet in the literature) have any reasonable comparison spectra become available. These are subject to the limitations described in part 'a' of this chapter. Chapman and Salisbury (1973) have attempted correlations between the asteroidal spectra of Chap- man and McCord and the meteorite spectra of Salisbury but the results have been hindered by the inadequacy of the latter data set. Johnson and Fanale (1973) have approached the same problem and have within the limits of their work achieved significant results.

References

Adams, J.B. (1968) Lunar and martian surface: Petrologic significance of absorption bands in the near infrared. Science 15 145 3 47

II-9

Bobrovnikoff, N.T. (1929) The spectra of minor planets. Lick Obs. Bull. 18 18

Budnikova, N.A. (1953) Vestn. Leninvr. Gos. Un-ta 8 8, 84 Chapman, C.R. (1972) Surface properties of asteroids. MIT Phd thesis

, T.V. Johnson and T.B. McCord (1971) A review of spectrophotometric studies of asteroids. NASA SP-267, p. 51.

, T.B. McCord and T.V. Johnson (1973) Asteroid spectral reflectivities. Astron. Jour. 78 126

and J.W. Salisbury (1973) Comparisons of meteorite and asteroid spectral reflectivities. Icarus 19 507-522 Fischer, V.H. (1941) Farbmessungen an kleinen planeten. Astron. Nachr. 272 127 Gehrels, T. (1970) Photometry of asteroids. In Surfaces and interiors of Dlanets and satellites (A. Dollfus, ed.) Academic Press, London.

, E. Roemer, R.C. Taylor and B.H. Zellner (1970) Minor planets and related objects. IV Asteroid (1566) Icarus. Astron. Jour. 75 186 Hapke, Bruce (1971) Inferences from optical properties concerning the surface texture and composition of asteroids. NASA SP-267, p. 67 Hunt, G.R. and J.W. Salisbury (1970) Visible and near infrared spectra of minerals and rocks: I Silicate minerals. Modern Geolovy 1 283 (1973) Visible and near infrared spectra of minerals and rocks, VIII Meteorites. In preparation.

Johnson, T.V. and F.P. Fanale (1973) Optical properties of carbonaceous chondrites and their relationship to asteroids. J. GeoDhv. Res. In Press. Kitamura, M. (1959) A photoelectric study of colors of asteroids and meteorites. Pub. Astron. Soc. Janan 11 79 McCord, T.B.: J.B. Adams and T.V. Johnson (1970) Asteroid Vesta: Spectral reflectivity and compositional implications. Science 168 1445 48

II-10

Recht, A.W. (1934) Magnitude and color indices of asteroids. Astron. Jour. 41 25

Sandakovo, E.V. (1955) Concerning color indices of minor planes. Astron. Circ. Acad. Sci. USSR no. 163

(1959) Interpretation of color indices of minor planets. Publ. Kiev Univ. Astron. Obs. 8 (1962) Concerning the color indices of minor planets. Publ. Kiev Univ. Astron. Obs. 10 Sharonovo, V.V. (1954) Astron. Zh. 31 442

(1958) Uch. Zap. Leninr. Gos Un-ta 273 120

(1961) Izv. Komissii to Fizike Planet 3 74

Sytinskaya, N.N. (1955) Meteoritika 13 65

(1960) Izv. Komissii Do Fizike Planet 2 59

(1965) Experiment in colorimetric comparison of asteroids and terrestrial rocks. Soviet Physics - Astronomy 2 100 Watson, F.G. (1938) Small bodies and the origin of the solar system. Phd thesis, Harvard Univ., Cambridge Mass.

(1940) Color and magnitude of asteroids. Harvard Collesre Obs. Bull. 913 3 49

III) Snectral Reflectivity of Natural Silicate Materials: Internal Effects a) Introduction

The conditions which give rise to or modify the manner in which photons of any particular wavelength are reflected from a solid surface can be divided into two distinct groups. The first group, treated in this chapter, can be termed internal variables: that is, parameters which are primarily a function of crystal fields effects. The second group, treated in the next chapter, is characterized by external factors and include particle size distribution, packing and other physical parameters of the material. The characteristics of a spectral reflectivity curve (slope, presence and position of absorption features) may be considered to be controlled almost exclusively by the internal parameters. External parameters can modify the curve by decreasing the contrast (amplitude of the variation) or slightly changing the slope, but they cannot produce the curve independently. b) Crystal Field Theory A knowledge of crystal field theory (CFT) is necessary to understand the processes which produce the characteristic spectral reflectivity features that occur when light interacts with a silicate surface. No attempt will be made to present an exhaustive review of crystal field theory. Instead., the reader is referred to an excellent text on the mineralogical II111-2

applications of CFT by Burns (1970a). Only a brief review of the subject is presented here so that a reader unfamiliar with the field will be able to follow the logic which leads to some of the subsequent assumptions and conclusions. Furthermore, no discussion will be made of Ligand Field Theory of Molecular Orbital Theory (more complex and exact versions or CFT) since the additional precision gained makes no significant change in the validity of the conclusions reached. For a discussion of these topics the reader is referred to Figgis (1967), Ballhausen (1962), Orgel (1970) or Dunn et. al. (1965). Crystal Field Theory can be considered as a combination of structural mineralogy, quantum theory and wave mechanics and is specifically concerned with physical systems. In this paper, we will deal only with CFT as it applies to the transition metals of the first transition series (Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Cu), which are characterized by partly filled 3d-orbitals in any of their commonly occurring oxidation states. Since iron (Fe2+) is the most common of these metals found in the unoxidized mineral assemblages typical of meteorites, we will be primarily concerned with it. A detailed discussion of the characteristics of the transition elements can be found in Cotton and Wilkinson (1966). The 3d-orbitals are mapped out by solving the Schrodinger wave equation for quantum numbers n=3 and 1=2, to produce five orbitals (quantum numbers mi = +2, +1, 0) which have identical 51 III-3

radial dependence but differ considerably in their angular distribution (see Figure III-1). In the absence of an electro- static field or in the presence of a spherically symmetric field, these five orbitals have identical energies. In a coordination site (cation surrounded by a regular array of anions) certain of the orbitals of the transition metal cation more closely approach the surrounding anions. This increases the interelectronic repulsion and splits these orbitals to higher energies. The d-orbitals in an octahedral environment (six fold coordination) are shown in figure III-2 and the energy level dia-

grams for several environments are shown in figure I1-3. Ferrous iron (Fe2 +) has a d6 electronic configuration. Under normal conditions (high spin) five of these electrons with parallel spins occupy the five d-orbitals while the sixth

electron is antispin and occupies the lowest energy (ground state) orbital. In anunsplit configuration this antispin electron is not preferentially located in any orbital. In any other environment, there is splitting of the d-orbitals and the sixth electron is located in the lowest energy level in the groundstate. This sixth electron can interact with

(absorb) a photon of the appropriate energy (corresponding to the energy differences of the split orbitals above the ground state) and undergo a transition to a higher energy orbital. The decay of the excited state will produce a photon of very nearly the same energy as the exciting photon (Frank-

Condon Principle) but in a random direction. 52 III-3a

Figure III-I

's' orbital

'p,' orbital 'pY' orbital 'p,' orbital

"4 x y .4. J :, - y 49 x• x

Boundary surfaces of atomic orbitals. The boundaries represent angular distribution probabilities for electrons in each orbital. The sign of each wave function is shown. The d orbitals have been classified into two groups, t, and e,, on the basis of spatial configuration with respect to the cartesian axes. (Reproduced and modified from: W. S. Fyfe, G4eodmistry of Solidr, McGraw-Hill Book Co., New York, 1964, figure 2.5, P. g9.) Figuye III-2

3d-Orbitals in an Octahedral Coordination Site.

- Log, /= I s- I -

- C -- - I - G

* I /', * 7 e I

dz2 d xz.yZ dxydxz,dyz III-3c 54

Figure III-3 (Burns,1970a)

d,7y d

I I

r- I------Ida,, dy. dd d- ----

energy NJ d t

Relative energy levels of d orbitals of a transition metal ion in octahedral co-ordination.

1 . d

dd.

I= _

J-......

square planar octahedral tetragonal monoclinic Relative energy levels of d orbitals of a transition metal ion in coordination sites of various symmetries. The z axis in each site is the axis of highest symmetry and usually corresponds to the axis of distortion of a regular octahedron. 55 III-4

In an ideal crystalline solid for a given set of ligands, the energy of the transition is mainly determined by the particular cation and the symmetry of the site in which it re- sides. However, Burns et. al. (1966) showed that the center of an absorption feature due to a single crystal field transition appears to shift slightly depending in the relationship of the polarization vector and the crystal structure in the mineral Gillespite. They attributed this shift to vibronic effects preferentially oriented due to the square-planar coordination about the transition metal cation. A discussion of the relationship between the site symmetry elements or group theory and CFT can be found in Cotton (1971). However in a real mineral crystal structure the dimension of the sites and hence the magnitude of the splitting fluctuates due to thermal lattice vibrations. Since the magnitude of the site distortion is proportional to the energy available to displace the ions against a restoring force, the number of sites with a given splitting will vary as the thermal energy of the material. This produces absorption features that are generally Gaussian in shape. Additionally, any real crystalline solid has a myriad of structural defects such as vacancies, fractures of the crystal or offset in the lattice layers, which produce a range of distortions of the site symmetry elements and which create a dispersion of nominal splitting energies. The effect of these factors is to produce broad (= 2-3000 cm "1) absorption features in the spectra of light which has interacted with a 56 III-5 transition metal ion in a silicate mineral structure.

c) Crystal field effects measured in minerals Burns (1965, 1970b), White and Keester (1966, 1967), Bell and Mao (1973) and Runciman et. al. (1973a,b) have measured the transmission spectra of oriented sections of a number of silicate minerals, including olivine. Burns (1966) has noted that the spectra of light polarized parallel to any one of the crystallographic axes of olivine differs from that of light polarized parallel to any other axis (figure 111-4), and that the features can be treated as summations of three individual gaussian absorption features (Burns, 1970b). In the olivine transmission spectra Burns (1970b) assigned the first and third O O features (Band 18550-9150 At Band IIlt 11000-12800 A) to transitions within Fe 2 + in the M(1) site (z D4h symmetry) and the second feature (Band IIs 10400-10800 2) to a transition in the M2 site (approximately C3, symmetry with the cation off- center). Burns also noted that in a solid solution series such as the Mg-Fe olivine series, the centers of these three bands migrate toward longer wavelength (lower energy) as the series becomes more iron rich. Since the Fe2+ ion is significantly larger than the Mg2+ ion, the increasing substitution of iron in the olivine structure expands the structure. This increases the dimensions of the cation sites and thus decreases the interelectronic repulsion and the splitting energy which shifts the transition toward lower energy and longer 1 WAVENUMBER (C4 ) 13000 10000 T000 13000 o000oo o000100o 1000lo 00 C WsVeatnber(cm-1 ul Il II I Jil 1t1- 25,01 20,01 15,000 , 10000

0 10

r-w

S,4Kx) 1(0,KIO 15,000 I I I£ ,I I I wavelength(A) 13000 1000 0 00 70000ooo 10000 7000o ** " Polarized alorption spectra of fayalite (specimen10) ...a spectrum; wavounseI tewl) spectrum; -- v spectrum. (optic orientation: a-b; P-e; -ya. The polarized absorption spectra of olivines resolved Into component'Gaussian- shaped bands. The figure shows each polarized spectrum of three olivines (specimens 11, 5 and 2) separated with a Dupont model310 curve resolver into three overlapping absorption band. tionb~rnds. 58 III-6 wavelength. The dimensions of the unit cell of fayalite

(Fe2SiO 4 ) are approximately 6% larger than those of forsterite

(Mg2SiO4 ) exhibiting this expansion of the crystal structure. It was also shown that this shift toward lower energy is nearly linear with the changing composition (figure III-5).

Figure III-5 (Burns, 1970b)

12000 1 23 4 5 67 8 9 101

11000 "

S10000

9000

8000 1 2 3 4 5 67 8 9 101

0 10 20 30 40 50 60 70 80 90 100 mole % Fe2SiO 4 Compositional variations of absorption maxima in the polarized spectra of MAg"-Fe olivines. El 3 spectra (84.50-9050 A); O a spectra (8550-9150 A); X y spectra (10400-10800 A); A 0 spectra (11000-12400 A); * published data for other forsteritic olivines. C Clark (1957); N Farrell and Newnham (1965); W White and Keester (1966); F Fukao d at (1968) S Shankland (1969).

Runciman et. al. (1973b) have presented an alternative interpretation of the three bands in the one micron olivine absorption feature which contrasts sharply with the interpretation by Burns (1970). Runciman and his colleagues suggest that all three bands must be the result of Fe 2 + transitions in 59 II-7

the M(2) site. Since they consider that transitions in a

centrosymmetric site (eg MWO) would be an order of magnitude less intense than in a noncentrosymmetric site (eg. M(2)),

they conclude that bands I and III cannot be due to transitions in the M(1) site as suggested by Burns. They admit that they

are unable to provide a simple explanation for bands I and III. While Burns (1973) has presented several very persuasive arguments against the Runciman et. al. interpretation, we shall

consider both of these interpretations in the light of subsequent work later in this paper and attempt to determine which interpretation more clearly matches the observed relationships.

The overall slope of the visible and near infrared spectrum

is presumed to be the result of metal-oxygen charge transfer absorptions centered in the ultraviolet region of the spectrum with wings into the infrared. For simplicity, this charge transfer effect is considered to be similar to crystal field transitions except that the electron in question is exchanged with another ion. Consequently, the energy required is generally higher and the half-width is much broader due to the irregularity of the natural crystal structure. For most silicates the high blue and ultraviolet absorption is due to this effect. Work discussed in the section on meteorite reflectivities indicates that over the range measured, the slope of the continuum is linear in energy. 60 III-8 d) Aneular variations of photon interaction Since the work used to establish the tenents of crystal field theory in relation to the observed physical characteristics of natural mineral systems has involved transmission spectra exclusively, the correspondence between these phenomena is well understood for oriented crystals. However reflection spectra are composed of two components, specular (first surface reflection at an interface between mediums with different indices of refraction) and transmitted, with the transmitted component carrying essentially all of the 'signal'. This crucial transmitted component is composed of a multitude of rays passing through randomly oriented crystals and is thus the sum of all possible transmission spectra. Any attenpt to understand the reflection spectrum of a mineral surface in terms of crystal field theory must be based on knowledge of the relationship between the interacting photon and the electron in the ground state orbital and in particular its variation with anzular rotation. As far as the direct relationship between CFT and the reflection spectrum of a mineral is concerned, the literature on the subject simply does not exist. However some work describing the angular variation of certain transmission spectra does exist. As was noted in section III-c, the shape and intensity of the one micron absorption feature of olivine varies with the orientation of the crystal relative to the direction of polarization of transmitted liaht. For an olivine specimen, Burns (1966) 6], 111-9

showed that when the plane of polarization of the transmitted light was rotated between crystallographic axes about the

third axis the shape of the absorption feature varied smoothly between the axial spectra (figure 111-6). However, no par-

ticular attempt was made to interpret this effect. It can also be noted that in the polarized transmission spectra of an olivine the X -spectrum has an absorption maximum several times more intense than that of either the oc or B -spectra. However the reflection spectra of an olivine specimen (figure III-7) shows only a relatively weak maximum

Figure III-7 (Hunt and Salisbury, 1970) Soectral Reflectivity Curves of Olivine

o M oo4 --

Mg. 4 Fe.6 / / g, Mg. . F\e.

z r FAYALITE Fel.o

1.0 1.5 2.0 2.5 WAVELENGTH IN MICRONS

at the position which corresponds to Band II. This indicates that the summation of transmission spectra producing the reflection spectra cannot be a simple linear average of the III-9a 62

Figure III-6 (Burns, 1966)

7 4 * 5 30' 6 50 8 15 995 30* 10

S ie 8 ,o 2 14 i 12 ," Waveltenqt(0-' A) r ---X

18 I5 17 300 16 45* IS 30' 14 15

Variations in the polarized absorption spectra of fayalite with changes in orientation of each vibration axis. Spectrum 1: x vibration axis horizontal east-west; z vibration axis vertical; y vibration axis horizonta! north-south. Spectra 2-6: Rotation about the y axis brings the z axis into horizontality (east-west) and the x axis vertical. Spectrum 7: z vibration axis horizontal east-west; y vibration axis vertical; x vibration axis horizontal north-south. Spectra 8-12: Rotation about the x axis brings the y axis into horizontality (east-west) and the z axis vertical. Spectrum 13: y vibration axis horizontal east-west; x vibration axis vertical; z vibration axis horizontal north-south. Spectra 14-18: Rotation about the z axis brings the x axis intd horizontality (east-west) and the y axis vertical. The angle marked on each spectrum is the tilt angle to the horizontal. 63 III-10

three axial polarized spectra. This implies that the strong maximum exhibited by Band II in the X -spectrum must be quite localized in solid angle distribution.

e) Aneular variation in nhoton interaction with a mineral - Experimental

In order to determine the exact nature of the angular variation of the shape and intensity of the absorption feature in a silicate, in this case an olivine, an extensive program of detailed measurement and examination of the transmission spectra of an oriented olivine crystal was undertaken. A thin section of an olivine crystal (Fa 9 9 ) was provided by Dr. Roger Burns for this measurement. The crystal was cut approximately perpendicular to the 'b' axis of the crystal. The spectral measurements were made on a Cary 17 spectrophoto- meter made available to this experimenter by Dr. Burns. The olivine crystal was placed in a universal stage and the crystallographic axes located. Using the universal stage, any arbitrary light path with respect to the crystal axes through the thin section could be established. With the polarizer (a calcite Nicol prism) in the light beam between the source and specimen, the universal stage could be rotated to measure the polarized spectrum at any orientation of the vibration vector of thetransmitted light with respect to the crystal structure and orthogonal to the chosen light path (rotation axis of the universal stage). The sum of the individual 64 III-10a

Figure III-8

Rotation axes and Light Paths with respect to the Crystallographic Axes.

II I II I I II II

05=3000

The five 0-directions represent the paths of transmitted light with respect to the crystallographic axes. The 5's are also the axes of rotation of the polarization vector of the transmitted light. The photon vibration vectors for a given 0 will lie in a plane orthogonal to the 0 direction spaced at 10 degree intervals. 0=90 coincides with the 'b' axis. 0=600 & 200 li in the 'bc' plane 300 either side of 'b'. 0=240 & 300 lie in the 'ba' plane. 65

polarized spectra measured during a rotation about a given axis should be equivalent to the unpolarized spectrum of light transmitted along that axis. For these measurements, five rotation axes (light paths through the crystal) were chosen and the polarized spectra measured every 100 of rotation about each of these axes. The five axis chosen are shown on figure 111-8 with respect to the crystallographic axes. Axis -d90ois the 'b' crystallographic axis, Axes =60o and b=1200 are in the 'ab' crystallographic plane 300 either side of the 'b' axis, and Axes 6=2400 and

=3000 are in the 'be' crystallographic plane 300 either side of the 'b' axis. Thus, while previous measurements had been made for polarized spectra with rotation axes coinciding with the crystallographic axes, we could make the more general measurement of off-axis and out-of-axial-plane transmission spectra. Spectra were measured for nearly 100 different crystal orientations for the wavelength range 6000-18,000 2 (16,500- 5500 cm-1). All spectra were recorded at high spectral resolution (5-20 R) by strip chart recorder and reduced by hand. A number of interspersed spectra taken through the same microscope, universal stage and glass slide setup were sub-

tracted from the mineral spectra to remove instrumental effects.

The polarized absorption spectra spaced at 100 intervals

of rotation (9) about each of the five axes are shown on

figures III-9a,b,c,d,e for =-90O, 600, 1200, 2400 and 300o III-lla

Figure III-9a

9=900 I PVV il'c'

0 r-i 0

1 . 14 c

0.50 P=00 -d PVV: Il'a' 4 o

Wfenubef X -1 I . I .I I veu1 e xl ?c0. ') -1 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 3 0 TO 90 DEG.. PHI = 90 Variation of absorption feature in olivine spectrum during rotation of photon vibration vector (PXV) about 'b' crystallographic axis from 'a' (9=0 ) to 'c' (9=90 ). III-lib

Figure III-9a

9=900 PVV= ' '

@=1800 PVV= 'a '

I I I I I I t I I i I I I I I I 1 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 3 90 T70 180 DEG., PHI = 90 III-lic

Figure III-9b

O 0=600,e=9 0 PVV='bc'-plane, 300 left of 'c'

e=00 PVV= a'

III I It 111111111i 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 3 0 TO 90 DEG.. PHI = 60 III-lid

Figure III-9b

0=60', 9=900

PVV='bc'-plane, 300 left of 'c'

9=180=00 PVV= ' a '

1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 3 90 TO 180 DEG., PHI = 60 70 Ill-lie iA 5=1200 Figure III-9c 9e=90 0 PVV= 'bc'-plane, 300 right of 'cI

PVV='a'

l 1I1111_ II,II I 1.8 1.5 1.4 1.2 1.0 0.8 0.6 0.4 0.2 3 90 To 180 DEG., PHI = 120 III-llf f =1200 Figure III-9c =90 PVV='bc'-plane, 1 i 300 right of 'c'

I !I

3 0 TO 90 DE.. PHI = 120 3 0 T( 90 OEG. Pt-f = 120 III-llg

Figure III-9d

;=2400, 9=900 PVV='ac'-plane, 300 left of 'c'

9=00

I I I I I I I I I I I [ I I I I 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 3 0 TO 90 DEG, PHI = 240 III-11h

Figure III-9d

0=2400, 8=900 PVV='ac'-plane, 300 left of 'c'

9=1800 = 0

1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 3 90 TO 180 DEG., PHI = 240 III-lli 74

Figure III-9e s6=300 0 9= 900 PVV=' ac' -plane, 300 right of 'c'

e=0 0

L L 1 , I I I , - I, I I L i I 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 3 0 TO 90 OEG., PHI = 300 III-11j 75

Figure III-9e 0=3000 8=90 0 PVV=' ac' -plane, 300 right of 'c'

e=1800=00

I I I I .I I I I I . I I I I - I I I 1,8 1.6 1.4 1.2 1.0 0.8 O.G 0.L 0.2 3 90 TO 180 DEG.. PHI = 300 76 111-12

respectively. In each figure pair, for a given axis $ (labeled either '0 to 90 Deg.' or '90 to 180 Deg.'), the ampli-

tude of the curves increase toward e = 900, The true X-spectrum (photon vibration vector parallel to c-axis) is found only

at the $=900, 9 =900 orientation, while the true ~ -spectrum is found at 6 =00 (or 1800) for )=900, 600 and 1200 . The vertical scale is loglO (absorbance) where each division is 0.1 units. The continuum has been removed from each spectrum by the sub- traction of a continuum of constant slope (in energy coordinates) which was calculated independently for each spectrum. The value of this continuum slope for all spectra averaged +0.0251/1000 cm-1 with an average variation of +0.0017 and a full range of variation of _0.0040. This continuum sE e is very nearly constant and certainly exhibits no strong directional dependence. An examination of these spectra quickly reveals that the maximum intensity reached by the spectra at 8 =900 (photon vibration vector parallel or subparallel to the c-axis), attributed to band II (Burns, 1970), is very directionally dependent. The maxima reached at 6=1200 and 3000 are both significantly stronger than that reached at --90o, while those at '=600 and 2400 are both significantly weaker. The maxima corresponding to bands I and III also show a variation but not nearly as extreme. The intensities of the maxima of the absorption feature ( =900) for each of the rotation axes (6) are given in Table III-1. This observation should dispel any II1-13 77

Table III-1 Intensity of Absorption Feature at Maximum (0 =90) as a Function of Rotation Axis (d)

0.D.* Trans. Absorbed 90' 2.4 0.4% 99.6% 6o0 1.4 4.0% 96.0% 120" 2.95 0.1% 99.9% 240 1.27 5.3% 94.7%

300' 2.7 0.2% 99.8%

* O.D. (Optical Density) = Loglo(Io/I). O.D. is proportional to the number of absorbing elements encountered by the photon flux.

lingering doubts concerning the orthogonality of the axially polarized spectra. Even with only two polarization directions completely mapped out, it is clear that the vector direction of the maximization or minimization of the transition feature is not coincident with the crystallographic axes. Thus in general, the three orthogonal axial spectra (All a, lib, I Ilc) do not constitute 'end-member' spectra, and intermediately oriented polarized spectra cannot be derived by some linear summation process from these three spectra. The intensity of Band II appears to maximize in the X -spectrum only because the parameter which controls this transition maximizes in a direction 78 III1-14

closer to the c-axis than to either of the other two crystallographic axes. The nature and orientation of this parameter will be discussed in greater detail later in this chapter. Since the individual transitions as measured by Burns

(1970b) and implicit in any type of crystal field model (see section III-b above) have Gaussian shapes in energy space, the resolution of the measured spectra into their individual

Gaussians allows one to determine the variation of the parameters with respect to the orientation of polarized light in the crystal structure. To accomplish this, a computer program based on the exact solution of the Gaussian equation was applied to these spectra. The results, shown in table 111-2, present a confusing picture. If the crystal field model and assignments by Burns (1970b) are used to explain the observed transitions, certain conditions should be observed in these results: a) the centers of the three transitions should remain at nearly fixed positions (subject only to relatively small shifts as a result of vibronic effects, Burns et. al. (1966)) and b) the halfwidths of the bands will be nearly constant irrespective of rotation. The model defined by Runciman et. al. (1973b) implies Gaussian distributions (line shapes) for the transitions but is sufficiently vague, especially with respect to the spectral regions designated Bands I and III, as to preclude the establishment of testable criteria.

An examination of Table I-2 shows that the deviations of the values for the center position and halfwidth for band 79 III-14a

Table 111-2

Band I- Bqnd III Cen. H-Width Cen. H-Width 900 00 10490cm - 1 1630cm -1 7470cm-1 900cm- 1 if 10 10460 1660 7500 90o '9 20 10390 1700 7490 890 if 160 10480 1680 7450 890 ifg rN 170 10400 1670 7460 900 ( 180 10440 1660 7480 60 0 10460 1620 7530 930 if 10 10590 1590 7560 940 if 20 10570 1610 7540 930 if 160 10780 1540 7400 850 if 170 10610 1600 7490 91o if 180 10510 1610 7520 920 120 0 10770 1410 7490 890 " 10 10660 1520 7410 860 20 10580 1580 7510 890 160 10570 1620 7460 870 ft 170 10500 1640 7510 910 *U 180 10740 1510 7510 900 * 240* 0 11060 1390 7460 890

if* 10 10670 1560 7450 870 20 10730 1530 7470 870 if 870 if 160 10730 1490 7400 170 10700 1500 7420 880 180 10610 1540 7420 890

300 0 10680 1520 7470 880 10 10740 1510 7420 840 20 10680 1550 7440 850 160 10650 1540 7460 880 170 10730 1530 7440 870 180 10780 1490 7410 850

Averages 10625 1520 7470 890

Deviations (+) 115 70 35 20 III-14b 8Q Table 1TI-2 Cont.

Band II

Cen. H-Width 900 700 9270cm -1 1770cm--1 80 9150 1240 90 9230 760 100 9210 1050 110 9270 1550 60 70 9280 1670 80 9230 1290 90 9110 1400 100 9120 1450 110 9140 1610

120 70 9330 1810 80 9210 1130 90 100 110 9260 1360 240 70 9200 1670 80 9310 1260 90 9250 100 9290 1260 110 9190 1460 300 70 9230 1670 80 9200 1170 90 100 9170 1220 110 9280 1720

Averages 9220 1400

Deviations (+) 210 II1-15

I are much larger, In the fitting program, the errors are correspondingly very much larger. The variations for band

II are also very large and the fit is poor. The pattern which emerges suggests that more than three transitions are contributing in a significant way to the olivine absorption feature, suggesting that perhaps a fourth transition at an energy intermediate between those of bands I and II, contributes to the spectra. To evaluate this possibility, the spectra were resolved on a Dupont model 410 curve resolver provided by Dr. Burns. The results obtained were indeterminate since, over the rotation positions, bands with fixed centers could not be defined precisely.

An alternative approach was undertaken to determine the nature and magnitude of the angular variation of the spectra. This involved taking the difference between adjacent (in terms of rotation about a given axis) spectra and noting the location, direction and magnitude of these changes. The results are displayed on figure III-10a,b,c,d,e for the five different axes. Since the difference between two Gaussians with the same center and halfwidth but different amplitudes is also a Gaussian shape, one would expect to see three regions of Gaussian shape (or sums thereof) varying. The typical noise level on these differential spectra is approximately 0.1 vertical units (as plotted) except between 8000 and 11000 cm- 1 near e = 900, where the noise level approaches 1.0 vertical units. Since the effective edges of the absorption feature in our III-15a 82 Figure III-10a

Variation of Olivi. Absorption Feature between adjacent T insmission Spectra (8's)

90 - 90 -100 2

80 -70 100- 110

70 - 60\ - 110- 120

SO-50- 120- 130

50 - 40 130- 140

140- ISO

150- 160

160- 170 I14 ! - 1i I

10 - 0 170-180 1 I 1 -- I I I i I I 1

1.8 1.6 1.4 1.2 1.0 0.6 0.6 0.4 0.2 1.p 1.6_ 4 1.2 1.0 0.8 0.6 0.4 0.2 Wavenumber (xl0O cm ) PHI = 90 83 III-15b

Figure III-10b

90 - 80 90 -100 I I

80 - 70 100-110

70 -60 110- 120

120-130 50 - Lo

130- 140

40 - 30 1t- 150 I II{- 1 1

30 - 20 150- 160

20 - 10 160- 170

0 10 - 170- 160 1 11. 1--t- 1 IA A PI I I -

1.8 1.6 1.4 1.2 1.0 0.6 0.6 0.4 0.2 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 PHI = 60 OZI !* IHd

E'D hO 9.0 910 0*1 z*1 hW 91 911 gI :)- Z-1 tI 9-,4L1gI iI t I 1 I 09 OLI

I I I I I I +41 ~---.l-

I t IvJy.4±- I I

0?4- 091 -Ohl

OFSI -02i1 Oh a s

Os - 09

OL -09

LI 09- 06

ZoQT-III aanbTa

OST-iii III-15d 85

Figure III-10d

90 -100

80 -'70 100-110

70 -60 110- 120

120- 130

50 - 50 130- 10

40 - 30 140- 150

30 - 20 150- 160

20 - 10 160- 170

11 -0 170- 180

1.8 1.6 I.4 1.2 1.0 0.8 0.6 0.4 0.2 1.6 1.6 1.4 1.2 1.0 0.8 0.6 O.4 0.2 PHI = 240 86 III-15e

Figure III-10e

90 -B0 L 1\LIII i I I I 90 - 10 -

80 - 70 100- 110i

70 -60 110- 120 /

120- 130 /

1110- 150

I l~-4' i-h4-J 1 1 1

I20 Ir4I-- 10 i T 1tI I 160- 170 ,.l

10 -0 170- 180 -

1 1 1 1 i1 1 1 1f 1f 1If

1.6 1.6 1.4 1.2 1.0 0.6 0.6 0.4 0.2 1.5 1.6 1.4 1.2 1.0 0.0 0.6 0.4 0.2 PHI 300 87 III-16

olivin6 specimen are at 14,000 and 6000 em-1 , this interval is of prime concern and for convenience it is subdivided into two distinct regions, (a) 11,000 to 14,000 cm-l the variations in this region are not as intense as those over the rest of the feature but they display a very distinctive pattern. This is in the region where on the measured spectra the intensity decreases as 8 approaches 900, as shown by the inward movement of the high energy edge of the feature on figure 111-9. At low 6 's (00o-200, 1800-1600) no significant trend can be defined, however at intermediate 9's (300-700, 1500-1100) significant features, predominantly negative, appear. Especially for = 600 and 1200, these variations are strongly indicative of several features all of approximately the same halfwidth (500 cm'l) interacting to produce the pattern shown. It is difficult to define the individual features but a reasonable estimate would suggest five or six features spaced at about 500 cm'l intervals outward from 11,500 cm'1 . To obtain these transitions from the two sites of the olivine structure would require a significant redefinition of the parameters of the sites. A more viable hypothesis is that the features are related to vibronic coupling effects. (b) 6000 to 11,000 cm-l This is the spectral region which contains the three bands described by Burns (1970b). The location of these features are indicated by the notations at the bottoms of the figures. The general fluctuation in the shape of the variations, and especially in the shape and width of the region of maximum III-17 variation (8000 - 10,000 cm l), is strongly indicative of more than three simple features overlapping (if the curves were produced by three transitions at the indicated energies, the basic shape of the variation curve would vary in a predictable patterns broad when the rates of change are nearly equal and narrow when band II is changing rapidly). The broadness or narrowness of the maximum change region seems to be independent of the amplitudes of the side transitions (I & III). One is thus forced to the conclusion that several extra transitions are involved and that they appear as this anomolous behavior in the shape of the variation curves. These additional transitions could be higher amplitude vibronic coupling features. This explanation is consistent with the observation that such features increase in amplitude (in the region where they can be clearly distinguished) toward the center of the absorption feature. The above results need not be viewed as a contradiction of the assignments of the transitions by Burns, but it does suggest that the discrepancies originate from an oversim- plification of the crystal site symmetry. If the asymmetry of the sites is considered, one can reasonably conclude that the degeneracy of the excited states in the M2 site will be removed leading to two possible transitions. This new splitting would be small and have the effect of creating two offset features near the locations previously assigned to each band. The energy level diagrams for this situation are shown on figure III-17a 89 Figure III-11

Energy Level Diagrams for the Olivine Absorption Feature

blg f ------lr -'---'ft alg- t C

b2g . e ff - t t e .. . t t 2g M. -A ii " a,a1 % (c')

- 5Eg .....SA1g5

_ST2g ..... 5B2g 5 M-O Eg 5 - A1 (d) (e) (f) """""""(f I)

Oh D3h C3v

Energy level diagrams for electronic states and 3d orbitals of Fe + in octahedral, tetragonal and trigonal coordination sites. (a). (b), (c) electronic configurations of the grouhd-states in sites with symmetry Oh, I)b (elongated along the tetrad axis) and Cj, (compressed along the triad axis). (d), (e), (f) crystal field states corresponding to the ground-states shown in (a), (b), and (c), respectively.

Figures c' & f' represent the energy level diagrams for the M(2) site (fig. c &f) with the degeneracy of the highest energy level removed to allow two closely spaced stransition to account for the observed characteristics of the Band II transition as discussed in the text. III-18

III-ll. The interpretation presented by Runciman et. al. (1973b) is not inconsistent with these observations if only because they are unable to explain the spectral features in most of the region of absorption around one micron. Alternatively, a situation may have arisen where the crystal field model simply lacks the sophistication to deal with the true situation at the precision required. If this is the case, a comprehensive description of the processes would involve a much more complex molecular orbital model. However, until the CFT model demonstrates itself to be hopelessly in- adequate, we shall avoid this latter course of action.

f) Photon interactions Implications and Discussion A method of extracting ourselves to a degree from the morass of models is to examine the rotational variations of the polarization vector of the light of small (narrow) spectral regions to see if any coherent pattern emerges which can be utilized. The results are shown on figures III-12, for each of the rotation axes. A very suggestive pattern emerges for the 9000 - 9500 cm-1 (Band II) region. The absorption in this region varies sharply, with a prominent lobe directed toward

) = 1000 (the amplitude of the lobe depends on the axis). There also appears to be a weak secondary maximum located between G = Oo and 300, giving rise to an inflection in the curve near 0 = 300, which is present in the amplitude vs. angle plots for this region for all give axes. The entire Figure III-12a Variation of Olivine Absorption Feature with Rotation Spectral Region: 13,000-14,000 cm-1

Il- 10 'S14.1 IIIINI t 3I0tIJ- PHI t 0 'tL11IR RHNI;L AC1TUAL tLHO tGhttHnMPLITUOE - O.116 ACtllaL fRO Of fiFf AMPtITUDO,, 0.04 SCLLI U A FHLIO l 5bU0.000 SCALED BY A FACTOROF 50.000

PHJ - 90 bPLCIHAL HRiNGE ACTUAL ZERO OEGREE AMPLITUOE - 0.05 SCALEDBY A FACTOROF 50.000

1 PHi) 60 SPCTHRAL RANGE 1 13906- 13072 CI- PHI - 240 SPEL H1L RANGE a ACIUAL LEROOEGREE AMPLITUDE * 0.05 ACTUAL ZERO OGRHEE AMPLITUDE = 0.06 SCALEO BY A FACTOR OF 50.000 SCRLEO BT A FACTOR OF 50.000 Figure III-12b m. Variation of Olivine Absorption Feature with Rotation Spectral Region: 11,000-11,500 cm-1

-- -'-- -~ N ,--~~~--~ -

N *~~~' "" N

111129- 11(150CK PHI1 300 SPFCTHRL HRHNhfI 1.4 11tJO (C PtN = 120 SPE'ITR( RANC t HCTUALtEHO O(FGHL 14MPLITUOE, 0.46 ACIUAL ZLRO OEGRIE RMPLITJUE- 0.43 SCALFO BY A FACTOR OF 6.000 SCA EO BY A FACTOROF 6.000

Pil . 911 ;Pf TH 1 H1N(; t 11 l. !1 1 5(1y r I RLI1HI /tOI 1110(, I t f llt 1111 SyRtlfOHT H llf'IIR of h.ll1ll

11429 - i10'0CKI PHI = 60 SPFI'TRRLRU PHI = 240 SPECTRAL RRNGE 1 ACIUAt ?FRO OFCRFf AMPLITUOF= 0.49 RETUAL7ERO OEGRFE RAPLI 'JOE 0.47 'SCARED8T A FACTOR Of 6.000 SCAIEO BY A FRCTOROF 6.000 Figure III-12c .Variation of Olivine Absorption Feature with Rotation Spectral Region: 10,500-11,000 cm-1 Spectral Region: 10,500-11,000 cm

PH1 '10 1~'(l.ritlHHNGIF 109?9- 0l)ti C" PHI - I120 I'W L Mll10 1Nl,.I 10929 -I 15112 CII H1:Illll01II Il;Hf PL IiT E • 0. 47 RCIIL LHl OfIGtII HNI'LIIJOt0 0.4 1IIO BY R FACI11tOF 5.000 SCItO BA110 OtFlllll. 110l

PIll ,' !Ill it'41 IHIII HIIN(I II4.l1 1 I'I' WIlf ofHI M I f1iillI1 th 1 I lOi I illi - 0.1mi WH I) iUt N l I(n li t 5 1iii

N. \

/ N

PHI 60 SPECIFRI HRN.L 10929 -10582 CM 1 1 PHl - 240 SN'(HRL I 09'9 -10582 CM- RCIUAL IFR O IF ARMPLITIOF- O0.8 RANGE ACTIUL20 0HOEGRHEE HMMPLIITUOE SCAIFO BIYA FRCTOROF 5.000 0.49 SCRIF0 BY A FACTOROF 5.0 '0 Figure III-12d

"I,mI Variation of Olivine Absorption Feature with Rotation -1 Spectral Region: 9000-9500 cm

" MiI *- 300 SPECTAHL HHNGE1 9479 -9050 CiI ACTURL ZEHO OFGREE AMPLITUE - 0.39 ACTUALZERO OEGREE AMPLITUOE= O.8 SCALEOST A FACTOR OF 3.000 SCALED YTA FACTOROF 3.000

t PHIJ 90 ;PEtlCRi HAN( I 9 '9 -9150 ct ARCTUALZERO DEGRIE AMPLI/UOE - 0.39 SCAL.O BY A FAfClRH(f 3 0410

1 PHJ 60 SPECTRALAT PHJ - 290 SPECTRALRANGE 1 9479 -9050 C t ACTUALZERO DEGREEAMPLITUDE - 0.43 ACTUALZERO DEGREEAMPLITUDE , 0.43 SCALFOBY A FACTOROF 3.000 SCALEDOBT FACTOROF 3.000 Figure 1II-12e uL Variation of Olivine Absorption Feature with Potation -1 Spectral Region: 7000-7500 cm-1

4 1 PHI - 300 SPECTRARL RANGE 74119 -7010 CKI PHI - 120 SPECTRALRANGE 1 7491 - 7016 CW ACTUAIt ZRO 0EGREE AMPLITUE 0.42 RC1TU ZERO OEGREE AMPLITUDE * 0.44 SCALE0BY A FACTOR OF 6.000 SCALEOST A FACTOR OF 6.000

H \

PHJ 90 :I'LL1HRAHIONIt 7491 - 1018 C1-1 A( IU)1 lHf 01 ;11 1 HMiI'llT I1117t . 0.1 SARIO BY FiTH(10 O 6.0001Ii

1 1 PHI = 60 SPECTRRLRANGE t 7401 -7018 CK- PHI] 240 SPFCTIHL RANGE t 7491 -7018 CM ACTUAL. ERO OEGREE AMPLITUOE - 0.47 ACTUAL ZEROOEGREF AMP ITUOF - 0.43 SCALfO BY A FACTOR OF 6.000 SCALE BT A FACTOR OF 6.000 6. 111-19

center of the absorption feature (8000 - 10,000 cm-l) is dominated by this pattern of variation which maximizes at about 9500 cm-1 . The variations of this spectral region in both 0 and / indicate that a very strong maximal lobe is directed approximately toward G 1000 and *1200. In his discussion of crystal field splitting and molecular orbital hybridization Orgel (1970, p. 175) noted that for an octahedral site flattened along a direction perpendicular to one of its faces (such as the M2 site), the groundstate is a pure dz2 orbital oriented such that its major axis passes through the center of this face (more or less) with the higher energy states being hybrids of the other four orbitals. In the olivine M2 site, this axis of compression is oriented approximately through the centers of the facets of the site subperpendicular to the a-axis (see figure 111-13). Since the groundstate dz2 orbital for this site will be oriented with its major axis parallel to the axis of compression, the major lobe of this orbital will be oriented subparallel to the

'a' crystallographic axis but tilted significantly (20-300) toward the 'b' crystallographic axis. Thus the dz2 orbital will be oriented with its major lobe toward 5 oo00and 6 -.1200. If one considers that a polarized photon is composed of a magnetic field vector (parallel to the polarization direction) and an electric field vector (perpendicular to the polarization direction), it can be quickly seen that the variation in the intensity of the transition (band II) mimics the change in III-19a 9-7

Figure III-13

The M(2) Cation Site in Olivine

(Birle et.al., 1968)

43 Si 3.40 3.40

0(3) 400 3.66 50 M(1)* M(1)3. 2 00 3.59 3

57 0(2) 0(4) S 93 2 82 Si 3.28 72 127 07 3.27

00. 3.59 M(1)e M(1e 50 3.23 100 3_60 0(3) 28 I'C'

43 Sl 3.40

'a'

Interatomic distances in the olivine M(2) site. Two digit numbers adjacent to the ion identification symbols refer to the vertical coordinates of the ion. The site (octahedron) is flattened along an axis through the centers of the front and back faces on this figure. The crystal axes are indicated. 9-8 111-20

shape of the probability surface of the ground state orbital as seen by the electric field vector. That is to say, as the polarization is rotated about the b-axis from the a-axis through the c-axis, the electric field vector is rotated through the maximum lobe of the ground state orbital. These results imply a direct coupling between the electric field vector of a photon and the radial probability of the ground-

state orbital. This can be stated in the form of the following rule governing angular variations in transition featuress

The probability of a photon of the appropriate energy interacting with and being absorbed by an electron in the groundstate orbital of a transition metal ion in a

given site in a silicate structure is proportional to the magnitude of the angular probability of that orbital in the direction of the photon electric field vector. Thus for a given transition, the variation in the magnitude

of the absorption feature with angle will map out a cut through

the ground state orbital of that site. The implications of this observation are truely stagger-

ing. We have mapped out a 'mathematical fiction', an electron distribution defined on a purely theoretical basis from quantum theory. This technique for the detailed examination of the variation in the transition features of a material can be used to test the assignments of crystal field tran-

sitions. In this regard a reexamination of the interpretations of the crystal field spectra by Burns (1970b) and III-21

Runciman et. al. (1973) may be undertaken. Burns's assignments of an Al-- E transition in the M(2) site to account for the maximum in the X - spectrum seems to be confirmed,

Runciman et. al. seem to accept this assignment as well.

However, the assignment of the maxima in the ok andA -spectra to transitions from an Eg groundstate to Alg and Blg excited states in the M(1) site is not so securely based. The be- -1 havior of the absorption feature in the region 13000-14000 cm-1

(figure III-12a) is distinctly different than in the region

7000-7500 cm 1- (figure III-12e). Since these are attributed to the same groundstate orbital they should behave in a similar manner. However this is inconclusive because the amplitude of the variation in figure III-12a is a much smaller effect (notice the scaling factors) than exhibited in figure III-12e, where the figure is dominated by spill-over from the center transition. As a result the nature of the variation in the region designated band III, cannot be determined. However, the variation for the region designated band I is quite consistent with a dxy -type electron distribution as is implied by the assignment by Burns. Since Runciman et. al.

(1973b) do not provide an alternative interpretation, we must acknowledge that the assignment by Burns (1970b) is the more nearly correct interpretation. The question can be settled more definitely by examining the polarized spectra of olivine using the a-axis as the rotation axis. This program will be 111-22 IQQ

undertaken in the near future.

g) Summary and Conclusions A detailed study of the angular variation of the trans-

mission spectra of oriented olivine crystals has lead to a

rule governing the relationship between photon interaction with a transition metal silicate and the vector orientation

of the photon with respect to the silicate structure. This rule implies that the transmitted component of the reflection

spectra is the sum of the components due to each transition

multiplied by the relative number of transitions. Thus the reflection spectra of a pure mineral phase can be understood

in terms of crystal field theory and the characteristics of

the transmission spectra of a mineral can be converted into the reflection spectrum by a simple rule. Alternatively, the

reflection spectrum contains all the necessary information to determine the transitions involved in its production. Finally, a method of mapping out the groundstate orbital for a particular transition is defined, which should be applicable to any type of crystal field or molecular orbital transitions.

References

Ballhausen, C.J. (1962) Introduction to Lizand Field Theory, (McGraw-Hill)

Birle, J.D.* G.V. Gibbs, P.B. Moore and J.V. Smith (1968) Crystal Structure of Natural Olivines, Am Min, 3, 807 III-23

Burns, R.G. (1965) Electronic Spectra of Silicate Minerals, Applications of Crystal Field Theory to Aspects of Geochemistry, Ph.D. Diss., Univ. Calif., Berkeley, California

(1966) Apparatus for measuring polarized absorption spectra of small crystals, J. Sci. Instrum. 43 58-60

, M.G. Clark and A.J. Stone (1966) Vibronic Polarization in the e electronic spectra of gillespite, a mineral containing iron(II) in square-planar coordination. Inorganic Chemistry 5 1268

(1970a) Mineralogical Aoplications of Crystal Field Theory, (Cambridge University Press)

(1970b) Crystal Field Spectra and Evidence of Cation Ordering in Olivine Minerals, Am Min , 51, 1608

(1973) The polarized spectra of iron in silictess Olivine. A discussion of neglected contributions from Fe + ions in M(1) sites. (submitted to Am. Min.)

Cotton, F.A. (1971) Chemical Aptlications of Grout Theory, (Wiley-Interscience)

and G. Wilkinson (1966) Advanced Inorganic Chemistry, (Interscience)

Dunn, T.H.; D.S. McClure and R.G. Pearson (1965) Some Asnects of Crystal Field Theory, (Harper and Row)

Figgis, B.W. (1962) Introduction to Ligand Fields (Interscience) Hunt, G.R. and J.W. Salisbury (1970) Visible and Near-Infra- red Spectra of Minerals and Rockss I) Silicate Minerals, Modern Geology, 1, 283 Orgel, L.E. (1970) Introduction to Transition Metal Chemistry, (Methuen) Runciman, W.A.; D. Sengupta and M. Marshall (1973a) The polarized spectra of iron in silicates. I. Enstatite, Am. Min. 58 444-450

(1973b) The polarized spectra of iron in silicates. II. Olivine, Am. Min. 58, 451-456 White, W.B. and K.L. Keester (1966) Optical Absorption Spectra of Iron in Rock Forming Silicates, Am. Min., 51, 774 1Q2 11I-24

(1967) Selection Rules and Assignments for the Spectra of Ferrous Iron in Pyroxene, Am. Min., 2, 1508 Yoder, H.S. and Th.G. Sahama (1957) Olivine X-Ray Determina- tive Curve, Am. Min. 42, 475 VOLUME 2

A SYSTEMATIC STUDY OF THE SPECTRAL REFLECTIVITY

CHARACTERISTICS OF THE METEORITE CLASSES

WITH APPLICATIONS TO THE INTERPRETATION OF ASTEROID SPECTRA

FOR MINERALOGICAL AND PETROLOGICAL INFORMATION Chapter IV

Spectral Reflectivity of Natural Materials: External Effects

a) Introduction This chapter will discuss parameters such as particle size, packing, illumination angle and particle (mineralogical)

mixing, which serve to modify the spectral reflectivity characteristics of the surface as discussed in the previous chapter.

Adams and Filice (1967) have made a study of these parameters for a number of silicate rock powders and their work is a good background for this discussion. These parameters cannot them-

selves account for spectral bands, such as the one micron feature of olivine, but under specific conditions these features

can be essentially removed from the spectrum. Each parameter will be discussed separately and in reasonable detail, but references will be cited for more complete information on any

particular aspect. Necessarily, much of this discussion is qualitative to a large degree. However this discussion should

define the problem of asteroid surface properties well enough within the present level of knowledge. It is important that these considerations be discussed because they directly involve the applicability of this interpretative technique. To ignore them leaves the validity of any mineralogic or petrologic determinations made by this technique open to question. b) The nature of light reflection - Dielectric materials At the interface between two media with different indices of refraction, light can be reflected or transmitted. IV-2 LQ4

The nature of this refraction-reflection effect can be well understood in terms of Fresnel's equations. This material has been covered in a number of texts (Stone, 19631 Born and Wolf,

1964; Garbung, 1965; plus most texts on optics). The most extensive mathematical treatment of the problem of reflection

from a particulate surface has been undertaken by Aronson et al (1973) in which they abandon the traditional Mie and Rayleigh

scattering approaches in favor of a ray tracing method which

takes into account the average particle size, shape and packing. They have compared their calculated models with equivalent materials measured in the laboratory and have found good agreement. Unfortunately their work has been done in the region of

10 microns and longer wavelengths so it is not directly applicable to this study, but their basic approach is illustrative. Also they have worked only with minerals of marginal interest to this study (e.g. A1 2 03 ). c) Particle size

It is clear to anyone who has spent any time pounding rocks, that pulverization tends to increase their albedo.

Adams and Filice (1967) made a systematic study of the albedos of a number of rock types. They showed that the change of albedo with size is strongly dependent on the rock type

(figure IV-la) but that the general curve shape (position and relative intensity of features) is preserved over a wide range of particle size (figure IV-lb)

These effects can be understood by consideration of IV-2a

Figure IV-la

12 11 10 9 8 7 6 .5 4 3 2 1 I I , ! ...... , . , , . 20 50 o00 500 1000 Mean Partice Diameter,p Variation of albedo with mean particle diameter. Sample numbers labeled 1 through 12 appear above micron scale. Of the samples studied, Pisgah basalt shows the least change in albedo with pulverization. Mono Craters obsidian shows the most change. (Adams and Filice, 1967) Figure IV-lb

OBStIIAN R4YOLITE TUFF so (YucouMn)

10.

II5

,.3

04 7 LO LS 20 Wo"OongtN.#L

. . .. Fig._ 3a. .,. , Fig. 3b.

47 GA iRO, BASALT I:5an arcm; If

9 20 10 j] 5 nF -4 r I ... .. I 44 Q 10 L .0 .4 or LD.S ID Waveolnth, WoweI#ngLhl. Fig. 3c. Fig. 3d. Figure shows the spectral reflectance curves representative of (a) glsses (b) crstalline acidic rocks, (c) and (d) crystalline basic-ultrabasic rocks. Numbers after curves refer to the particle size fractions that are given in Figure 1. 106 IV-3 the reflection-absorption processes involved in the imposition of the spectral features onto the transmitted light. The probability of a photon being reflected increases as the number of encounters with interfaces between media of differing indices of refraction increases. A decrease in particle size increases the chance that any photon not reflected at the first surface, will be able to traverse this particle and strike another interface without being absorbed. Hence the reflectivity (albedo) of a particulate surface increases as the particle size and/or optical density decreases. The intensity of spectral features is controlled by the mean path length of a photon through the material and by the optical density of the material at different wavelengths.

From the preceding chapter (especially figures III-4 & 6) it can be seen that the optical density of an olivine (or pyroxene) at 0.7 microns is five to ten times less than that a 1.0 microns. This means that for even rather large particles (on the order of a millimeter) a 0.7 micron photon is very likely to simply pass through the particle and be reflected at the next interface. Thus at 0.7 microns the material has a high albedo. However at 1.0 microns, even a very small particle size ( 30 microns) has a high probability of absorbing the photon so that it is not available for any second interface reflection. This contrast preserves the absorption features in the reflection spectrum over a large range of particle sizes. IQ7 IV-4

It is only when the extremes of the particle size

distribution are considered that the spectral features are absent (or very weak). For very large particles (a few milli- meters) the mean path length is so long that even low optical density photons are likely to be absorbed, making the albedo

low (relatively) and the features weak or absent. This case can be ignored with reeard to asteroid surfaces since a) the average particle size should be smaller than is required to produce this effect (see Chap. V) and b) the typical crystal

size in meteorite specimens is much smaller than this ( 5 100 microns) and crystal-crystal interfaces will act to cause re- flections. For very small particle sizes ( 10 microns) even the high optical density photons are likely to be able to pass through the particles without significant absorption and be reflected, resulting in a high albedo for the material and weak features. The variations in the spectral reflectivity of a particulate surface with particle size is shown schematically on figure IV-2. It should be noted that the optical density of pyroxene is approximately an order of magnitude higher than that for olivine. In the course of this work, a number of particle size distributions for several meteorite specimens were utilized and the albedo effect was clearly noted (table IV-2). However, IV-4a S08

Figure IV-2 Schematic Representation of the Variation in the Spectral Reflectivity of a Particulate surface with Particle Size.

-P-4

I-'H

r() rWQ)

Wavelength

The absortion feature is assumed to be triangular in shape. The effect of decreasing particle size (or optical density) is a general increase in the reflectivity of the surface. The intensity of the absorption feature maximizes for a size (or optical density) where the di±ference between the relative opacities of the material in and out of the absorption feature is the greatest. IV-5 109

Table TV-1

Spectral Reflectivity (0.56Am) of Various Meteorites

Particle Size

Coarser 0.117 0.122 0.296 Mixed 0.215 0.228

Finer 0.125 0.147 0.241 0.250 0.353

the relative reflection spectrum is generally not strongly affected, the major change being a slightly higher reflectivity at longer wavelengths (relatively redder) but the position and relative intensity of spectral features are unaffected. These results and the work by Adams and Filice suggest that for silicate materials particle size distributions larger than approximately 30 microns do not show significant variations in the relative reflection spectra due to particle size variations but that very small particle size distributions will show distinctly different spectra. To get a physical impression of these results, the typical particle size in the lunar soil is on the order of 50 microns. d) Particle racking effects Adams and Filice noted that the effect of packing a rock powder more tightly is to increase the albedo slightly in the visible and to decrease it slightly in the infrared

(figure IV-3). Adams and Filice suggest that since the O0 Figure IV-4 (Adams and Filice, 19671

co0 1.26

Figure IV-3 I.'Z .22 -

t.1 -

1.14 -

0 1.10 BASALT (Little Lake)

1.0 1 1 1 1 1 10 20 30 40 50 60 70 80 90 i, deg I, deg

I I I I I I -

1.30 1

1.18 Wavelength,1L

Spectral reflectance curves illustrating ef- 1.14 fects of different packing of <40-,a powders.

OBSIDIAN IMono Croters) S I I I I 20 30 40 50 10 20 30 40 50 so TO SQ (Adams and Filice, 1967) ' i, deg I, dog R/IB variations with nigle i for sifted powthrs of sizofractionl 12 (<40 p). Iv-6 tighter particle packing increases the effective optical path length, the powder is less reflective in the spectral region (visible and ultraviolet) where the optical density of silicate materials is highest. The magnitude of the variation due to this effect is about 4% and it is unlikely that this maximum variation can be achieved in actual practice, since similar surface conditions (no atmosphere and a particle flux) should lead to similar particle packing on the individual asteroid surfaces. Therefore, in practice this effect is almost certainly ignorable. In any case it shows no evidence of effect on the presence of spectral features. e) Illumination anile effects Adams and Filice noted that the ratio of reflectivity of a particulate silicate surface at different wavelengths

(0.7 microns, red or R and 0.4 microns, blue or B) varies as the illumination angle (the angle between source, surface and detector) changes. This ratio, R/B, increases slightly toward

300 and decreases thereafter. The magnitude of the variation depends on the type of material (figure IV-4). Adams and

Filice correlate this effect with the variation in the optical path length through the powder, noting that the R/EB ratio is highest where the path length is longest and the blue is relatively more absorbed. Since the phase angle over which a belt asteroid can be observed is about 300, this effect which is small over this range of illumination angles, can be con- IV-7 sidered negligible. The variation of albedo with illumination angle is quite well understood in terms of shadow effects by the particles on the surface. This effect is clearly noticeable on the lunar surface at different phase of the moon. f) Mixing of mineral phases This effect is the most significant of the external effects since it has implications both for the measured spectra and any modeled spectra which are used in interpretation. An examination of the spectral reflectivity curve for an ordinary chondrite (figure IV-5), which has slightly more olivine than pyroxene in the meteorite, shows that the strong absorption features at 0.95 microns is almost entirely due to the absorption in the pyroxene phase. The broad feature due to olivine, centered at 1.0 microns, is not a strong contributer to the whole rock spectrum even though this is the more abundant phase. It is therefore clear that the two phase are not contributing equally to the spectra but that some other effect besides relative abundance is operating. From discussions of this effect with other individuals concerned with this problem, particularily McCord, Adams, Pieters and Charette, the following model emerges. For a given mineral (ignoring for the moment reflection and scattering effects) the relative strength of an absorption feature is a function of the mean path length of a photon IV-7a

Figure IV-5

SPECTRAL REFLECTIVITT (SCALED TO 1.0 AT 0.56 MICRONS)

0(D tJ r

0C) n

tIH-D H- CD,r ( rt-

0 H oc

U,lu w 0

p g H (D ri H- a0 - I-1 t 0 D 0 LCpi(D C) ruR~) S U,1 o- ct 0

ALFIANELLO L6 14 9 10 PTROXENE (HYPERSTHENE - JOHNSTOWN) 010 15 14 29 OLIVINE (CHASSIGNT) CHRS 27 12 31 ANORTHITE FELD (LUNR 12063 - AM.1971) AN 0 0 0 114 IV-8

through the material in relation to the variation in the optical density of the material over the whole wavelength range under consideration. It is clear that the optical density of a material is greater in an absorption feature (Df) than it is in the region outside the feature, the continuum (Dc), that is Df > Dc. If the mean path length of a photon, related to the particle size, is very large relative to Dc, then no light at all will be transmitted at any wavelength and the material will have a flat dark spectra. If the mean path

length is short with respect to Dc but long with respect to Df, the light will be transmitted in the continuum but absorbed in the feature to produce a very strong (relative) feature.

If the path length is short with respect to both Dc and Df there will be no strong absorption and the spectra will be bright but featureless. These cases are shown schematically on figure IV-2 for a hypothetical mineral phase. In the case of olivine and pyroxene, the optical density of pyroxene is about an order of magnitude higher than that of olivine. This can be noted from work on transmission spectra by Burns (1970) where the thickness of thin sections used to obtain spectra of olivines are about ten times thicker than those used for pyroxene of the same Fe/Mg ratio. The result is that in a mixture of olivine and pyroxene with ~ 100 micron crystals and/or particles, the pyroxene grains given significant features while the olivine grains are nearly transparent at all the wavelengths measured. Since this is the IV-9 115

size range of grains typically found in lunar and meteoritic materials, the pyroxenes will control the spectral features observed in these materials. Therefore, to make theoretical calculations of whole rock (compositionally whole) spectra from measured spectra of pure mineral components, one must consider the optical depth/grain size ratios for the individual components. Figure IV-6 shows a schematic production of a reflection spectra of an ordinary chondrite (45% olivine, 45% pyroxene, 10% feldspar) with the optical density of the individual phases considered. While this model is not to be considered as a serious model but rather as an indicator of the process involved, it is nevertheless indicative that this basic approach is correct. The effect of adding an opaque phase to the mix has been discussed by Johnson and Fanale (1973). They studied the carbonaceous chondrites which contain dispersed opaque carbon compounds. In their model they added lampblack (carbon) to montmorillonite (a hydrous layer lattice silicate such as is found in these meteorites) as shown in figure IV-7. It is clear that the opaques present have very high optical densities and thus they dominate the spectrum. The water features in the montmorillonite spectra are completely absent from the bombination spectrum. This same effect has been noted in the lunar soil where the albedo and the contrast of the spectral features have been lowered by the presence of dark glass fragments and agglutinates (see chapter V). IV-9a )LL6 Figure IV-6

Schematic Production of Meteorite Spectrum from Component Mineral Spectra.

(Continuum is assumed to be linear in energy spacel

PYR

OL 1 .4

I I I I

1 n 1_5 2.0ww v 2.5 0.5 .v a Wavelength (Microns) IV-9b

Figure IV-7

Production of Artificial C-Type Chondrite Spectrum.

100

0.5 1.0 1.5 2.0 2.5 3.0 WAVELENGTH, /,m

CJohnson and Fanale, 19731 IV-10 118

Thus the reflection spectrum of a particulate mixttre

of several mineral phases is a function of the phases present, the amount of each phase and the ratio of the optical density to particle size for each phase as well as the amount and distribution of opaque phases. One of the goals of future work should be to define this functional relationship analy- tically. g) Metal ohases

The preceeding discussion was carried out with regard to dielectric materials, particularly silicates, which are the predominant terrestrial and lunar rock forming minerals as well as a substantial fraction of the mineral phases of the meteorites. But also present in the meteorites (and presumably, the asteroids) are metal phases composed of iron, nickel and cobalt. These conductina minerals must be treated quite differently than the dielectric phases.

In a perfect conductor, incident electromagnetic radiation will, by virtue of the varying magnetic field, generate a localized current such as to just counteract the EM wave. Upon the collapse of the current, an EM wave is generated which appears as a reflected light ray. These effects are described by Maxwell's equations. For real metals, which are not perfect conductors, the situation is more complex. At low frequencies (61000 cm'l) these metals still behave very much like perfect conductors. However at the higher frequencies of the visible and near-infrared radiation, 119 IV-11 119

the free electrons in the metal (conduction band electrons)

are unable to respond rapidly enough to the time varying

magnetic fields of the incident light to accomplish perfect

cancellation and reradiation. Therefore, in general, the

reflectivity of metals decrease toward shorter wavelengths,

although the particular variation of reflectivity depends on

the particular metal. For a number of metals measured in this study, including aluminum, gold, nickel and iron, the spectral reflectivity curves are shown on figure IV-8. The behavior of these metals is thus understandable as a reflected component and an absorbed component with no transmitted component. A discussion of the reflection of metals in terms of the number of free electrons car be found in Burns and Vaughan (1970). They attribute the break in the slope of the reflection spectra of metals such as gold or copper (accounting for their color) to transitions between occupied outer d-level orbitals and unfilled s-p hybrid orbitals. They also noted that the reflectivity of a metal was correlated with the number of free electrons per unit atom, the greater the number of the electrons the higher the reflectivity. The addition of these metallic phases to a silicate mix, must be treated quite differently from the opaque phases which they superficially resemble (very high optical density). Since these metals have no transmitted component but only the strong reflected component, they do not dominate the spectra such as the opaques. They may alter the spectra: the higher IV-lla

Figure IV-8 SPECTRRL REFLECTIVITT (SCRLED TO 1.0 AT 0.56 MICRONS) o a *)l .1 1 1 11 _

i i 00 i i 'i i i l i

ru w 4 I-

I II I I I I I I I I I I I METRL: GOLD 0.433 STD 28 1 1 METAL: ALUMINUM 0.505 MET 18 2 2 METAL: NICKEL 0.365 MET 18 1 1 METAL: IRON 0.321 MET 18 3 3 121 IV-12

slope and reflectivity of the enstatite chondrites versus the enstatite achondrites may be attributable to the metal present in the former meteorites (see chapter 8).

h) Conclusion Of the physical variations which will effect the

spectral reflectivity of a surface, petrological variations are the most important in determining variations in the spectral

reflectivity. However these variations are not linear with the amount of each phase present. Particle size variations have a strong effect on the albedo of a surface, but do not

strongly affect the relative reflection spectra except at the

extremes of the size distributions. The presence of dielectric opaques, such as carbon, disseminated throughout the material

tend to dominate the spectra, removing features and resulting in a dark, featureless spectra. The presence of conductive phases, such as the metallic minerals, do not dominate the spectra but can significantly modify the total reflectivity as well as the relative reflectivity.

References

Adams, J.B. and A.L. Filice (1967) Spectral reflectance 0.4 to 2.0 microns of silicate rock powder. J. GeoThy. Res. 72 5705 Aronson, J.R.: A.G. Emslie and E.M, Smith (1972) Development of a theory of the spectral reflectance of minerals. Applied Optics 12 2563-2584 122 IV-13

Born, M. and E. Wolf (1964) Principles of ptics, Pergamon Press Burns, Roger G. (1965) Mineraloical Anplications of Crystal Field Theory Cambridge University Press

and D.J. Vaughan (1970) Interpretation of the reflectivity behavior of ore minerals. Am. iTn. j5 1576 Garbung, M. (1965) Optical Physics Academic Press, New York

Johnson, ToV. and F.P. Fanale (1973) Optical properties of carbonaceous chondrites and their relationship to asteroids. J. Geophy. Res. (in press)

Stone, J.M. (1963) Radiation and Optics McGraw-Hill 123 CHAPTER V

A.tero'l !urface: Physical and Litholozical Characteristics

a) IntreCtion Vtry little is known about asteroid surfaces from direct measurements. All information concernine the surfaces of these

bodies comes from two sources: (1) The first is obtained by consideration of the population of objects and the dynamics of interactions. (2) The second group is inferred from the

character of light reflected from the surface: wavelength dependence, phase and polarization effects as well as the thermal

flux and radar reflectivity of the asteroidal surfaces. From these sources a picture of the surface characteristics can be

built up. A review of much of the work and thought concerning the nature and texture of asteroidal surfaces can be found in Chapman (1972). This chapter will, for the most part, consist

of an integrated review of the results of pervious workers in this area.

b) Surface Texture and aterials: Dynamic Considerations The manrer in which a mineral phase imposes a characteristic sinatulre onto the reflection spectra is influenced by

the particle sJfe distribution (see Chapter IV) of the material.

Thus it is neceF:ary to determine, to the extent of available knowledge, the i-xture of the surface of the astroidal bodies, whether one is 5faling with a bare-rock surface, a surface of

very fine powd~e or some intermediate case.

From ever 'he most cursary examination of the size-number "24

V-2

(or analogously the magnitude-number) relationships for the asteroid population statistics, it should be clear that the number of bodies increases very rapidly as the size decreases

(ANcAD2 ). On these grounds it is reasonable to posulate that a substantial portion of the mass of the belt is represented by small particles. Thus in the asteroid belt, the flux of small particles is very high relative to the flux in interplanetary

space. This reasoninz is supported by direct measurements of the population density of small particles in the belt measured by

the meteoroid detectors carried by the spacecraft, Pioneer 10, which passed through the belt on its way to Jupiter in late 1972.

The importance of such a distribution of small particles

lies in their interaction with the larger bodies in the belt which can be studied telescopically. An examination of any cata-

log of asteroid orbital parameters (eg. Ephemeris of Minor Planets) shows a significant amount of scatter in the elements

of the orbits. Thus one obtains a picture of a complex pattern of (in theory) intersecting orbits. The relatively small number of large (observable) asteroids excludes, for all practical purposes, collisions between these on any time scale except geo- logic. The case is quite different for the probability of collision between these larger bodies and the smaller particles in the belt. The smaller particles will show at least the same degree of dispersion in their orbital elements as the larger bodies. Thus it is reasonable to picture the surface of any 125 V-3

of these larger asteroids as beina constantly bombarded by a flux of smaller particles as the larger body intersects or is intersected by the orbits of these smaller particles.

Certainly the closest analogs to asteroidal bodies (small

airless bodies) about which we have direct information, the Martian moons, Deimos and Phobos, show evidence of intense bom- bardment. The Mariner 9 images of these moons show a high den-

sity of craters with diameters down to the resolution limit of the pictures (Pollack et al., 1972). Since in the case of

lunar craters (again on an airless body), this distribution of craters continues to increase rapidly in number down to the

smallest sizes, it is reasonable to conclude that such is also the case for Deimos and Phobos and by extension for the asteroids as well. In the case of the asteroids, however, the flux unto the surface should have a much greater number density, although the Pioneer 10 measurements suggests that the actual number density is less than that postulated on the basis of the population of larger objects (Van Houten et. al., 1970). It is necessary to consider the nature of the interaction to determine their effect on the surface. Marcus (1969) has discussed accretion and erosion on airless bodies in an environment of hypervelocity impact. Dohnanyi (1971) has modeled collisional interactions in the asteroid belt based on realistic distributions of asteroidal orbits and has arrived at an average interaction velocity of 5 km/sec. This agrees with the value V-3a 126 Figure V-I (Marcus, 1969)

0.1 1.0 10 10 PLIANETARYESCAPE SPEED ,, KMI/SEC

100 200 500 1.000 2,OCO 5.000 10.000 20.000 3 PLANETARY RADIUS, EM.FOR P = 46/Cn Fl. . Ratio of ma. lost from planet to mass of projectile (basalt target). ' \ 10 0 .-

0. 10 10 100 0.1 I0 10 i PLANETARY ESCAPESPEED Ve,KM/SEC PLANETARY ESCAPE SPEED ve.KM/SEC. C'- t, tit I t1

Bit . as A U a

. Ratio of mass lost from planet to mass Ratio of mass lost from planet to mass of projectile: unbonded sand target, hypothetical of projectile: unbonded sand target, hypothetical velocity distribution 2. ' velocity distribution 3. V-3b

Figure V-2

Velocity Distributions of Ejecta from Hypervelocity Impacts

G(V)

C.i 0.2 0.5 10 2D 5D 10 20 0.5 10D 2.0 5.0 10 EJECTA SPEED V, km/sec EJECTA SPEED V, km/sec Fraction of mass ejected at speeds greater than i for hypothetical ejecta velocity Differential ejecta velocity dis- distributions. tribution.

CGault et.al, 1963) 128 v-4 calculated by Piotrowski (1953). Two limiting cases exist depending on the ratio of the escape velocity from the surface of the body (Ve) and the impacting velocity of the particle onto the surface (Vi). a) V i , Vet In this case the impact velocity is primarily an effect of the gravitational acceleration of the particle by the asteroid. Since the collision with the surface is quite inelastic (energy is dissipated as heat and in fracturing the materials) not enough energy is retained in a kinetic form

(as ejecta) to achieve escape velocity for any significant fraction of the mass of the impacting body such that most or all of this impacting material is retained on the surface of the impacted body. In this case the surface of the impacted body would consist of a framental mixture of infallen material and underlying 'bedrock' with the infallen material composing a significant fraction or the main part of the surface material. Thus in general the observable surfaces of these bodies would exhibit petrologies which tend toward an average petrology for the small bodies interacted with (and probably for the small particles of the belt as a whole). b) Vi >Ve, In this case the impacting velocity would be primarily due to differences in orbital velocities with only a small component due to gravitational acceleration by the larger body.

Since the energy available on impact is proportional to the second power of the impact velocity (kinetic energy), the collision can be quite inelastic and still eject a substantial amount of 129 V-5

material with velocities greater than Ve . Marcus (1969) has calculated models for the loss rate from a surface as a function of the ratio (Vi/Ve) for various surface textures based on experimental data by Gault et. al. (1963). Although the results by Marcus (fiure V-1) are idealized cases, they do provide quantitative constaints on the asteroidal surface conditions.

Marcus has shown that ratio of ejecta mass to impacting particle mass is high for typical asteroid sized bodies at moderate interaction velocities, resulting in a net loss of material to the body. Gault et. al. (1963) showed that a substantial fraction of the ejecta from a hypervelocity impact is ejected at low velocities (figure V-2). This material ejected at less than escape velocity will be returned to the surface. Thus the surface of an asteroidal body subject to these conditions will again be fragmental but will be composed mostly or entirely of underlying material, and given reasonably hiEh flux rate, will tend to be of relatively 'fresh' underlying material. To determine which case will predominate on a particular body, one must consider the relationship between the escape velocity of the body and the typical orbital interaction velocities of the particles in the belt which impact on the body. Table V-1 contains the calculated escape velocities for spherical bodies with a range of radii and densities. The sphere was chosen because it has the highest escape velocity per unit surface area per unit mass of any natural shape (a surface which is always convex or flat). All other 'always convex' shapes have higher V-5a13

9..k#.Aa 0 .- a'p I*., 0.-I.4 0# a. 0o''D

0 -~.9 M %.d.1 00 W 0 C 4*% ~ 0I 'a 040%^%^%^ *~~ 4P. ~~'49.46 ~~~ N'.%,-b#%403 ^J 0d 00 0 @ ~~ d

......

016 "SAl A)%^m.M-A~ 9. 0a iP644 :ot

~O*0A- A) J~ J . . .J3. p. ...4 '4 ... 4 0 . *. * '4.

'4JJS. ftta0***j*

*-.AO-.0A0--a*I U V- 0 3 O 0', m 3d , -. 0 W- ChN% *Mik m0ft'J'4 .to0 lA '09 %

-Qp~".aR -4.10 '4-9e ,. a m-w-)-30. 0 Z- 2 0.A-A,* 0. cmZO3 Q%.2

r_ 0 0*#. ZA ftm 4N15Z14 -' "%.a o m-.. a 001 =4 OAA 4r%*W. 10 W""' ~ ~-N. - ~ 0-.a -43 0-J I9. 4 .. A %-0-to30-~.u .010 -2Oat 9allowM 0..109.1*a%

-% a 3

N) u.'3 cD0.. ' a0Oij.J0IQ rQr4 3,Q .~9 0 .- Ml.Jt a 0' .04%d pR V-6

Ve at all points on their surface than the sphere for the same mass and density. Thus the spherical case is the limiting case and produces the highest escape velocity for a particular mass. The calculations were also made using the assumption of homo- geneous bodies. Since variations of the density in a body which are not radially symmetric, are as likely to increase Ve at one point on the surface as they are to decrease it at some other, these type of variations can be ignored when dealing with the generalized body. Table V-2 gives the typical densities of materials which might be encountered at this distance from the sun.

Table V-2 Densities of Solar System Materials

Material Density (gm/cm3 )

'Ices'(H 2 0,CH4, NH3 ) 0.8-1.0 Carboneaceous Chondrites, Cl 2.20-2.42 C2 2.57-2.92 C3 3.40-3.78 Meteoritess Stones 2.95-3.90 " Stony-irons 4.60-4.90

" Irons 7.70-7.90

Accepting the above figures for typical impact velocities and typical escape velocities as reasonable estimates, it should be obvious that for all but the largest asteroids, 32 V-7

case 'b' (Vi$OVe) will dominate the surface characteristics. The diameters of asteroidal bodies are a matter of some

dispute, since most measurements in the past have been made by means of micrometer measurements of the apparent size of the objects in the sky, where these diameters are equivalent or

smaller than the typical atmospheric 'seeing'. Since the image

size was so small, large uncertainties could be introduced into the calculation of the diameter. (A description of the most

sophisticated utilization of this approach and the degree of uncertainty remaining can be found in Dollfus (1971a).) Re-

cently a new technique involving the infra-red emmisivity of these bodies (Matson, 1971a; Morrison, 1973; Cruikshank and Morrison, 1973) have produced new diameter determinations which should have significantly smaller uncertainties. These are reproduced for several asteroids in table V-3 which also gives

Ve and the ratio Vi/Ve for each body listed. It is clear that Table V-3 Dynamic Characteristics of Nine Agteroids with IR Measured Diameters Asteroid Albedo Radius(km) V (m/sec) V g Ceres 0.07+.01 500+40 80-8oo 3 *. Pallas 0.09+.01 270+20 320-430 15 -11 Vesta 0.22+.03 260+20 310-410 16 -12 Davida 0.047.01 176T12 200-270 25 -18 Euromia 0.15+.02 125+8 140-220 36 -23 Juno 0.15+.02 115+7 135-180 37 -28 Hebe 0.16+.02 10076 120-i60 42 -31 Nemausa 0.05+.01 72+8 80-110 63 -45 Eros 0.06+.02 13+2 10- 25 500 -200 * Escape Velocity (Ve) calculated for density range 2.5-4.5 ** Ratio of impact velocity to escape velocity calculated assuming Vi = 5.0 km/sec •** Asteroid iameters and albedos from Cruikshank and Morrison (1973) V-8 133

only in the very large bodies is there any question that case 'b' will control the surface parameters, and that even in these

large bodies it will probably dominate. Matson (1971a, 1971b) showed that at least one very faint

asteroid (324 Bamberga) with an apparent magnitude nearly 4.0 magnitudes dimmer than 1 Ceres, which would suggest a very small object, has a very low albedo and is actually one of the largest

asteroids. Even if this pattern is common, i.e. low albedo ob-

jects, leading to a gross underestimation of the asteroid diameter, it is still very unlikely that any bodies can be found in the belt where case 'a' dominates the surface characteristics.

While the particle-size distribution (PSD) on the surface of an asteroid cannot be calculated directly, the relative PSD can be approached by the use of scaling arguments such that this information can be used with some assurance to test the degree of probable particle-size effect on the reflection spectra of the material. In his work on the impact mechanism at Meteor Crater,

Arizona, Shoemaker (1963) showed that the behavior of materials during hypervelocity impacts can be understood in terms of shockwave propagation through the materials involved. In this model the material being ejected is accelerated to high velocities by the pressure gradient across the schockwave. The frarmenta- tion of the material is also an effect of this gradient, and in a particular type of material, the 'mean'. particle size resulting is inversely proportional to the gradient of the shockwave through this portion of the material at the instant of material 134 V-9

failure. In this model the gradient of the shockwave, the magnitude of the shockwave and the velocity of propagation of the shockwave through the material are all proportional to the impact velocity. Thus for a lower velocity impact the character of the shockwave is such that the mean particle produced is larger than the mean particle produced in a higher velocity impact. Additionally in the higher velocity impact, these smaller particles leave the site at a correspondingly higher velocity. Since the lunar surface is that of an airless body subject to a particle flux, it can be used in a scaling argument to determine the relative particle size distribution (PSD) for the asteroid surface. Each of the manned moon missions has returned soil material from its landing site. Several of the grain size distributions are shown on figure V-3. In these soils, the median grain size tends to be about 75 microns (half the mass is in grains which are larger and half smaller). With such a grain size distribution the effect on the reflection spectra is small (see chapter IV). For the lunar case, the impact velocity &s 15-30 km/sec or approximately three to six times higher than in the asteroid case. It is therefore reasonable to expect the PSD on an asteroidal surface to be at least as coarse as that on the lunar surface and probably more so. Such a distribution will not modify the relative spectral reflectivity from such a surface to any significant extent. Recent work on the spectral reflectivity of lunar surface Figure,V-3

Grain Size Distributions for several Lunar Soils

.01

Sco -4'3 -2 -1 0 2 3 4 5 6 7 6 9 g0 * - -Re. 44 r sa'mple.. .- la--a 0 L--Glocial tisl #3; Apollo 15 S70.- Desertresiduol soil 60 #1: Apollo 11 o- 40 0 / 30

g to [**

%o.0 8.0 8 0.5 0.12600312 0.0078 0.00195 so Groin size (mm)

o i1 t ,90 - m[ 814141 70- #2: Apollo 14' d

50 40 - 14149 999/ 3 9 30 4148 16 4 1 .25 .0625 15.6# . # 0 Grainsize (mm) S, / 1 163 Grain sizedistributions for repre- 4 w sentativesoil samples from stations2, 6, S 7, 8, 9, and 9a and the LMN site plotted .0- 14156 on a probability scale.Most of the curves 0 ,0'- * 140259 are very similar, and the range is plotted / here as a shadedband. The solid line 2.0- ,/ represents the coarsest grained soil from 1.0-- station 9a, and the dashed line rcprescnts 0. , soil from station 8. Median grain sizes 160 &0 4.0 2.0 1.0 0.5 0.25 0.125 0.063 0.031 range from 42 to 98 um. All soil samples Grain size (nim) are poorly sorted. Most of the soils are slightly finer graincd than soils from other landing sites. Sources: #1) Duke et.al. (1970) ; #2) LSPET (1971); #3) LSPET (1972) 136 V-10 material (Adams and McCord (1971a,bs 1973); Conal and Nash (1970) Bell and Mao (1972a,b; 1973) has indicated that the major

factor in the absence of deep absorption bands in the reflection spectra of soils is the presence of a large quantity of glass and agglutinates (glass welded aggregates) in the soil. Figure V-4, from Adams and McCord (1971b), shows the effect of sys-

Figure V-4

35 1 1 , o 12063,79 30 - Whole-Rock Powder a b +20% Gloss

S10 c 55% Gloss .- d 12070,ll Fines"Surface 020 12063,79

0.5 1, 1.5 2.0 2.5 WAVELENGTH (Mm) Spectral reflectivity of glass made from 12063 whole-rock powder; mixtures of rock powder plus 20% glass and plus 55% glass; compared with curve of Apollo 12 surface fines.

tematically adding, to crushed crystalline lunar roe-Rockcks, glass of the same composition. This decreases the contrast of the spectra, wiping out the absorption feature. Adams and McCord (1973) have examined the problem of agglutinates in the lunar soil (Apollo 16) and have concluded that the darkening Of the ;I37 V-11

soil and the absence of spectral features can be attributed primarily to these agglutinates. The question naturally arises as

to how much these effects will modify the reflection spectrum of

an asteroid surface. A scaling approach, from the lunar case, can be used to

approximate the glass and agglutinate content.of asteroidal sur-

face materials. Since the impact melting of a material is an effect of the inelasticity of the impact process, the relative

proportions of glassy material produced on these two surfaces by impacting material will have the same ratio as the impact

energies. Because the impact velocity onto the lunar surface is only slightly less than an order of magnitude greater than

that onto an asteroid surface, the energy available to be dissi-

pated in the melting of material is about two orders of magnitude

higher in the lunar case. Thus per unit mass flux onto a surface, one to two orders of magnitude more glass and agglutinates will be formed on the moon. Two other factors must be considered to make this picture more realistic. The first is that the flux onto an asteroid surface is much higher than onto the moon which would seem to indicate a higher glass content on a

typical asteroid surface - than on the lunar surface. However, the mass loss from an asteroid surface should

overcome the effect of this higher flux. Since the most highly

shocked (and hence the most likely to be melted) material has been subjected to the maximum gradient in the shockwave, it is

the most strongly accelerated material and is the least likely ,38 V-12 material to be retained on an asteroid surface. The second, and perhaps -most important factor, is the threshold effect. That is, that a certain minimum impact velocity is required to produce glass. Horz et al (1971) report experiments which seem to indicate that in the case of small particles, a minimum velocity of 10 km/sec is required to produce the glass lined pits seen in lunar rocks. It is not clear whether this is also the case in the production of the glass fragments found in the s6ils, but it seems to be a reasonable estimate. Since typical impact velocities are 5 km/sec, this would imply that glass is a rare component on most or all asteroid surfaces. The conclusion from this information is that one would expect that the amount of glass formed in the high flux environment of an asteroid surface will be quite small and very little of this amount will be retained on the surface. Therefore from dynamic considerations the following picture of a typical asteroid surface emergess a) the surface is fragmental rather than solid and subject to a high flux of 'low' velocity particles, b) the surface lithology will be essentially that of the underlying material,

c) the patticle size distribution of the surficial material will be coarser than that for the lunar surface and will have little effect on the spectra, v-13

d) the glass content of the surface will be quite low and will not significantly affect the reflection spectra.

c) Surface Texture and Composition - Observational Data Veverka (1971) has indicated that the phase coefficient

(the increase in the amount of light reflected from a surface to an observer as the phase angle (sun-object-observer angle) decreases) cannot be interpreted unambiguously for a typical

asteroid surface. He has indicated that the phase coefficient

may depend as much on the photometric properties of the surface material as on the surface roughness of the body. Since the latter had generally been assumed to be the primary factor, previous work needs to be reinterpreted before it can supply meaningful information on the surface conditions of the body being studied. Polarization variation with phase angle is an effect of several characteristics of the surface, texture and mineralogic composition, but fortunately in this instance work has been done to unravel this ambiguity. Dollfus (1971b) from laboratory comparison of polarization spectra of several asteroids concluded that a) the surfaces were fragmental rather than solid on even the smallest body studied (Icarus, 1 km dia), b) that the relatively deep negative branch of the polarization curves was in- consistant with a glass rich soil such as the lunar soil, and c) that definite compositional differences were indicated. Chapman et al (1973) have carried out the most extensive program of spectral reflectivity measurements on asteroids. The 40 V-14 variation in the shapes of the curves, shown on figure V-5, is strongly indicative of lithological variations between the

individual bodies. This supports the second conclusion reached in the previous section, that the surface material depends on the individual object rather than on any belt wide surface de- posit. The strong features in some of the asteroid spectra and the relatively low slope (not steeply reddened) is strong evidence that glass or agglutinates do not constitute a spectrally important component of the surface material. These observations support the general conclusions reached in the previous section concerning asteroid surface conditions.

Radar measurements of asteroids are still at a very primitive stage. Only two asteroids have been detected by this techniques 1566 Icarus (Pettengill et al, 1969; Goldstein, 1969,

1971) during the close passage of this body in 1969, and 1685 Toro (Goldstein, 1973) during a recent apparition. However this technique has been applied to the lunar surface with good results. With the resurfacing of the giant Arecibo radio tele- scope dish, many of the larger asteroids will be well within the working range of this technique. d) Conclusions

Based on dynamic considerations and supported by observational data, the surface of a typical asteroid seems ideally suited for application of remote mineralogical interpretation techniques. The characteristics of the surface are such that V-14a 141 Figure V-5

Spectral Reflectivity Curves of Several Asteroids (Chapman et.al., 1973) I 1 -0 lu7 543 S EI A 4 S 4 S3Ut.

10 U'

5 10- _ ,fitI i ri- o

t ,,1I - o t j i jt

SLO f ti

04--j2 0 04

5 0 i o. 0. ' '. 09 u 03 0 07 05 5

MELEm6M 1AVELEGTH 0m k42 v-15 variations in the spectral reflectivity are the result of mineralogical and petrological variations rather than physical parameters.

References

Adams, John B. and Thomas B. McCord (1971a) Alteration of lunar optical parameterss Age and composition effect. Science 171 567 (1971b) Optical properties of mineral separates, elass and anorthosite fragments from Apollo mare samples. Proc. of the 2nd Lunar Science Conference 2 2183 (1973) Vitrification darkening in the lunar highlands and identification of Decartes material at the Apollo 16 site. Proc. of the 4th Lunar Science Conference

Bell, P.M. and H.K. Mao (1972a) Crystal-field effects of iron and titanium in selected grains of Apollo 12, 14 and 15 rocks, glasses and fine fractions. Proc. Third Lunar Sci. Conf., Geochen. Cosmochem. Acta SunDi. 2 55-553 (1972b) Zoned olivine crystals in an Apollo 15 lunar rock. The Apollo 15 Lunar Sazmples (editors J.W. Chamberlin and C. Watkins) pp. 26-28 Lunar Science Institute (1973) Optical and chemical analysis of iron in Luna 20 plagioclase. Geochim. Cosmochim. Acta .7755-759 Chapman, Clark R. (1972) Surface properties of asteroids. MIT Phd Thesis

_ Thomas B. cCord and Torrence V. Johnson (1973) Asteroid spectral reflectivity. Astron. Jour. 78 126 Conel, J.E. and D.B. Nash (1970) Spectral reflectance and albedo of Apollo 11 lunar samples Effects of irradiation and vitrification and comparison with telescopic observation. Proc. Apollo 11 Lunar Sci. Conf. Geochim. Cosmochim. Acta Suppl. 1, 2 2013 Cruikshank, D.P. and D. Morrison (1973) Radii and albedos of nine asteroids. Abs. of 3rd Annual DPS-AAS meeting, Tucson, p. 95. V-16 143

Dohnanyi, Julius S. (1971) Fragmentation and distribution of asteroids. NASA SP-267, p. 263 Dollfus, Audouin (1971a) Diameter measurements of asteroids. NASA SP-267, p. 25 (1971b) Physical studies of asteroids by polarization of the light. NASA SP-267, p. 95. Duke, Michael B. Ching Chana Woo; Melvin L. Bind; George A. Sellers and Robert B. Finkelman (1970) Lunar soils Size distribution and mineralogical constituents. Science 167 648 Ephemerides of Minor Planets, Institute of Theoretical Astronomy, Academy of Science, USSR Frondel, Clifford; Cornelis Klein Jr. and Jun Ito (1971) Mineralogical and chemical data on Apollo 12 lunar fines. Proc. 2nd Lunar Science Conf., p. 719 Goldstein, R.M. (1969) Rada observations of Icarus. Icarus 10 430

SP-267, p 165* (1971) Asteroid characteristics by radar. NASA SP-267, p. 165,

(1973) Astron. Jour. in press Htrz, F.; J.B. Hartung and D. Gault (1971) Micrometeorite craters on lunar rock surfaces. J. Geophvs. Res. 76 5770 Housley, R.M.; E.H. Cirlin and R.W. Grant (1973) Characteriza- tion of fines from the Apollo 16 site. Lunar Science IV (ed. J.W. Chamberlain and C. Watkins) b. 381, Lunar Science Institute LSPET (Lunar Sample Preliminary Examination Team) (1971) Pre- liminary exhmination of lunar samples from Apollo 14. Science 173 681 (1972) The Apollo 15 lunar samples A preliminary discussion. Science 175 363 Marcus, Allen H. (1969) Speculations on mass loss by meteoroid impact and formation of the planets. Icarus 11 76-87 Matson, Dennis L. (1971a) Infrared observations of asteroids. NASA SP-267, p. 45 (1971b) Infrared emission from asteroids at waveleneths of 8.5, 10.5 and 11.6 microns. doctoral diss., Califor. Inst. of Technology v-17I44

Morrison, D. (1973) Determination of radii of satellites and asteroids from radiometry and photometry. Icarus 19 1 NASA SP-267 (1971) Physical studies of minor planets (T. Gehrels, ed.), Proc. of 12th Colloq. Intl. Astron. Union, Tucson Pettengill, G.H.s I.I. Shapirol M.E. Ash; R.P. Ingalls; L.P. Rainville: W.B. Smith and M.L. Stone (1969) Radar observations of Icarus. Icarus 10 432 Piotrowski, S. (1953) The collision of asteroids. Acta Astron. Ser. 5 115-138 Pollack, J.B. et al (1972) Mariner 9 television observations of Phobos and Deimos. Icarus 1 394 Shoemaker, Eugene M. (1963) Impact mechanisms at Meteor Crater, Arizona in The Moon, Meteorites and Comets (Middlehurst and Kuiper, ed.) University of Chicago Press van Houten, C.J.s I. van Houten-Groeneveld; P. Herget and T. Gehrels (1970) The Palomar-Leiden survey of faint minor planets. Astron. Astrophys. Suppl. Ser. 2 339-448

Veverka, J. (1971) The physical meaning of phase coefficients. NASA SP-267, p. 79. 145 Chapter VI Meteorites, Mineralorg and Petrology a) Introduction This chapter consists of a review of the physical nature of meteoritic material. While this information may be quite redundant for many readers, there nevertheless exists a substantial segment of the astronomical and planetary science community who have only the barest acquaintance with meteoritics but who would also find this thesis of interest. This chapter is aimed at providing these individuals with a single con- densed body of information on the meteoritic materials which are likely to affect the spectral reflectivity of these materials. This chapter should provide enough background on this subject to make subsequent discussion meaningful. Those individuals who are conversant with meteoritics should either. briefly skim or skip altogether this chapter. References are provided for more complete information on any particular subject. Meteorites are defined as any naturally occurring solid body which reaches the earth's surface on its own accord from beyond the earth's atmosphere. This definition will include such diverse materials as the metal and stony meteorites, cometary debris (dust), 'ices' (solid solutions of water, methane, ammonia and hydrogen sulfide) and possibly glasses

(tektites) but excludes man-made objects such as decayed satellites or returned lunar samples. However for purposes of 146 VI-2 this study, only the more restricted group of stony or metallic meteorites will be considered. These have been divided into three general classification groups on the basis of gross compositional characteristics. Irons (Siderites) are composed of nickel-iron, iron sulfide and occasional silicate inclusions. This group is subdivided on the basis of nickel content, metal crystal size and structure and comprises about 6% of all known falls.

Stony-Irons (Siderlites) are composed of approximately equal portions of metal and silicates mixed together. This group is subdivided on the basis of silicate composition and textural relationships, and comprises about 2% of all known falls. Stones (Aerolites) are composed primarily of silicates with occasional NiFe minerals and troilite. This group is subdivided into two major classes (chondrites and achondrites) on the basis of the presence or absence of chondrules (spherical silicate inclusions). These are. further subdivided on the basis of metal content, silicate composition and textural relationships. This group comprises 92% of all known falls. b) Meteoritic minerals Before proceeding with a description of the individual meteorite types it is necessary to define the suite of meteoritic minerals and to acquaint the reader with the mineralogical terminology to some extent so that subsequent definitions will 147 VI-3 -

be more meaningful. A current, more detailed -study can be

found in Mason (1972). A mineral is classically defined as a naturally occurring, inorganic solid substance with a definite set of chemical and physical properties. Thus each mineral is characterized by a certain proportion of elements arranged in a specific crystal structure. Many of the meteoritic minerals are found in terrestrial rocks but a number are unknown in earth rocks and as such tend to give evidence of the conditions under which the. meteorite parent bodies were formed. Tables VI-1 and VI-2 (Tables I and II from Mason (1972)) list these minerals. The meteoritic minerals can be divided into silicates and non-silicates (primarily metals and sulfides). The non- silicates are composed of iron, nickel, phosphorus, cobalt, carbon and sulfur in various combinations. The major non- silicates ares Kamacite ( 4C-NiFe) is a nickel-iron alloy with a fairly constant composition of about 5.5% nickel which crystallizes in a body centered cubic lattice. Taenite ( K -NiFe) is a nickel-iron alloy of variable composition. 27 to 65% nickel, which crystallizes into a face-centered cubic structure. Troilite (FeS) is present in practically all meteorites in about equal amounts (6%).

Schreibersite (Fe, Ni) 3P) is a universal accessory mineral in iron meteorites and is common in small amounts in many Table VI-1 (Mason, 1972) ma Meteoritic Minerals (as of 1962)

Calcite CaCO3 Accessory in Ccl and Ccll Ccl, II, III = carbonaccous chondrites, Types I,li, 111; Dolomite CaMg(CO3), Accessory in Ccl Ce = enstatite chondrites; Ae = enstatite achondrites Quartz SiO2 Accessory in some eucrites Name Formula Occurrence and Ce SiO Accessory in some stones, Kamacite a-(Fe,Ni) In irons, stony-irons, and Tridymite 2 stony-irons, and irons most chondrites Cristobalite Si0 Accessory, mainly in Cc Taenite 7-(Fe,Ni) As for kamacite 2 Copper Cu Common as an accessory Ilmenite FeTiO3 Accessoryin manystones andstony-irons Diamond C Present in ureilites Graphite C Commonaccessory in Spinel ' MgAI204 Accessorymainly in Cc irons and some stones Magnetite Fe3 04 Accessoryin Cc Chromite FeCr Accessoryin most Sulfur S Accessory in Ccl 204 P Accessoryin irons, stony- meteorites Schreibersite (Fe,Ni)3 C Irons, and somechondrites Chlorapatite as(PO4)3 CI Accessoryin many Cohenite (Fe,Ni)sC Accessoryin many irons meteorites andin Ce Whitlockite CraMgIH(PO4)7 Accessory in many *Osbornite TiN Accessory in Ae meteorites *Farringtonite in some Troilite FeS Present in most meteorites Mg3 (P0 4 )2 Accessory *Oldhamite CaS Accessoryin Ce and Ae pallasites Gypsum C in Ccl and Ccll Pentlandite (Fe,Ni), Ss Accessory, mainly in Cell :aSO4.21120 Accessory and Ccill Epsomite igS04 .71120 Prominentin Ccl Bloedite Ae, and Mg(SO) .41H20Accessory in Ivuna(Ccl) *Daubreelite FeCr, S4 Accessory in Ce, ( 4 2 many irons Olivine Mg,Fe)2SiO4 Commonin stonesand stony-irons Chalcopyrite CuFcS2 Accessory in Karoonda (Ccill) Orthopyroxene (Mg,Fe)SiO3 Commonin stonesand stony-irons Pyrite FeS2 Accessory in Karoonda (Cclll) Clinopyroxene (Ca,Mg,Fe)SiO3 Commonin stonesand stony-irons Sphalerite (Zn,Fe)S Accessoryin Ce and some irons Plagioclase (Na,Ca,XAI,Si)4 08 Commonin stonesand stony-irons '*Lawrencite (Fe,Ni)C12 Accessory in some meteorites Serpentine (Mg,Fe)Si 010(OH)sMatrix of Ccland Cell (or chlorite). Magnesite (Mg,Fe)COz Accessory in Ccil

I as I VI-3b ;49 Table VI- 2-

Minerals discovered in meteorites since 1962 (an asterisk indicates those not known to occur in terrestrial rocks). Ccl, II, I1 = carbonaceous chondrites, Types 1, II, III; Ce = enstatite chondrites; Ae = enstatite achondrites Name Formula Occurrence Awaruite Ni Fe Accessory in Odessa iron and Allende (CcIll) *Lonsdaleite C Rare, in ureilites Chaoite C Rare, in ureilites

*Haxonite Fe23 C Accessory in many irons *Barringerite (Fe,Ni) 2P Accessory in Ollague pallasite *Perryite (Ni,Fe)s (Si,P)2 Accessory in Ce and Ae *Carlsbergite CrN Accessory in many irons *Sinoite Si N2 O0 Rare in some Ce Pyrrhotite Fe,.xS Accessory in Ccl Mackinawite FeS, .x Common as an accessory Heazlewoodite Ni 3 S2 Accessory in Odessa iron *Niningerite (Mg,Fe)S Accessory in some Ce Alabandite' (Mn,Fe)S Accessory in some Ce and Ae *Brezinaite Cr 3 S4 Accessory in Tucson iron Djerfisherite K3 CuFe,2SI4 Accessory in some Ce, Ae, and Toluca iron 2 *Gentnerite Cus Fe 3Crj, Sas Accessory in Odessa iron Rutile TiO2 Rare accessory Hercynite (FeMg)Al 2O4 Accessory in some CcMII Accessory in Hibonite CaAl, 2 019 some CclI and CcllI Perovskite CaTi03 Accessory in Cclll Accessory in Murclison Whewellite CaC20 4 .H2 O (Ccll) Accessory in some stony-irons *Stanfieldite Ca4 (Mg,Fe)s (P0 4 )6 *Brianite CaNa2 Mg(P0 2) Accessory in some irons Rare accessory in some irons Graftonite (Fe,Mn) 3 (PO4 )2 *Panethite (Ca,Na) 2 (Mg,Fe) 2 (P0 4 )2 Accessory in Dayton iron Sarcopside (Fe,Mn)3 (P0 4 )2 Rare accessory in some irons *Ringwoodite (Mg,Fe) 2 Si04 In Coorara and Tenham chondrites In Coorara and Tenham chondrites *Majorite Mg30(gSi)Si 3 012 Wollastonite CaSiO3 Accessory in Allende (CcIII) *Ureyite NaCrSi 2 0 6 Rare accessory in some irons Rare accessory in a few irons Potash feldspar (K,Na)AISi3 Os Nepheline NaAISiO4 Accessory in a few chondrites Accessory in some CcIII Sodalite Na8 A16Si6 0 24 Cl2 *Merrihueite (K,Na) 2 Fes Sil2 03o Rare accessory in Mezo-Madaras chondrite *Roedderite (K,Na) 2 Mgs Si 12 O30 Rare accessory in Ce and irons iron *Yagiite (K,Na) 2 (Mg,Al)s (Si,AI)1,2 0 30 Rae accessory in Colomera Richterite Na2 CaMgs SisO22 F2 Rare accessory in a few irons, and in Abee (Ce) In chondrules in CclIl Melilite Ca2 (Mg,Al)XSi,Al)z20 Zircon ZrSiO4 Rare accessory (CclII) Grossular Ca3 A12Si3 012 Accessory in Allende in Allende (Cclll) Andradite Ca3 Fe2 Si3 O,2 Accessory Rhbnite CaMg 2 TiAl2 SiO Accessory in Allende (Ccill) 150 VI-4

chondrites.

Graphite (C) is a common accessory mineral in iron meteorites in the form of grains or nodules or as a diffuse amorphous form in some meteorites. Some of the carbon in a few meteorites has been converted to the diamond structure either by static pressure in the parent body or by shock.

The major silicate minerals ares

Olivine (Mg,Fe)2Si 4 ) is an essential component of most meteorite groups. The composition is usually of the magnesium

rich variety, from 15 to 30 mole percent Fe2Si04 (Fa), although olivines ranging to 65% are found in the achondrites.

Orthopvroxene ((Mg,Fe)SiO 3 ) is the most common silicate mineral in meteorites after olivine. Meteoritic orthopyroxenes are

divided according to the molar percent of the FeSi0 3 (Fs) components enstatite (less than 10%), bronzite (10-20%) and

hypersthene (more than 20%). Care should be taken with these definitions since they differ from the definitions commonly used by mineralogists. Orthopyroxene is crystallized into an orthorhombic structure. Clinopyroxene ((Ca,?g,Fe)SiO3) is the monoclinic form of pyroxene and is generally found as the calcium poor (pigeonite)

or calcium rich (augite) forms although the iron free form

(diopside) is found in certain types of meteorites.

Plagioclase ((Na,Ca)(Al.,Si)Si 2 08 ) is a common constituent of most.chondrites (5-10%) and some achondrites and is generally VI-5 151 the sodium-rich member of the plagioclase series (oligoclase).

The meteorite Serra de Mage consists mainly of plagioclase.

SerDentine ((Mg,Fe) 6 Si 1 40 1 0 (OH) 8 ) is a sheet silicate mineral, an alteration product (hydrated) of the ferromagnesium minerals olivine or pyroxene which makes up the groundmass of the type I and II carbonaceous chondrites. The above list comprises the major meteoritic minerals. A large number of accessory or rare minerals also exist but the reader is referred to Mason (1962, 1972) for a discussion of these. c) Meteorite Tynes: Irons The irons can be grouped into a sequence based on structure which is clearly related to the proportion of nickel in the metal. Figure VI=l shows the frequency of nickel compositions in a number of irons. Hexahedrites are so named because they are generally made up of large crystals of kamacite with a cubical habit (cube = hexahedron). The nickel content in the metal phase in these objects ranges from 4 to 6% with the lower values (less than 5%) probably in error due to older less accurate analytical methods. The peak in the nickel distribution (Figure VI-1) at about 5.5% is due to this type of meteorite. A polished, etched surface of a specimen of this type exhibits an array of fine parallel lines (Neumann lines) which are the result of crystal twinning on a trapezohedral face probably due to Oct. ,Nicke--r-ch Atacites

0 4 8 12 16 20 24 28 32 36 40 60 % Ni in metal

The frecqicwcy of distribution of nickej content In analyses of Iron meteorites (after Yavnel, 1958). N - numlwr of analyses.

_ I II ------53 VI-6

mechanical shock deformation of the object at relatively low temperatures. With increasing nickel content (approaching 6%), the hexahedrites grade into the coarsest octahedrites. Octahedrites show an orientation of kamacite and taenite bands parallel to octahedral planes forming the Widmastatten pattern when a surface is polished and etched. The range of nickel contents is from 6 to 14 percent. This group of meteorites is subdivided on the basis of the width of the kamacite bands. The Rose-Tschermak-Brezina classification recognizes five types, Type Band Width in mm Coarsest Octahedrites 2.5 mm Coarse Octahedrites 1.5-2.5 mm Medium Octahedrites 0.5-1.5 mm Fine Octahedrites 0.2-0.5 mm Finest Octahedrites 0.2 mm The nickel content varies systematically with band sizes Coarser (6-8%), Medium (7-9%) and Finer (8-14%), The coarsest octahedrites grade down into the hexahedrites and the finest octahedrites grade up into the nickel-rich ataxites. Nickel-Rich Ataxites occur as the nickel content of the metal increases, narrowing the bands of kamacite, until at between 12 and 14% Ni, they become extremely narrow and discontinuous, at which point the Widmanstatten structure no longer exists. The mass of these meteorites consists of plessite (very fine grained eutectoid intergrowths of kamacite and taenite) except in the very nickel rich specimens (> 25%) which. consist of VI-754

taenite with small inclusions of kamacite generally arranged in a trigonal pattern. Additionally two other rare types of irons have been recognized. Nickel-Poor Ataxites (which have been included with the hexahedrites on the basis of nickel content) have a very fine grained structure and are thought to have been produced from hexahedrites by thermal metamorphism. Sorotiiti which is represented by a single specimen and consists of approximately equal amounts of nickel-iron and troilite. Thus the irons can be seen as a continuum of compositions with most specimens found at the approximate compositions for the solidus-liquidus relation in the nickel-iron system. In addition to the omnipresent nickel-iron minerals, there is also present troilite in varying amounts as scattered nodules in the coarsest octahedrites decreasing in the finer octahedrites and nearly absent in the nickel-rich ataxites and hexahedrites. Additionally occasional silicate inclusions are found in the irons including olivine and pyroxene masses, feldspar crystals and even reasonably well preserved chondrule structures. A detailed discussion of the silicate inclusions in the iron meteorites and their implications concerning the origins of these bodies can be found in Bunch et. al. (1970). Work on certain inclusions in the Campo del Cielo iron suggests that the overall distribution of silicate inclusions in the VI-8 meteorite parent bodies may be much higher than is generally suggested by the irons (Lewis, personal communication 1972). Lovering et. al. (1957), Wasson (1967) and Wasson and Kimber- lin (1967) have discussed a chemical classification of the iron meteorites based on the "quantization" of gallium-germanium concentration in these meteorites. Such groups seem to be genetically related and the different groups suggest different parent bodies. However, such subtle differences in the compositions of these meteorites is unlikely to create differences in the spectra of these materials. d) Meteorite Typess Stony-Irons

This class of meteorite is the least common group of the three major types in. terms of observed falls to the earth's surface. It can be divided into two major groups and two minor groups, which, since members of this type are composed of approximately equal portions of metal and silicate phases, are defined on the basis of structural relationships between these phases and on the basis of the composition and structure of the silicate phase. The stony-irons seemingly form a distinct family with unverified but highly suggestive affinities to several other types of meteorites.

Pallasites (olivine stony-irons) are composed of a metal matrix with olivine inclusions. There can be significant variations in the ratios of the two phases and at least one pallasite

(Brenham) has fragments which range from approximately equal VI-9 156 amounts to completely metallic specimens with a few that show patches of each in the same fragment. The olivine inclusions range from complete or subrounded crystals to angular or brecciated fragments of crystals. Yavnei (1958) divided pallasites into two subgroups on a chemical basis, the first with olivine compositions near Fa1 3 (Fal 0 -Fal 6 ), 55% metal with a nickel content of 10%; and the second with about Fal9 olivine, 30-35% metal with 15% nickel in the metal. The metal phase generally shows Widmanstatten patterns. Mesosiderites (pyroxene-plagioclase stony-irons) consists of a mixture, perhaps mechanical, of metal and silicate neither of which forms a definite matrix. The metal phase consists of irregular blebs which show indications of having formed in situ after the observed physical relationships were established. The nickel content of the metal phase averages 8% (7.4-8.8%) for ten of these meteorites which were described by Powell (1969), and show some Widmanstatten structure. The silicate phase consists of a breccia of orthopyroxene (Fs 2 0 -Fs 4 0, mostly near Fs32) and calcic-plagioclase (An80-An 9 8 ) with minor pigeonite (Fs 3 5Fn 6 0Wo5 -Fs 5 2En 3 9Wo9 ) and olivine (Fa 20 -Fa 4 0 ) in a very heterogeneous mineral distribution with significant textural variation from point to point in the same specimen. A detailed study of the petrology and chemistry of the silicate and metal phases of the mesosiderites can be found in Powell (1969, 1971). The two rarer stony-iron groups, each represented by a VI-lO 157 single specimen are, Siderophyre (bronzite-tridymite stony-iron) consists of a metal groundmass with granular inclusions of orthopyroxene with minor tridymite, the two phases being present in approximately equal amounts. The metal contains about 10% nickel and shows

Widmanstatten structure, and the orthopyroxene is about Fs 2 0 . The single representative of this class is the Steinbach meteorite. Lodranite (olivine-bronzite stony-iron) consists of a friable aggregate of granular olivine (Fal3 ), orthopyroxene (Fs?7) and nickel-iron (9% Ni), with approximately equal amounts of these three phases. The single representative of this subgroup is Lodran meteorite. e) Meteorite Tyress Stones - Chondrites The chondrites are so named because of the presence of small spherical silicate inclusions, chondrules, in these meteorites. In terms of observed falls, the condrites are by far the most common type of meteorite, constituting more than 85% of such falls. Whether this abundance is representative of the population of the small solid bodies of the solar system or simply the reflection of a few favourably located (such as in an earth crossing orbit) source bodies has not been determined, but the very fact that these are the most commonly available meteorites has allowed investigators to define the group characteristics and the variations in these characteristics with some degree of certainty. In this work VI-11 158

a significant amount of attention will be devoted to this

group. These meteorites have been divided into five subgroups

based on the relative abundance and composition of the mineral phases present: enstatite, olivine-bronzite, olivine-hyper- stene, amphoterites, and carbonaceous chondrites. With respect

to the bulk composition, in all but one subgroup of the ordinary chondrites (L-type) the proportions of the nonvolatile

elements are very nearly constant. Especially in terms of the total amount of iron present, these four subgroups behave at least superficially, as if they were produced from the same starting material which was then reduced (or oxidized) to varying degrees. The degree of oxidation or reduction is reflected in the ratio of FeO to total iron as is shown on figure VI-2, where the four groups lie along a straight line. The most characteristic, although not universal, feature of the chondrites is of course, the presence of chondrules, spherical silicate bodies, on the order of Imm in diameter, which are usually composed of crystals of olivine and/or orthopyroxene. The orthopyroxene chondrules are generally made up of radial crystals which in some cases may be so small as to appear amorphous. The olivine chondrules consist of one or several crystals or fragments of one crystal. A detailed description of the structures in chondrules can be found in Mason (1962) and Wood (1963). Compositionally, the chondrules are often very similar to the appropriate crystals VI-1la I59

Figure VI-2

* enstatite chondrites a ohvine-bronzite chondrites *+ olivine-hypersthene chondrites o , olivine-pigeonite chondrites 0 carbonaceous chondrates

rio-1 A

820 is

10 ...... 0 1* 0 * 0 ".0 4i 5 !"

0 5 10 15 20 25 30 Weight per cent oxidized iron

Helationhip ltveen oxidized iron and iron as metal and sulfile in analyses of oberved falls, illustrating the separation into distinct subgroups and the variation within the subgroups (.Mason, 1962). VI-12 6a

in the groundmass of the meteorite. Another very common structural feature in many chondrites are dark veins shot through the bulk of the specimen, which have been variously interpreted as frictional heating, injec- tion of melt, troilite concentration by hot sulfur-containing gases (Anders and Goles, 1961), but the most consistent explanation (Fredricksson et al, 19631 Heymann, 1966) based on experimental evidence, suggests that shocks between 150 and 800 kilobars will microfracture the matrix of the meteorite, darkening it without changing the composition. At low shocks the veins appear and at the higher shocks the whole meteorite is blackened to produce the so-called black chondrites which will be discussed below. Van Schmus and Wood (1967) described a chemical-petrological classification system for the chondrites in which they expanded the basic chemical (mineralogic) subgroups on the basis of varying petrologic (implied metamorphic) properties, such as the degree of inhomogeneity of silicate mineral compositions, evidence of recrystallization of the structure and the relative abundance of the volatile phases. The petrologic grade can be defined as the degree to which the mineral assemblage approaches equilibrium. They defined six petrologic grades and defined a set or criteria to determine the grade which is given as Table VI-3. Figure VI-3 shows the population density of each chemical-petrologic type. These types will be rt

Table VI-3

Summary of Characteristicsof Chemical-PetrologicalChondrite Subtypes VI-13

Figure VI-3 Population of the Van Schmus and Wood Chemical- Petrological Subtypes of the Chondrites.

Petrologic type

1 2 3 4 5 6

E El E2 E3 E4 E5 E6 - - 1 4 2 6 C C1 C2 C3 C4 C5 C6 4 16 8 2 -

Chemical H H1 H2 H3 H4 H5 116 group - - 7 35 74 44

L L1 L2 L3 L4 L5 L6 -- 9 18 43 152

LL LL1 LL2 LL3 LL4 LL5 LL6 - - 4 8 7 21

Number of examples of each meteorite type now Imknown is given in its box.

(Van Schmus and Wood, -1967)

considered in our description of the individual subgroups of the chondrites. Enstatite Chondrites (E-type) are composed principally of enstatite and/or clinoenstatite (40-60 wt%), kamacite (20-

28 wt%), troilite (7-15wt%) and some plagioclase (5-10 wt%). They characteristically exhibit a very high degree of reduction which is reflected by a pyroxene phase uhich is very nearly pure MgSiO 3, containing as an upper limit 1 mole percent

FeSiO 3 and in most cases probably less than 0.1%. Additionally the presence of elemental silicon in the metal phase plus the presence of nitrides of titanium and silicon indicate a very high degree of reduction. The metal phase averages about 6% nickel and may contain as much as 3% elemental silicon in VI-14

solid solution. The plagioclase is oligoclase within a restricted compositional range between An 1 4 and An2 0 . In general appearance these meteorites are very dark in color and very dense with blebs of metal, about imm in diameter

scattered more or less uniformly throughout the groundmass.

The general blackness of these specimens may be due to the distribution of an opaque phase such as carbon of sulfides throughout the groundmass or it may be due to a general. shock blackening as described above. There are sixteen known meteorites of this type and for detailed information on several aspects see the following workst description of class and individual specimens, Mason (1966)t mineralogical and chemical relationship, Keil (1968)1 composition and structure of the feldspars, Van Schmus and Ribbe (1968); composition of the pyroxene phase, Binns (1970): and composition of the olivine phase, Mason (1963, 1967). Olivine-Bronzite Chondrites (High-Iron or H-type) are so named because the pyroxene present is bronzite and are also called, along with the hypersthene and amphoteritic chondrites, the ordinary chondrites. These meteorites are composed of olivine

(25-40 wt%), orthopyroxene (20-35 wt%), nickel-iron (16-21 wt%) with plagioclase (5-10 wt%) and troilite ( 6 wt%). Since this class is quite common, enough statistics have been compiled to define the range of variations of the characteristics of each mineral phase and each will be treated separately in some detail below. VI-15 164

The olivine phase constitutes between 25 and 40 weight percent of these meteorites and according to Mason (1963, 1967) and Keil and Fredricksson (1964) the average olivine composition from individual specimens of this type ranges between Fal 4 -Fa 2 1 (Figure VI-) with most in the range Fal8.19. However Dodd and Van Schmus (1965) and Dodd et. al. (1967) pointed out that in certain of the ordinary chondrites, especially those which show the smallest degree of recrystallization, there is a significant variation in the composition of the olivine (and pyroxene) from grain to grain in specimens of unmetamorphosed and slightly metamorphosed meteorites (see figure VI-5a, 5b), as opposed to the relatively uniform olivine (and pyroxene) composition in a highly recrystallized chondrite (see figure VI-Sc). In the least recrystallized case, these meteorites have been termed the unequilibrated ordinary chondrites and there exists a continuum between these two extremes. This is the basis of the Van Schmus and Wood chemical-petrologic classification, where the olivine (and pyroxene) show significant variation in a low grade (H4) specimen while being quite uniform in the high grade (H6) specimen.

The pyroxene phase comprises between 20 and 35 weight percent of these meteorites and averages between Fsl5-18 (figure VI-6) and exhibits the same degree of disequilibrium in certain meteorites as does the olivine described above

(figure VI-5). A detailed description of the structural VI-15a

Figure VI-4 (Keil and Fredricksson, 1964)

H- GROUP L-GROUP LL -GROUP

15 17 19 21 23 25 27 29

MOL PER CENT Fe + Mg Frequency distribution of the mole per cent ratios Fe/(Fe + Mg) in olivine for the chondrites

Figure VI-5a (Dodd et.al., 1967) OLIvi,.ES PYROXEN'ES

20 Krymka A (103) 20 Kryr ka A (85)

p.h Mfaa L , .06a;, 1 - -

0 10 20 30 40 0 10 20 30 40 Wt. % Fe Wt. % Fe Fig. 1. Iron contents of olivines and low-calcium pyroxenes in four unmetamor- pbosed or slightly metamorphosed ordinary chondrites. The number of mseasure- ments for each sample is in brackets following the name. VI-15b Figure VI-5b (Dodd et.al., 1967) OLIVINES PYROXENES

40 Parnallee (94) 40. Pornollee (104)

20 2c )IA.!I Baorratto (i06) Barratta (79)

i 2C

r"m 10 20 30 40 10 20 Wt. % Fe Wt. % Fe Iron contents of olivines and low-calcium pyroxenes in two moderately mnequilibrated ordinary chondrites. The number of measurements per sample is in brackets following the name.

OLIVINES Ficmue V1-Ec PYROXENES . ...I ... II

Tennasilm (94) Tevinasilm (114) 1 60 40 40

2( 201

0 I ______I' 60C 6 SModoc (90) Modoc (46)

40 4C

I 0 J &P) - 0 -0 30 U &%0 Iron contents of olivines and low-calcium pyroxenes in a nearly equilibrated ehandrite (Tennasiln)- and a completely equilibrated chondrite (Modoc). The number of measurements per sample is in brackets following the name. VI-15c 167

Figure VI-6

(Keil and Fredricksson, 1964)

24 H-GROUP L-GROUP 20

12 I LL-GROUP 8

13 15 17 19 2'1 23 25 Fe MOL PER CENT Fe + Mg

Frequency distribution of the mole per cent ratios Fe/(Fe + Mg) in rhombic pyroxene for the chondrites VI-16

state and composition of the pyroxenes in these meteorites can be found in Keil and Fredricksson (1964) and Binns (1970).. The metal phase comprises about 17 weight percent

(14-19 wt%) of the mass of these meteorites and contains about 9 mole percent nickel (Keil and Fredericksson,.1964). The metal is found in small grains (10-1000 microns) which are generally either kamacite or taenite with a few (10-20%) grains which contain both minerals. The taenite grains are generally zoned with 45-55% Ni at the.edges and 25-35% Ni at the center of the grains (Wood, 1967; Taylor and Heymann, 1971). The feldspar phase of the H-type chondrites occupies a

very small compositional range at about Ab8 2An1 20r 6 with about a + 1 mole percent range in An and Or concentration. Structurally the feldspar crystals are very small, less than 10 microns, except in those specimens which exhibit extensive recrystallization. In general outward appearance, meteorites of this class are light gray in color (with the obvious exception of the black chondrites), with a signiflcant amount of metal visible. In many specimens, especially those of the lower metamorphic grades, large (1 mm) chondrules are evident. The lower grade specimens also show some significant inhomogeneity with a mottled appearance of different colors(shades of gray) or 7different grain size which makes many of them resemble breccias. VI-17 A6.9

Black veins and troilite crystals are also found in many of the ordinary chondrites. Olivine-Hypersthene Chondrites (Hypersthene or Low Iron or L-type) are characterized by the presence of a 'hypersthene' pyroxene which ranges from Fsl8 to Fs22 and a significantly lower amount of metal (7?% vs 17%). Other minerals present in quantity are olivine, plagioclase, and troilite plus a small amount of diopside (calcium rich pyroxene). Statistically these seem slightly more abundant than the other chondrites.

Olivines 35-60 weight percent, Fa 2 1 -Fa 2 5 with the same type of variation seen in the H-type.

Pyroxenes 25-35 weight percent, Fsl8-Fs2 2

Feldspar: about 10 weight percent, Ab84An 100r6 with +1 mole percent in the An and Or contents. Metals about 7 weight percent (4.4-11.7), 14-15 mole percent nickel, structure of kamacite-taenite similar to that of the H-type chondrites. In outward appearance, the L-type chondrites are very similar to the H-type except for the significant difference in the amount of metal. Amphoterites (Soko-Banja type or Low Iron-Low Metal or LL-type) were recognized as a discrete group from the L-type chondrites by Keil and Fredricksson (1964) on the basis of significantly lower metal content (about 2*5%) with an iron-rich pyroxene

Fe 2 3 -Fs2 4 .

Olivines 35-60 weight percent, Fa2 6-Fa29

Pyroxenes 25-35 weight percent, Fs2 2 -Fs2 5 VI-18 170

Feldspar: about 10 weight percent, Ab 86An00Or4 with +1 mole percent An but Or from 0.9 to 5.2 mole percent. Metals 2.5 weight percent (1.7-3.2), 28 mole percent nickel Additionally these meteorites contain about 6 weight percent troilite plus a small amount (about 1%)of diopside. The ordinary chondrites (H, L and LL) can thus be divided into three distinct groups characterised by metal content, olivine and pyroxene composition. Black Chondrites are ordinary chondrites as far as the metal and olivine and pyroxene compositions are concerned, primarily L-type, but are exceptionally black in color, having a visible albedo of between 3 and 5% reflectance. It was shown by Fredricksson et. al. (1963) and Heymann (1966) that there was no systematic chemical or mineralogical differences between the black veins in certain chondrites and the rest of the mass, and that the veins consist of extremely fine grained crystal fragments. Perhaps, this microbreccia acts as a light sink, allowing photons to enter but preventing easy reflection out of the material. It was also shown that the effect could be produced by shocking specimens at between 150 and 800 kilobars for a fraction of a microsecond, the lower value producing veins identical to the natural ones, and the higher value producing the black chondrites with a gradation between these two effects.

Heymann inferred from the decrease in .gas retention ages of the progressively darker ordinary chondrites, that the VI-19 darker gray (to black) a chondrite is, the more shocked it is, and that a continuum exists based on shock history between the white and -lack ordinary chondrites. On the basis of bulk composition and mineral phase relations of the members of this series, this concept is reasonable. However, for many purposes in this paper, the black chondrites will be considered as a separate subdivision of the ordinary chondrites. Carbonaceous Chondrites (C-type) are characterized by the presence of carbonaceous material in a form other than free carbon (graphite or diamond), by the abundance of water in hydrated mineral phases and by a characteristically high Mg/Si ratio (; 1.0). Physically, these meteorites consist of a very fine grained, dark gray to black groundmass with lighter colored inclusions and/or chondrules in varying abundances.

Carbonaceous chondrites are also characterized by a very high degree of disequilibrium in the mineral assemblage,for example, with the coexistence of elemental sulfur, sulfides and sul- fates. Additionally the olivine and pyroxene phases are quite inhomogeneous and have a wide range of compositions as is shown in figure VI-7. Wiik (1956) proposed a chemical classification of the carbonaceous chondrites on the basis of variations in certain constituents: OLMNE% PYROXENE: Fiqure VI-7

(Wood, 1967b) i HARIPURA

I s I.u eaiw6.a . La.S. OLIVINE: PYROXEF

COLDBOKKEVELD

Ti NOGOYA LI.. Palo so 0 8 A a go as

oli I I• MIGHEI S

POLLEN

0 20 40 60 0 10 MOLE% F.*S1O04 %FeSiO S ...... a. . L. 13 ' Histograms of five moro C2 chondrito (and one C3)t 20 40 0o MOLE% % FeS4 Fe2 SiO4 8 Histogramns of mineral collmposition in fivo of tho C2 chondrites stuLdiel. Each intorval of comnposition is 0)5 m11o0'. fayalito (I1ceZ804) or rtfrroilito (0ASio 3) wide. "Por couentof points" rCepresIntAltIby cach vortical bar is calculated relative to all points analyzed in each chondrito, both olivino and pyroxone. VI-20 173

Si02 Mr0 C Hg S Type I 22.56 15*21 3.54 20.08 6.03 Type II 27.57 19.18 2.46 13.35 3.16 Type III 33.58 23.74 0.46 0.99 2.21 Van Schmus and Wood (1967) classified the carbonaceous chondrites on the basis of metamorphic feature (see Table VI-3) which agrees with several exceptions with the -rouping of Wiik. Van Schmus (1969) divided Wiik's Type III into two subgroups based on the structure of the condrules. These class- ifications and the meteorites themselves are described in a review paper by Mason (1971). Type I (Cl) are low density (2.20-2.42), friable azrre- gates of poorly crystallized hydrated magnesium-iron silicates with no chondrules but with some small inclusions. The groundmass is in general a clay mineral, montmorillanite, and serpentine such as have been identified in the Orgueil (Type I) chondrite by Bass (1971). These are soft, porous meteorites generally and resemble nothing so much as dried mud. Struc- turally they are brecciated and contain abundant opaque minerals, carbonates, sulfides, and hydrated iron oxides with some phos- phates and sulfur. Mason lists five members of this type. Type II (C2) are medium density (2.57-2.92), more solid with variable amounts of chondrules and grains of iron-poor olivines and pyroxenes in a fine grained groundmass composed of chlorite or serpentine and containing very little free metal VI-21

except in two exceptions (Al Pais, 1-2%1 Rennazo, 12%). The chondrules are generally light in color and consist of iron-

poor olivines and enstatite or clinoenstatite. Mason lists 16 members of this group.

Type III (C3) are higher density (3.40-3.78) with a

groundmass consisting of fine-grained iron-rich olivine

(Fa4 0 -5 0 ) and a metal content ranging between 0 and 6 percent with two exceptions (Coolidge and Kainsaz). This type has more

abundant chondrules and is divided on the basis of chondrule structure. Type III-V (C3V), Vigarano subtype, have large Aspongy' chondrules in an abundant fine grained black groundmass.

This subtype has systematically less total iron and more Ca and

Al. Mason lists 9 members of this subgroup. Type I-0 (C30), Ornans subtype, have more chondrules which are dense and small. Mason lists 6 members of this subgroup. f) Meteorite tvDes - Achondrites The achondrites cannot be considered as a coherent group in the sense of having a single unifying characteristic such as the irons (composition) or the chondrites (chondrules) but are a quite heterogeneous group of stones considered together because they lack the characteristic chondrules of the chondritic stones. They commonly exhibit a much more coarsely crystalline structure than do the chondrites and tend to show, in terms of composition and mineralogic character, appearances quite analogous to certain terrestrial igneous rocks. Additionally, the metal phase 175 VI-22

is almost completely absent. The achondrites have been subdivided into two groups on the basis of calcium content: calcium-poor (0-3% CaO) including aubrites (enstatite), diogenites (hypersthene), chassignites (olivine) and ureilites (olivine-pigeonite); and calcium-rich (5-25% CaO) including eucrites (pyroxene-plagioclase), howardites (pyroxene-plagioclase), angrites (augite) and nakhlites(diopside-olivine). This whole class tends to be rare, both because members are rare in falls and because they are difficult to distinguish from terrestrial rocks making finds uncommon. Eucrites (pyroxene-plagioclase monomict breccias or basaltic achondrites) have been studied in detail by Duke and Silver (1967) who distinguished these objects from howardites on the basis that the fragments of the breccia are of a single composition('monomict) rather than of several distinct basaltic-type compositions(polymict). Mineralogically, the eucrites contain approximately equal amounts of pyroxene and plagioclase with traces of quartz and accessory nickel-iron, troilite and ilmenite. The pyroxenes are generally the olinopyroxene, pigeonite (low calcium, monoclinic pyroxene) with some sub-calcic augite with a composition of 10 mole percent Cs (CaSiQ3 ) and about 60 mole percent (48-70) Fs (FeSiO3 ). The plagioclase varies in composition from An 80 to An9 5. The nickel content in the metal phase is very low, ranging from 0.1 to 1 percent. Petrologically, the -eucrites are composed of gray to black lithic fragments of essentially identical mineralogic composition contained in a dark VI-23 ;7C gray groundmass made up of a mixture of plagioclase and pyroxene. The shape of the lithic fragments varies from angular to rather well rounded and in size from fine to quite coarse (~ 1 cm). Duke and Silver list 26 eucrites making them the most common type of achondrites.

Howardites (pyroxene-plagioclase polymict breccias or basaltic achondrites) are physical mixtures composed of lithic fragments of several distinct compositions (polymict). Mineralogically, these meteorites are composed of pyroxene and plagioclase, in a proportion of about 3s1, with traces of olivine and accessory nickel-iron, troilite, chrcmite and ilmenite. The pyroxenes are both orthopyroxenes and clinopyroxenes (2,1) with the composition quite variable (Fsl5-Fs7 0 ). The plagioclase is anorthite ranging mainly between An8 5 and An 9 0 .' Olivine comprises about 1% of the mass of the howardites and ranges in composition between Fa 8 - 3 0 . The lithic fragments occur in a wide variety of colors including yellow, yellow-green, black, brown and white contained in a light gray groundmass. Duke and Silver list 17 howardites, making them the second most common type of achondrite. Other Basaltic Achondrites. A separate subclass of the eucrites termed the shergottites (Shergotty) are characterized by the presence of a glassy variety of plagioclase termed maskelymite, presumably produced from plagioclase by extreme shock. Thus the shergottites probably represent merely a structural variation on the eucrites. A single specimen, Serra de Mage, while classified as a VI-24 177 member of the eucrites, is composed mainly of anorthosite feldspar which makes it significantly richer in this phase than the typical eucrite and perhaps should be classified as a distinct group with a basaltic anorthosite composition. In any case, the existence of a specimen this rich in anorthite suggests that the composition field of meteorites might reasonably be extended to anorthite-rich or exclusive compositions. The basaltic achondrites as a group seem to represent some sort of magmatic process (basaltic composition) acting in a gravitational field; layering or bedding of crystals in certain eucrites, such as is seen in terrestrial magma chambers. Additionally the existance of at least one eucrite of basaltic anorthosite composition, which in terrestrial and lunar occurance is indicative of gravitational separation (rising) due to density differences which concentrates this phase near the top of the chamber, supports this concept. Anzrites (augite achondrites) are represented by a single known specimen, Angra dos Reis, which is composed of more than 90% augite with accesory olivine and troilite but no feldspar. The composition augite is (Ca 0o 5 0 g0. 2 7 Feo0 1 5 A10 0 5 Ti o.o3Nao.o0 0 0 1 )

(Si 0 *87 A10 . 1 3 )03 . Angra dos Reis is thus virtually a monominer- allic rock and is structurally unbrecciated. It is analogous to relatively rare terrestrial augite-pyroxenite rocks. Physically fragments are of a deep wine-red color, with moderately coarse crystals ( !Z 41mm). Nakhlites (diopside-olivine achondrites) are represented by two VI-25 ,78

specimens, Nakhla and La Fayette. Mineralogically they are composed of about 75% diopside (Cs 3 9 , Fs24) and about 15% olivine

(Fa66 ) with plagioclase (An5 5 ) and augite. Nakhla is a green- ish mixture of olivine (green) and dicpside (brown) crystals (Z- mm) and is not brecciated. The feldspar, augite and mag- nitite occur in a eucrite-like assemblage. There are virtually no analogous terrestrial igneous rocks of this composition. Aubrites (enstatite achondrites), named for the first member of the group-Aubres, are composed mainly (90%) of very nearly pure magnesium enstatite (0.04% Fe) with clinoenstatite, diopside, olivine (Fa0 ), plagioclase (An25 ), oldhamite (CaS) and osbornite (TiN). The amouht of metallic phase (10% Ni) is on the order of 2%. A detailed study of the enstatite from these meteorites was carried out by Reid and Cohen (1967). Generally these objects are light in color, coarse in texture and highly brecciated. The Cumberland Falls aubrite contains irregular fragments of black chondritic material (shocked?). The extremely low iron content of the mineral phases in these meteorites as well as the presence of CaS and TiN, argue for very reducing conditions during the formation of these objects. No terrestrial analog exists for these meteorites. They are 9 members of this class. Dio-enites (hypersthene achondrites) are composed mostly of hypersthene (Fs20 -3 5 ) with some plagioclase (An80 ), troilite and chromite with accessory nickel-iron and quartz (tridymite). Structurally they are highly crushed and brecciated with large angular fragments of hypersthene in a groundmass of crushed and VI-26 ;79

broken hypersthene. Terrestrial analogues are quite common and occur as layers in large magma chambers. This meteorite class has 8 known members. Chassignites (olivine achondrites) are represented by a single

known specimen, Chassigny, which is composed of 95% olivine

(Fa3 2 ) and approximately 4% chromite with accessory pyroxene, plagioclase, plagioclase glass and rare nickel-iron. It is

nearly unbrecciated. It is analogous to terrestrial dunites which are accumulated on the floors of magma chambers from the first heavy minerals (olivine, chromite) to crystallize out of the cooling basic magma.

Ureilites (olivine-pizeonite achondrites) are probably the oddest of the stony meteorites,. being rich in carbon (1.5-4.1%) in the form of diamond, graphite and organic matter, an extreme disequilibrium assemblage. These meteorites are composed of large olivine grains (Z 85% Fa 1 5 - 2 5 ), clinopyroxene (pigeonite

S-10% Fs1 5-2 0 ) plus kamacite (1.5-4% Ni), troilite, chromite and carbon. Structurally these meteorites are porphritic, containing elongated void spaces generally stretched in the same direction, similar to the elongated bubbles found in terrestrial lava flows. The intergrowth of the carbon materials as trains of a few microns size is considered to be evidence of a dynamic shock origin while in space, perhaps by collision or the breakup of the parent body. Two subtypes of the ureilites are defined on several petrologic criterias a) olivine grains are finer in the second type; b) twinning is more common in clinopyroxene in the VI-27 first types c) iron in the first type has a net-like distribution, in the second type as kamacite plates between silicate grains; and d) the size of the diamond-graphite intergrowths in the first type is less than 0.3 mm while they range to 0.9 mm in the second type. Vdovykin (1970) presents a review of the characteristics of these meteorites. Six ureilites are known. g) Alternate meteorite comrositions It would be highly unlikely that the range of meteorite compositions represented in terrestrial collections presents the full range of compositions of the solid non-volatile material of the solar system. The presence of meteorites such as the diogenites, chassignites, angrites and nakhlites, which'are pure or high'ly concentrated separates of single mineral phases, is an indication that processes which can produce such concentrations were active. Whether these processes can produce asteroid sized concentrations of these phases is another question but the possibility must be considered. Again, while no model is implied, the achondrites, in particular, strongly resemble the range of compositions produced by magmatic differentiation such as occurs within a cooling silicate melt in the presence of a gravitational field. Thus any concentrations of phases which could be produced by this type of process should be taken as strong possibilities. For example, anorthosites (anorthite feldspar-rich assemblages) would be implied, the presence of the very feldspar-rich encrite Serra de Mage supporting this VI-28 line of reasoning. It is therefore justifiable to discuss concentrations of nearly any phase present in meteorites as possible asteroid comparison material.

References

Anders, E. and G.G. Goles (1961) Theories on the origin of meteorites. J. Chem.. Educ. 38 58

Bass, Manuel N. (1971) Montmorillonite and serpentine in Orgueil Meteorite. Goechem. Cosmochem. Acta 351 139

Binns, R.A. (1970) Pyroxene from non-carbonaceous chondritic meteorites. Min. Mar. 21 649

Bunch, T.E.; Klaus Keil and E. Olsen (1970) Mineralogy and petrology of silicate inclusions in iron meteorites. Cont. Min. and Pet. 25 297

Dodd, R.T. Jr. and W.R. Van Schmus (1965) Significance of the unequilibrated ordinary chondrites. J. Geophy. Res. 70 3801

and D.M. Koffman (1967) A Survey of the unequilibrated ordinary chondrites. Geochem. Cosmochem. Acta 31 921

Duke, M.B. and L.T. Silver (1967) Petrology of eucrites, howardites and mesosiderites. Geochem. Cosmochem. Acta 31 1637 Fredriksson, K.1 P. DeCarli and A. Aaramle (1963) Shock induced veins in chondrites. Proceedings of the 3rd International Space Science Svymosium, Washintton 1962 (ed. W. Priester) New Holland Publishine Co., Amsterdam, p. 974. Heymann, Dieter (1967) On the origin of hypersthene chondritess Ares and shock effects of black chondrites. Icarus 6 189

Keil, Klaus (1968) Mineralogical and chemical relationships among enstatite chondrites. J. GeoDhv. Res. 22 6945

and Kurt Fredriksson (1964) The iron, magnesium and calcium distribution in coexisting olivines and rhombic pyroxenes of chondrites. J. Geophy. Res. 69 3487 182 VI-29

Lovering, J.P.: W. Nichiporuk; A. Chodos and H. Brown (1957) The distribution of gallium, germanium, cobalt, chromium and copper in iron and stony-iron meteorites in relation to nickel content and structure. Geochem. Cosmochem. Acta 11 263 Mason, Brian (1962) Meteoritics John Wiley and Sons, Inc. New York

(1963) Olivine compositionlin chondrites. Geochem. Cosmochem. Acta 27 1011

(1966) The enstatite chondrites. Geochem. Cosmochem. Acta 30 23- (1967) Olivine composition in chondrites - a supplement. Geochem. Cosmochem Acta 31 1100

(1971) The carbonaceous chondrites - a selective review. Mreteoritics 6 59 (1972) The mineralogy of meteorites. Meteoritics 2 309 Powell, Benjamin N. (1969) Petrology and chemistry of mesosiderites - I. Textures and composition of nickel-iron. Geochem. Cosmochem. Acta 33 789 (1971) Petrology and chemistry of mesosiderites - III Silicate texture and composition and metal- silicate relationships. Geochem. Cosmochem. Acta 35 5 Reid, Arch M. and Alvin J. Cohen (1967) Some characteristics of enstatite from enstatite achondrites. Geochem. Cosmochem. Acta 31 661

Taylor, G. Jeffrey and Dieter Heymann (1971) The formation of clear taenite in ordinary chondrites. Geochem. Cosmochem. Acta . 747 Van Schmus, W.R. and J.A. Wood (1967) A chemical-petrologic classification for the chondritic meteorites. Geochem. Cos- mochem. Acta 3~ 747 and P.H. Ribbe (1968) The composition and structural state of feldspar from chondritic meteorites. Geochem. Cosmochem. Acta 32 1327

(1969) Mineralogy, petrology and classification of type 3 and 4 carbonaceous chondrites. in Meteorite Research (ed.R. Millman) p. 480 VI-30 183

Wasson, John T. (1967) The chemical classification of iron meteoritess I. A study of iron meteorites with low concentrations of gallium and germanium. Geochem. Cosmochem. Acta 31 161

and Jerome Kimberlin (1967) The chemical classification of iron meteorites: II. Iron and pallasites with germanium concentrations between 8 and 100 ppm. Geochem. Cosmochem. Acta 31_2065 Wiik, H.B. (1956) The chemical composition of some stony meteorites. Geochem. Cosmochem. Acta 2 279 Wood, J.A. (1963) Physics and Chemistry of meteorites. in The Moon, Meteorites and Comets (ed. B.M, Middlehurst and G.P. Kuiper) Univ. Chicago Press

(1967) Chondritess Their metallic minerals, thermal histories, and parent planets. Icarus 6 1 (1967b) Olivine and pyroxene compositions in Type II carbonaceous chondrites. Geochem. Cosmochem. Acta 31 2095 Yavnel, A.A. (1958) Classification of meteorites according to their chemical composition. Meteoritika 15 115 (translated in Inter. Geol. Rev. 2 380) Vdovykin, G.P. (1970) Ureilites. Space Science Reviews 10 481 Chapter VII 184 Experimental Procedures

a) Sample selection The weakest link in most previous studies of the spectral reflectivity of meteorites has been in the quantity and variety of the specimens. This thesis involves by far the most comprehensive spectral study of meteorites thus far attempted. The largest previous study (Hunt and Salisbury, 1973) involved the measurement of 41 specimens, all but three being chondrites. Johnson and Fanale (1973) measured nine carbonaceous chondrites and an iron meteorite. The remainder of the work in the literature consists of scattered spectra of one or more specimens. This study, by contrast, involves the spectral measurement of more than 150 individual specimens from every class or subclass of the meteorites ( 2 36) except three. These exceptions are three types of stony-irons (two represented by single falls) which have quite predictible spectra (see chapter VIII). Additionally, maximum obtainable representation of each type were measured, up to 29 specimens of chondrite type L6. The types of meteorites and the number of specimens of each type are listed in table VII-1. The list of individual specimens can be found in Appendix I. Specimens were acquired from a variety of sources. An initial 45 specimens were obtained with the cooperation of Dr. Clifford Frondel and Mr. David Cook from the meteorite collection at Harvard University. These specimens were important both as a starting selection with relatively good VII-la Table VII-

Types and Numbers of Meteorites Measured

CHONDRITES CHONDRITES 8 Enstatite 19 Carbonaceous 2 E4 2 Cl 2 E5 5 C2 4 E6 6 c3v 6 C30 31 Bronzite ANCHONDRITES 3 H3 9 Eucrite 6 H4 4 Howardite 12 H5 1 Shergottite 5 H6 3 Aubrite 4 Diogenite 50 Hypersthene 1 Nakhlite 1 L3 1 Angrite 4 L4 1 Chassignite 6 L5 1 Ureilite 3 LSB 31 L6 IRONS &'STONY-IRONS 13 Ampheterite 1 Coarse Oct. 3 Medium Oct. 3 LL3 1 Finest Oct. 2 LL4 1 Mesosiderite 1 LLS 6 LL6 VII-2 type-coverage and as a basis of testing and refining the selection criteria and the measurement procedure. However the bulk of the specimens utilized in this study were obtained from the Field Museum collection in Chicago through the good offices of Dr. Edward Olsen. These specimens were carefully chosen and represent a body of high quality specimens which cover the entire petrological range of materials. Additional specimens were provided from the Arizona State University collection and the Yale University collection and were used to expand the coverage of certain classes. Control of sample quality is at least as important as wide coverage. Previous investigators, not well versed in petrology and weathering phenomenon, have ignored or downgraded this aspect of sample selection. The presence of rust stains in a specimen, however faint, is cause for two fold caution. Initially, it must be considered that any effect visible to the naked eye as a color change (i.e. rust color) will also effect the measured spectrum. Figure VII -l shows the change in the spectrum by the rusting (weathering, alteration) of a meteorite specimen. The spectral reflectivity of an iron oxide is also plotted on the same figure, and it can easily be seen how rusted spectra are produced. The rusted specimen is a much better reflector in the infrared and the reflectivity drops off very sharply toward the blue. This steeply reddened visible spectrum (and hence the red color) is a very strong indicator of alteration in the specimen, and is used in deciding 187 VII-2a

Figure VII-1

SPECTRAL REFLECTIVITY (SCALED TO 1.0 AT 0.56 MICRONS)

RLFIANELLO 0.272 L6 14 9 10 APT M(USTED) 0.137 L6 14 10 12 IRON OXIDE (HEMATITE) 0.323 HEM 0 0 0 vii-3 188 which specimens are used in the determination of class spectral characteristics (chapter VIII). The absorption features in the spectra are not, as a general rule, significantly altered in position by this effect, but the comparison of curve shapes and slopes would be invalid. The second aspect of concern, is that the presence of rust may be indicative of the alteration of other mineral phases besides metal. While the rusting of iron causes a sharp visible change in appearance, the weathering of olivine or pyroxene toward hydrous silicates may proceed to a significant extent before the change is evident to the eye. However any such change in the mineral assemblage along with its attendant ion mobility (especially transition metal ions) is certain to alter the spectral reflectivity of the material quite independent of the addition of iron oxides. In addition to changing pre- existing features, this process can add features, especially the water related absorptions, to the spectrum. A most dis- tressing example of this effect is found in the spectrum of the aubrite (enstatite achondrite) 'Bishopville' obtained from the collection at the Field Museum. On initial examination this specimen was unaltered, showing only a minor rust stain on an enclosed metal grain. However a spectrum (figure VIII-8) shows absorption bands due to water, which suggests that the material of this meteorite is no longer only enstatite but contains a hydrated phase of this mineral. Obviously this specimen (and others which exhibit these effects) must be used v89 VII-4

with caution (or excluded) in determining the class spectral characteristics of meteorite groups. A set of criteria were established for the selection of meteorite samples to be used in this studys a) specimens should be free of any visible alteration (rust) or contamination (by dirt, paint, etc.); b) specimens should be free of fusion crusts and old surfaces and thus, whenever possible, should be interior fragments. and c) specimens should be intact (a single fragment or a few fragments) to allow detailed microscopic examination for subtle alteration or contamination effects. While itwas not always possible to obtain specimens which met all these criteria, these are the standards by which individual samples were judged in determining the weight to give their spectra. As a general rule, it was assumed that an imperfect sample (within reasonable limits) was preferable to no sample, as long-as the nature and magnitude of the im- perfection were taken into account. The character of each specimen is noted in Appendix I. It should be noted that the condition (with respect to weathering) can vary substantially for specimens of the same fall in different collections (or separate fragments in the same collection). The condition of a specimen is a function -of all the conditions under which it has been handled. Two extreme cases of this ares a) one of the early carbonaceous chondrites (Alais?) was dipped in hot hog fat to preserve it, and b) the lost City meteorite which in the few hours between '190 VII-5 its fall and its recovery, had a very telling encounter with a canine interloper. b) Samnle DreDaration Since one of the parameters to be studied was the effect of particle size on the reflection spectra, powered samples were prepared for a range of particle sizes. Spectra were measured on materials with textures ranging from whole rock through coarse powders (200-400 microns) to fine powders (30-60 microns). The specimens were either mounted in aluminum trays

(powder) or in sample bottle (whole rock) with a cover glass identical- to that covering the MgO standard. Whole rock spectra were measured from freshly broken surfaces (unless otherwise noted) which were free of dirt or other contaminants. Cut surfaces were measured for several stones, but the results suggest that contamination occurred on these surfaces and their spectra were given low credibility ratings, Powders were prepared by crushing clean, fusion crust free samples in a clean porcelain mortar. In the case of very hard specimens (such as the black chondrites), a steel mortar was employed for the initial crushing. Each sample was examined at several stages of the crushing process for any sign of contamination. After each sample, the mortar and pestle were examined for material loss. In addition to external contamination, rust fragments or stained fragments were removed from the VII-6 S9_1

material at each stage in the crushing process whenever possible. The usefulness of this effort is questionable for the second

reason discussed in section 'a' of this chapter.

The resulting powder was placed in an aluminum tray and covered with a calibrated coverglass, and the specimen shaken

to establish a particle size gradient across the surface from finer (- 40 micron) to coarser ( 300 micron). Spectral measurements were made at a number of locations across this surface to determine the effect of particle size on the relative spectra. This specimen surface was also examined micro- scopically to verify that the proportions of mineral phases seen were similar to those in the original specimen and in that class of meteorite. The sample area measured in the spectroreflectometer was approximately 2mm by #mm and several measurements were made on each sample in order to insure that mineral- separate variations were not introduced into the spectra. The iron meteorites present a particular problem in measuring their spectral reflectivity since crushing these materials to obtain a powder is nearly impossible. Instead, one must use cut surfaces. This is probably no great handicap since a) the spectral reflectivity of metals are not dependent upon particle size (no transmitted component) until the size becomes so small that currents cannot be established, and b) it is very unlikely that the impacts on asteroid surfaces are of sufficient energy to produce metal powders. The cut surfaces of the specimens to be measurel ere cleaned prior to the

ILA VII-7

measurements to remove alteration material. Two etched specimens were measured but the resultine curves are anomolous both with

respect to metals and other iron meteorites. This is probably due to a phase or surface texture resulting from the action of the acid on the metal.

The stony-iron meteorites present even a more difficult

problem since they contain one phase (metal) which cannot be crushed and another (silicate) which can be crushed. Therefore, one must settle for cut surfaces which, unlike the cut surface

of the iron meteorites, cannot be readily cleaned. Additionally,

fresh surfaces cannot be broken and cutting requires a very concentrated effort even for a very small specimen. This problem must be overcome by modeling rather than direct measurement.

c) SDectroreflectometer - Description and Oneration

The instrument used to make the spectral reflectivity measurements of the meteorite specimens is.a Beckman DK-2A

Ratio Recording Spectroreflectometer made available to this investigator by Dr. John B. Adams at the West Indies Laboratory on St. Croix, U.S. Virgin Islands. This apparatus is described by Adams and McCord (1970). The instrument employs an MgO coated integrating sphere with three ports, one for the sample and another for the MgO standard used for comparison.

The third port is the location of the detector (visible - photomultiplier tube; infrared - lead sulfide cell). The apparatus operates by illuminating the sample and standard VII-8 alternately by use of a mirror-chopper arrangement and monitor- ing the AC voltage produced in the detector to determine the relative reflectivity. That is, if both the sample and standard have the same reflectivity, no alternating voltage will be detected and the reflectivity of the sample is defined as 100% (relative to the standard). If the sample reflects none of the incident light, the maximum alternating voltage will be measured and the reflectivity of the sample will be defined as 0%. The 100% and 0% levels are defined by placing MgO surfaces in both beams (100%) and by blocking the sample beam (0%) with an MgO standard in place. The results are outputed in two forms, as a plot'on a stationary chart recorder and as a digital output 6nto punched paper tape, the sample interval for the data on the paper tape being 50 2. Spectra were measured over two intervals depending on the detector useds 0.35-0.65 microns (photomultiplier) and 0.55-2.50 microns (lead sulfide cell), the curves being matched in the overlap region. d) Data reduction and display The paper tapes containing the digital data were read on a paper tape reader and converted to magnetic tape. The digital data were compared with the analog plots to correct any errors introduced by the paper tape punch (e.g. randomly dropped numbers). After the data were corrected, they were plotted, a single plot for all measurements of each specimen. These spectral reflectivity curves are contained in Appendix II VII-9 4

for all the specimens used in this study. The caption along the right edge of the graph contains several pieces of information concerning the plotted curvess a) the number refers to the plotted curve of the same number, b) the caption gives the name and any comments (sometimes truncated) concerning this specimen, c) the number refers to the albedo of the specimen at 0.56 microns, d) the three two-digit numbers are identifi-

cation codes referring to the - -. date, run number on that date and sample number, and e) the last set of alpha-numeric symbols refer to the meteorite class. The long term stability of the system and the MgO standards used for comparison, were checked at the beginning and end of each day by comparison measurement of a gold standard. No changes were detected during the measurements and none were detected in the processing of these comparison runs in the subsequent data reduction.

References

Adams, J.B. and T.B. McCord (1970) Remote sensing of lunar surface mineralogy, Implications from visible and near-infrared reflectivity of Apollo 11 samples. Proc. Apollo 11 Lunar Sci. Sonf., Geochem. Cosmochem. Acta SuDnl. 1 1937-1945 Hunt, G.R. and J.W. Salisbury (1973) Visible and near-infrared -spectra of minerals and rockss VIII Meteorites. In prearation. Johnson, T.V. and F.P. Fanale (1973) Optical properties of carbonaceous chondrites and their relationship to aster6ids. J. Geophv. Res. in press CHAPTER VIII Meteorite Class Snectral Characteristics

a) Introduction The class spectral characteristics (CSC) of a meteorite type are those parameters of the spectral reflectivity curves of this meteorite type which characterize these meteorites. These parameters will be shared by the spectra of all members of the group. From the discussion in preceeding chapters, these parameters can be explained in terms of the crystal field effects in the individual mineral phases and the physical mixing effects. In the discussion of each meteorite type below, these effects will be discussed wherever relevant. The only valid comparisons of observed spectra for mineralogical and petrological information must be made on the basis of group characteristics rather than on the spectrum of any particular specimen. Establishment of group characteristics gives assurance that one is seeing effects which are truly compositional in nature. In the subsequent discussion, the spectra of the meteorites of each type will be discussed in terms of the determina- tive spectral parameters and their relationship to the lithology of the system. It should be noted that meteorites which showed in the spectra signs of alteration (as discussed in chapter VII) are eliminated from the determination of class spectral characteristics or considered only as far as they do not deviate from the unaltered spectra. VIII-2

b) Eucrites, Howardites and Diopenites (Basaltic or Hypersthene Achondrites) 'These three meteorite types are considered together in order to contrast otherwise quite similar spectra created by the dominance of a single mineral phase (hypersthene) in all three types. Mineralogically these meteorites all contain pyroxene (hypersthene) and varying amounts of feldspar, these two phases being present in the following ratios (Pys Feld) ,

Eucrites - lilt Howardites - 3:1 and diogenites - 30:1. Table VIII-1 gives the modal and mineralogical compositions of the specimens used in this study for which definite data is available. The spectra of all three classes have the same overall appearances a steeply rising spectrum in the blue, a less steep or zero slope through the visible, a deep absorption feature centered at about 0.9 microns, a maximum in the reflectivity near 1.5 microns and a broad absorption feature centered at about 2.0 microns. All of these spectra also show a set of narrow (: 50 R) absorption features resulting from spin- forbidden transitions in the pyroxene (Burns, 1970, p. 83). Additionally, the encrites by virtue of their relatively high plagioclase content, show an inflection in the spectral reflectivity curve at about 1.3 microns corresponding to the plagioclase absorption. The 1.3 micron feature is absent from the howardites and diogenites. This plagioclase feature is the Table VIII-1 Modal Mineral Data for Eucrites, Howarditesand Diogenites (Duke and Silver, 1967)

Specimen Type %01 %Py %Plag %Other O1(%Fa) Py(%Fs) Plag (%An) Bereba Euc 0 nd nd nd 87 Jonzac Euc 0 nd nd nd 84-86 Juvinas Euc 0 56 40 4 62(p) 80 Pasamonte Euc 0 63 30 7 48-70(p) 86 Sioux County Euc 0 56 41 2 5 9 (p) 90 Stannern Euc 0 55 39 6 62(p) 80

Frankfort How 1 86 12 2 var var (o,p) 90-95 Pavlovka How nd nd nd nd var var 90-97 Petersburg How(?) 0 70 25 5 58-70 (p) 86 24-45(o) Johnstown Dio 0 95 3 24 90

* The structural type of the pyroxene is indicated by 'o' (orthopyroxene) or 'p' (pigeonite)

Euc = Eucrite; How = Howardite; Dio = Diogenite

01 = Olivine; Py = Pyroxene; Plag = Plagioclase Fa = Fayalite (Fe2+ Olivine); Fs = Ferrosilite (Fe2+ Pyroxene) VIII-3 main distinction between the spectra of the eucrites and howardites. It is interesting to note that Duke and Silver (1967), on the basis of structural relationships have classified the meteorites Petersburg and Nobleborough (Nobleboro) as howardites while in previous work, Mason (1962, 1967) on the basis of calcium content (and by inference, plagioclase content), had classified both as eucrites. Spectrally, both of them show the 1.3 micron inflection and would hence be classified as eucrites, emphasizing the point that the mineral assemblage of the material is the dominant factor in determining the spectral reflectivity. Eucrites (pyroxene-plagioclase monomict breccia) The spectral reflectivity curves of these meteorites are shown on figures VIII-la & ib, the separation being made into two distinct spectral types. The first type is characterized by a relatively steep slope (reddened) through the visible and a clearly defined maximum at about 0.7 microns. The second type has a relatively flat spectra in the visible and no sharp maximum. This second type shows a weakened (or perhaps broadened) feldspar feature and at least one meteorite of this type (Padvarninkai) is a sherzottite in which the plagioclase is in the form of a shock produce glass termed maskelymite. Whether this is also responsible for the difference in the spectral reflectivity of these meteorites is not clear. The centers of the two main absorption bands in these spectra (shown on figure VIII-2, a band-band plot such as devised by Adams, see Adams et. al. (1973),for ten eucrites (including Petersburg and Nobleboro) show the first band VIII-3a Figure VIII-la SPECTRRL REFLECTIVITY ISCRLED TO 1.0 :RT 0.56 MjCRONS1 199

- C0

Ulp U1 2 J- CZD ~rr

4 NOBLEUnO - Ng RUST 0302 HOW 27 2 70 5 PPSAMONTE - RSU 0.333 EUC 28 9 13 6 STANNERN POWDER 0.205 EUC 16 7 50 VIII-3b Figure VIII-lb 2Q0

SPECTRRL REFLECTIVITY ISCRLED TO 1.06T 0.56 MICRONS1

I7 I I I I I ih I I I I

R f-u)

PZI

Lu fl-

0 Li '-3

CD ftr

I I I II I I I I II l ~ I I f ,, f I I | I- • -- . .. -.. .. BEREBR - NO RUST FINER INICE) 0.257 EUC 28 38 50 PRIDVPRNINKRI - NO RUST ISHER) 0.264 EUC 3 q 0 SIOUX COUNTY - FSU 0.355 EUC 27 28 7 20o VIII-3c

Figure VIII-2

Band-Band Plot for Eucrites, Howardites and Diogenites

o H * Center of the Short Band (Microns) * _ o0 II[

H 0l

C) l'o

- 0

\ 0 o crt (D CD

(DO 202 ViII-4 center at 0.90-0.92 microns and the second band at 1.98-2.03 microns.

Howardites (pyroxene-plagioclase polymict breccia) The spectral reflectivity curves of four (three, if Petersburg is eliminated) howardites are shown on figure VIII-3. There is no feldspar inflection in these spectra, and the centers of the absorption features (figure VIII-2) are: first band, 0.88-0.90 microns and second band, 1.91-1.95 microns.

Dioenites (hypersthene achondrites) The spectral reflectivity curves for these meteorites (fizure VIII-4) have the same general shape as the howardites but show a differing degree of decreased reflectivity shortwards of the 0.5 micron spin-forbidden feature. The band centers (figure VIII-2) first band:

0.89-0.915 microns and second band: 1.89-1.91 microns. c) Anrites (augite achondrite) Since this class is represented by a single known fall, the spectral reflectivity curve for Angra dos Reis (figure VIII-5) being a single specimen cannot be considered to define the class snectral characteristic of a group. However the spectral reflectivity curve for this meteorite is quite consistent with that of an augite (figure

II-3) so that any meteorite of similar mineralogy will have a similar spectra. The spectrum is characterized by a very sharply increasine infrared reflectivity with two strong features centered at about 1.0 microns and 2.25 microns. d) Nakhlites iopside-ilivine achondrite) In this study one of VIII-4a

Figure VIII-3 2Q3

SPECTRRL REFLECTIVITY (SCRLEO TO l.0 RT 0.56 MICRONSI

un I-

- Fl

2 LS

-- rt (D En

t FRRNKFOfRT - V. MINOR RUST 0.273 HOW 2-? 6 58 2 LE TEILLEUL - NO RUST 0.299 HOW 2-? 9 t8 3 P:VLUVKR - NO RBUST 0.265 HOW 2- Lj 77 4 PETERSBURG POWDER V.V. RH 0.230 NOW 18 L4 7 VIII-4b

Figure VIII-4 204 SPECTRRL REFLECTIVITY ISCRLED TO 1.0 AT 0.56 MICRONS)

I I I I I I I I I I ir I I I I I I I

(n Q Y 0PC m

rn( rM.r- -h ZCO

L\ CD 0 (D U2o CD

H'@s rt

I II I_ _I I I JOHNSTOWN FINER POWDER 0.353 OJ1 t15 14 29 RODA - NO RUST 0.332 OO1 2-7 1I 76 TATAHOUINE - PfODER NO ALT 0.L423 O1 28 t 5 SHRLKR - NO ALT 0.231 DO1 28 43 72 2Q5 VIII-4c Figure VIII-5 Spectral Reflectivity Curve of Angrite.

SPECTRAL REFLECTIVITT ISCRLEO TO 1.0 RT 0.56 M]CRONSI

o I l I i I II' I I i -u

L -

_ . -( -/

_ /

- ] - I tt I .lt I I ,! I ! l I

• • = • I ANRAR DOS REIS - Ng ALT 0.055 ANG 1 2 68 2 ANGRR ODS REIS - NO PLT 0.053 ANG 3 668 206 VIII-5

the two known nakhlites (Nakhla) is represented so that the same objection applies as with the angrites. However, the spectral reflectivity curve (figure VIII-6) is quite consistent with the spectrum of a mixture of diopside and olivine (see figure II-3) such that it can be considered to be representative of this mineral assemblage. The spectra is characterized by a steep slope shortwards of 0.55 microns, flat between 0.55 and 0.7 microns, a strong feature centered near 1.0 microns, a very strong maximum near 1.7 microns and a strong broad' feature centered near 2.25 microns. The shift of the pyroxene absorption features toward lorner wavelengths (1.0 vs. 0.9 microns) and (2.25 vs. 1.9 microns) is quite consistent with the lower energy transitions 'in the more calcium-rich pyroxenes (Adams and Mc- Cord, 1972).

e) Chassignites (olivine achondrite) This is a single-specimen class, the spectral reflectivity curve (figure VIII-7) being consistent with that of an olivine (figure II-3). The spectrum is strongly reddened in the region below 0.5 microns, with a very broad strong feature centered near 1.05 microns with sidebands in the main feature near 0.8 and 1.3 microns (a typical olivine absorption feature, see section III-e) and is flat beyond 1.7 microns. f) Aubrites (enstatite achondrites) The spectral reflectivity turves for the three specimens of this meteorite type are shown VIII-5a

Fiaure VIII-6 2Q7

SPECTRRL REFLECTIVITY SCRLED TO 1.0 AT 0.56 MICRONS)

C n _- m (D Y0 O

Z ml"(D CD -rt-r1

n n z

z Ri SCpi ZHr C Z(D

Ri Ll

NAKnLA FINER POWDER 0.132 NFK 15 15 39 VIII-5b Figure VIII-7 208

SPECTRAL REFLECTIVITY ISCRLED TO 1.0 PT 0.56 MICRONS1 0 a-

UD rt

n I-

o - Zo3 --

f t

I II I I , I I I

t CHRSSIUNY - NO LLTERPTlrN 0.352 CHRS5 2-7 12 31 20Q9 VII-6

on figure VIII-8. In considering the class spectral characteristics, the merits of each specimen must be considered. "Cum- berland Falls' was chosen from the Harvard collection and showed

no signs of alteration before or after crushing or in the spectrum. 'Norton County' was supplied as a powder by Arizona State Univer- sity and exhibited a faint reddish stain (the meteorite is

generally white) indicating alteration. 'Bishopville' was chosen from the Field Museum collection and showed minor rust

around a metal fleck but otherwise appeared unaltered, however

the spectrum shows a water absorption near 1.9 microns. In addition both of the latter two specimens have steeply reddened

spectra through the visible indicating rust stain. Therefore only 'Cumberland Falls' seems unaltered and its spectrum is con-

sidered to be most typical of the aubrites. While the other two specimens cannot be used directly to determine the CSC of the

aubrites, their spectra are consistent with a 'Cumberland Fall' type spectra to which rust stain has been added. The spectrum of an aubrite is thus a featureless spectra (consistent with a very pure (FsO ) enstatite) with a slightly reddened slope in the visible and a flat or slightly decreasing infrared reflectivity. This slow decrease of the infrared reflectivity to- wards loneer waveleneths is not seen in the measurements of the reflectivity of enstatite as shown in figure 11-3. However the enstatite measured by Adams and Filice (1967), from which the figure was taken, was a terrestrial enstatite. Since such enstatite never approaches the purity (Fs0 ) of meteoritic en- VIII-6a

Figure VIII-8 SPECTRAL REFLECTIVITY ISCALEO TO 1.0 AT 0.56 MICRONS1 C ,- n

W w

C CCU -n

C n0

L CUMBELRND FRLLS [EN. ACMHN.! 0.232 RUB 1q Ilq 18 2 NORTON COUNTY - ASU, FAINT R 0.L424 PUB 28 I8 18 3 BISHCPVILLE - SOME ALT 0,1481 RUB 28 39- 51 211 VIII-7

statites, this aubrite spectrum may indeed be the true enstatite spectrum. This apparent inconsistency should be examined more closely in subsequent work.

. ) Ureilites (olivihe-pigeonite achondrites) This very interesting class is represented by a single specimen in this study be-

cause only 'Novo Urei' was available in the meteorite collections

searched, however every effort is being undertaken to obtain

other specimens, especially the newly fallen 'Havaro' ureilite. As in the carbonaceous chondrites (discussed in chapter IV) the

presence of diffuse carbon dominates the spectrum of this met-

eorite. The spectrum (figure VIII-9) has a reddened visible, a

flat red and near infrared, a weak (15%) feature near 0.95 microns and a slowly rising infrared reflectivity with a very weak broad feature near 2.0 microns. However this specimen

showed evidence of minor alteration, presumably of terrestrial origin, which might be the cause of the slope in the visible and the infrared. Although if this is the case the effect does not dominate the spectrum. However the spectrum of a urielite should probably be a bit flatter. This class is being expanded as described above, but the spectrum shown is consistent with the mineral phases present. h) Octahedrites and Nickel Rich Ataxites (Iron Meteorites)

Despite the problems inherent (see chapter VII) in measuring the spectra of these meteorites (figure VIII-10), the spectra are consistent with those of metals. A general pattern of de'- VIII-7a

Figure VIII-9 SPECTRAL REFLECTIVITY ISCALEO TO 1.0 AT 0.56 MJCRONS1 -

I I I I I I I I I I I I I I I I I I I

C CA CD 0o

t-hr

:D33

C (D zysi CI-° ) I-.

0 CD

I-

I I I I I I I I I II

I NOBV-UREI - MINOR RUST IXSI 0.706 UREI 2 1t 82 VIII-7b

Figure VIII-10 SPECTRRL REFLECTIVITT ISCRLED TO 1.0 RT 0.56 MICRONSI 2X3

0 p 1 1 1 1 1 1 1 1 1 1 1 I C

Lf Ign o *tn rt -i

C (D(D

I-A rn

0 ~3C2

ft-

02 toM

CJC fn CjU U - 01

- 1 l1 1 1 .. 01 CSE'Y CCONTY - CUT SUR 0.168 CicT 26 15 76 CHULRFINNEE - CUT SUR 0.173 MOCT 26 11 93 BUTLER - CUT SCATED SUR 0.289 FSTO 26 6 21 BABB'S MILL - CUT SUR 0.232 NRAX 26 'Il5 234 VIII-8

creasing infrared reflectivity with increasing nickel content, is consistent with the relatively low reflectivity of nickel as compared to iron. Two specimens (Coopertown and Juncal) were eliminated from this comparison because of evidence that the etching gave rise to anomolous features (a decreasing infrared reflectivity) inconsistent with a metal, and probably results from some phase prodced by the action of the acid on the metal. All the spectra of these meteorites are featureless and differ in the slope of their reflectivity curve in the infrared as a function of nickel content. A single iron (Odessa), measured by Johnson and Fanale (1973), exhibits this same pattern.

i) Yesosiderite (pyroxene-plagioclase stony-iron) The whole group of stony-iron meteorites present one of the most difficult measurement problems to be faced in this study (discussed in chapter VII). However, this problem is compensated for by the fact that these spectra are very easily modeled. They involve two well understood phases whose spectra add together quite simply to produce the compound spectra. The spectrum of a mesosiderite (figure VIII-11) is that of an octahedrite (7-9% Ni) with the absorption bands of a howardite imposed weakly on the spectrum. The spectra of pallasites (olivine stony-irons), siderophyre (bronzite-tridymite stony-irons), lodranite (olivine-bronzite stony-iron) or any other hypothetical stony- iron should follow this general pattern. Any measured spectra of these meteorite types should be considered carefully because VIII-8a

Figure VIII-11 215 SPECTRRL REFLECTIVITY 15CRLEO TO 1.0 RT 0.56 MICRONS5

CD I I I I I I I I I I I I 1 1 1

I I I I I I I I I I I I I I I I I 1 I I L VERRMIN WHOLE ROCK RA 0.177 MES 18 16 51 216 VIII-9 of the problems involved in obtaining a representative sample. Preferably, the modeling approach should be used to define the pattern of the spectral reflectivity curves for these meteorites. In the preceedin sections of this chapter, we have discussed the seventeen types of meteorites which comprise the non-chondritic classes of meteorites. In this study, we have utilized 31 specimens to measure the spectra of 13 of these 17 types. As discussed above, the four unmeasured types, a hexaza hedrite (an iron meteorite with less than 6% Ni) and three stony- irons, have quite predictable spectra. Thus in this thesis we have truly excellent coverage of these meteorites. By contrast, all previous measurements of these types involve only five specimens: the Nuevo Laredo eucrite (McCord et. al., 1970), the Odessa coarse octahedrite (Johnson and Fanale, 1973) and the Pasamonte eucrite, the Dapoeta howardite and the Nakhla nakhlite (Chapman and Salisbury, 1973; Hunt and Salisbury, 1973). There- fore, it is clear that the work in this thesis, with regard to these non-chondritic meteorites, is much mare extensive and complete than all of the previous work combined. j) Chondrites: Introduction The whole family of chondritic meteorites has been left until the last part of this chapter in order that the discussion for this very large, very complex group can be considered in light of the preceeding more restricted groups. The chondrites show not only a significant range of compositional variations but also exhibit a wide range of 217 VIII-10

metamorphic and structural types not seen in the other meteorite classes. (This may be the result of having a large number of chondrites and relatively few non-chondrites.) Each of these separate compositional and metamorphic classes should be treated as a distinct group, but in many cases due to the rarity of certain types they will be considered together, however as far as possible the gaps will be filled by extrapolating from other adjacent types.

k) Enstatite Chondrites These meteorites are composed of very

pure enstatite and metal (4:1) and the spectra (figures VIII-12a

& 12b) show this combination. The spectra are featureless (as both enstatite and nickel-iron) with a variable reddish slope in the visible and a very slowly increasing infrared reflectivity.

Because of evidence of rust alteration in the spectra of Indarch (E4) and St. 1Marks (ES), the steeply reddened visible spectra,

these two cannot be used in determining the CSC of the E-type

meteorites. The comparison of the remaining spectra with the spectrum of the enstatite achondrite 'Cumberland Fall' (figure VIII-8) is instructive. The only significant difference in the mineralogy of these two types of meteorites is in the amount of nickel-iron present (chon. 20%; achon. 2%). This extra metal in the chondrites is almost certainly responsible for the in- creased reflectivity of these meteorites in the infrared. The sirnificantly steeper slope in the infrared of Abee (E4), as compared to the three E6 meteorites, seems to be a real petro- VIII-10a

Figure VIII-12a 218 SPECTRRL REFLECTIVITY (SCRLED TO 1.0 fT 0.56 MICRONS1 n,

RBEE - MINOR PLT IXS1 0.065 EL4 3 -7 7tL4 INODARCH IXS ) 0.055 EUL 3 29 28 0.08eBE5 3 16 8 ST MRK'S - V RRRE RUST IXS) tm VIII-10b

Fiaure VIII-12b 219 SPECTRAL REFLECTIVITY ISCALED TO 1.0 PT 0.56 MICRONS)

I MVITTIS 0. 179 E6 3 10 76 2 KHAIRPUR - RUST ST 0.117 E6 2 11 38 3 PILLISTFER - NO RUST 0.162 E6 3 2q q 6 VIII-11 220

logical or mineralogical effect and not an effect of rust.. Also, it does not seem to be a direct effect of the metamorphic grade

since Indarch and St. Marks, with the additional reflectivity of the iron oxide, are still less reflective than Abee in the

infrared. If the metamorphic grade were responsible for the high reflectivity of Abee, Indarch and St. Marks should both be

significantly more reflective than Abee. Among the four un- weathered specimens, the metal/silicate ratio is, Abee, 0.49;

Hvittis, 0.391 Khairpur, 0.26 and Pillistfer, 0.37 (from Keil, 1968). The higher infrared reflectivity of Abee may quite possibly be the result of the relatively higher metal content of this meteorite. Hunt and Salisbury (1973) have measured the spectrum of Abee and show a similar spectrum.

1) Ordinary Chondrites (Bronzite - High Iron; Hypersthene - Low Iron; Amphoterite - Low Iron Low Metal) are the largest group of meteorites in falls and they differ primarily in the amount of metal and the composition of olivine and pyroxene phases. The spectral reflectivity curves all exhibit common characteristics, a reddish spectra in the visible rising to a broad maximum near

0.7 microns, a feature of variable strength (depending on the metamorphic grade) centered between 0.9 and 0.97 microns (depending on the petrologic type), a rising reflectivity toward 1.5 microns with a feldspar absorption causing an inflection in the curve around 1.3 microns and a broad relatively weak absorption centered in the region of 1.9 microns. 221. VIII-12

m) Bronzite Chondrites (High Iron) The spectra for H-type chondrites are shown on figures VIII-13a,b,c. The minima (centers) of the absorption features for the H-type chondrites are shown on figure VIII-14, the first bands 0.88-0.925 microns and the second bands 1.85-2.00 microns. Although none of the metamorphic subgrades of this type are as heavily represented as those of the L-type chondrites, the same pattern of the band intensity is followed, that is, the band is weakest in the lowest metamorphic grades and strongest in the highest grades. This is attributable to the increasing uniformity of the mineral assemblage (equilibrium) where the pyroxene phase more closely approaches a single mineralogical composition. The effect of such increasing uniformity of the pyroxene composition is to produce an increasingly narrower envelope of individual absorptions due to given pyroxene fragments. This results in a sharper and deeper absorption feature in the higher metamorphic grade meteorites of a particular class. Thus in general, an H6 spectrum has a deeper feature than an H4. However, it is relevant to note that the metamorphic grade of a meteorite in the Van Schmus and Wood classification is not a discrete choice, that is, a specimen may be classified as either of two adjacent grades in a large number of cases (Wood, 1973, personal communication). Thus it is quite reasonable that there is a degree of overlap between the band depths of adjacent grades. This effect is shown clearly for the case of the L-type chondrites (figure VIII-15) VIII-12a

Figure VIII-13a 222 SPECTRPL RErFLECTIVITY ISCALEDO TD 1.0 PT 0.56 M]CRONS)

Q -j

, O O 1

-I1 *

'U1

DO rO

CD (n i

L TIESCHT2 NO RUST 3 2 29 2- K MIXED N ,

i TIESCHIT - NO RUST 0.137 H3 26 29 25 W 2 CHP.NSK - MIXEQ NO RUST f.215 Hil L 1ff.l P VIII-12b

Figure VIII-13b 223 SPECTRAL REFLECTIVITY 1SCALED TO 1.0 PT 0.56 MICRONSI

I I I I I I I I 1 I i~ l i i i I I I

C c CD 0 rt Ii FJ

P1-rn< H-

-1C) Zn c 0n

Llj 0,(D I-a (D

711-- 11 I ll I l 11 111 CRSTFLIR FINER PWODER, IOJRK 0.125 H5 16 20 1-7 COLLESCIPOLI - V. RPRE RUST O.2q2 H5 2 11 56 PFNTFR - ASU 0. 257 H5 28 21 19 VIII-12c 224 Ficure VIII-13c SPECTRRL REFLECTIVITY [SCRLED TO 1.0 PT 0.56 MICRONS o r\

oCD

Lfl rt

mM _CH cj H SC 5ni .. rt"

-4

r C~1 C

rrj C3 -

2:i(D SC N R) t Wi K I Ki

LRNCON - VVR RUST 0.29L H61 I 35 26 QUEEN'S MERCY - CARBSER MINOR 0.227 H6 2 19 0 225 VIII-12d Figure VIII-14

Band-Band Plot for the Ordinary Chondrites

o. .D Center of the Short Band (Microns) o I

C) r CD

i1-> Ii0

~rr 'd ;C ,o. II Ii Ii I t-4 t-i' '

C~(D '-3(1) (D 226 VIII-12e Figure VIII-15

Effect of metamorphic grade on the strength of the absorption feature.

I i I I f I I I i I i - I

cr4

CII Q

• e e .. . *'.0 .."*

C (C) • 0 a ~ * 4 /r a ** go ,0 I

•r 1 t 0

1 l _ _ __ i I I t !. 1 g 227 VIII-13

n) Hypersthene Chondrites (Low Iron) The spectral reflectivity curves for the L-type chondrites are shown on figures VIII-16a, b,c,d and the minimums of the absorption features are shown on figure VIII-14t the first bands 0.90-0.95 microns and the second band: 1.875-1.975 microns. The distribution of the spectral reflectivity curves fdr the L6 specimens is of particular interest, since this is the croup with the greatest representation, 15 good specimens. An examination of figure VIII-16d shows that even with such a large number the range of dispersion in the spectral reflectivity curves is small, the envelope of all these spectra being much smaller than the amplitude variation across the spectral range.

o) Amnhoterites (Low Iron-Low Metal) The spectra of the LL-type chondrites are shown on figures VIII-17a,b and the centers of the absorption features on figure VIII-14 the first

bands 0.925-0.975 microns and the second bands 1.91-1.98 microns. The pattern of deeper features with increasing metamorphic grade is not clearly shown because of the small number of cases, but the pattern is the same as is shown by the L and H type chondrites. p) Black Chondrites As defined in chapter VI, these meteorites are highly shocked specimens with very low albedos (0.05-0.10). Of the four black chondrites shown on figure VIII-18, two are L5 (L5B), one is an undesignated L-type and the last is an H-type. Their spectra are all characterized be being quite flat VIII-13a

Figure VIII-16a 228 SPECTRRL REFLECTIVITY ISCRLED TO 1.0 RT 0.56 MICRONS

S I I I I I I I I I I I I I I I I I I-

Fn (D 0 I"

C< (I- ZCDZ:f

rq<0- H-

-rC-4

n co Ln IC ruC S0- ru

(,B2(D

P-- !d 0 t I K Ki.]1i i III i l l - MEYO-MRAORRS RRRE RUST 0. 103 L3 18 g 56 VIII-13b

Figure VIII-16b 229 SPECTRRL REFLECTIVITY ISCRLED TB 1.0 FiT 0.56 MICRONSI

rt0

0< -IED

I r ftQji- ctCD

m(D ISC H- n

0 I'I"ri-

I BRLD MOUNTAIN - NO RUST 0.188 LI 5 10 92 2 CYNTHINA FINE PBWDER 0.221 Lq 1-? 19 3 SFRATBV - ASU 0.207 LL 28 16 17 VIII-13c

Figure VIII-16c 2-0 SPECTRRL REFLECTIVITY ISCRLED TO 1.0 RT 0.56 MICRONSI 01r

(D 0rt

(D0

rt0 I-n

(lcD :I-(D - CD-J -sr n

-, C r:O1

-7 X

_13-,- 0

0 - Cl- ul

PUSSON FINER POWDER 1ST 0.247 2 PUSSON FINER POWDER [NO 0.,249 LI HUMESTERD CG1RSER PoWO 0.179i 22 KNYHMINF A 1WHLLE RBCK 0. 172 SHELBURNE RARE RUST 0.239 11

.1 ' VI l-13d

Figure VIII-16d 23 SPECTRAL REFLECTIVITY ISCALED TO 1.0 Al 0.56 MICRONS)

! I I I I I I I I I i I I I I I 1 I

rn (D rC)

I-'

(D E m CD r- nn rl

r z H' C G)LA rn

ru 0 t-4

(D n CDt,

:j~ Ln Ri (D

S(D 0 I i I I I I I I - I I I I I I I1 II • II ...... Ii iC- ALFIANELLf 0.272 LB 1L 9 10 0O ANDOVER - SOME RUST 0.314 LB 4 18 66 AUMRLE - NO RUST 0.331 LB q 21 37 BRUCERMEIM - ASU 0.328 LB 26 5 11 BUSCHCF - N2 RUST 0.365 LB W 31 -75 CRBE70 DE MRAY - FINER, NO R 0.343 LB 26 31 l4 CCLB fIWIST - MIXED V ,ARRE RUST 0.271 LB 3 13 89 DRAKE CREEK 0.351 LB 15 26 23 GCRGENTI - VR RUST 0.264 LB 2 11 47 LEECET - ASU 0.317 LB NERFT - RRE RUST 0. 51 L6 27 16 85 ST. MICHEL FINER PWC.R 0.321 LS 15 16 9 TCUl1NNES LA GROSSE - RRRE RUS 0. 295 LS 1 17 72 UTRECMT - MIXED MINOR RUST 0.273 L6 ' 16 0 !pVID - VR RUST 0.290 L6 4 12 61 VIII-13e Figure VII-17a 232 SPECTRAL REFLECTIVITY ISCRLED TO 1.0 AIT 0.56 MICBONS)

I' I I I 1 11 I I I I' 1 1 1I I rn ) a

20 in oC H r-IftY

(D N

| u cr- CL rf(D

cr/ 00

1 Iv tn r- cD

I~ I ~ I I I I I I I I m ul SKOI-BRNJR - MIXED NO RUST 0.272 LL4 L 33 19 OLIVENYi 0.370 LL5 114 1 6 VIII-13f

Figure VIII-17b 233 SPECTRRL REFLECTIVITY lSCRLED TO 1.0 RT 0.56 MJCRONS) aor O D 0

- l I

0-i

(D0

2 MNH - N RUST 10000) 0.22 LL6 2 3 83

---

1 I ! I I I I ! I I I I! I! I I VIII-13g

Figure VIII-18 234 SPECTRFRL REFLECTIVITY ISCALED 70 1.0 AT 0.56 MICRONSI

C) (DD

Sl-

-4

- 0 :m

2 PARAGOULU - NO BUST IX 0.085L5B 3 20 L- 3 SEVRUKOVO - NO RUST !XSl 0.081 L B I 23 2 i ROSE CITY - W BOX IXS) NO V UL 0.065 H 696 5 12 235 ViII-14 with subdued. features. Their shock origin may make them a more

important surface constituent of an asteroid surface than their abundance would otherwise suggest. These featureless, rather

flat spectra, also tend to decrease the uniqueness of any identification made by this technique, since they are spectrally similar to the ureilites and the carbonaceous chondrites. This problem will be discussed in more detail in the next chapter. q) Carbonaceous Chondrites The spectral reflectivity curves of the four carbonaceous chondrite types (C1, C2, C30, C3V) are

shown on figures VIII-19a,b,c,d which will be discussed separately below. However, the spectra of all members of this

class are dominated by the presence of a dielectric opaque phase (carbon) which causes a very low albedo and subdued or absent features. Much of this discussion parallels that of Johnson and

Fanale (1973) which has been considered in chapter IV. The increasing metamorphic grade of these types follow a change in the basic mineralogy from hydrous to anhydrous silicates.

Type Cl are the most primitive (hydrous) specimens of this class being represented in this study by two specimens, Orgueil and

Alais. Unfortunately, Alais being the first recorded meteorite of the carbonaceous type (having fallen in the early 1800's) has not been well preserved and spectrally shows very significant

alteration (the darkness of the material makes naked eye ob-

servation of this effect very difficult). The water related

feature at 1.9 microns in this spectra is almost certainly due VIII-14a

Figure VIII-19a 236 SPECTRRL REFLECTIVITY ISCALEO Ti 1.0 T 0.56 MICRONS1

- I I I I I~ I I ill I I I I 1 1 I I lrl Irn

CD Q ft I u1 mFJ 0r m0l-

6--a 10 :5 0 0CD 0C

0 'l-A Pn LrnTi t C c

Ri 0 a 0 z NoSC P-r mO c 2Y-. 0 0

C U1 U2

H mlr i --I I I I I I I --I I I I I- I -- I I I I RLA IS IXS ) 0.030 Cl 1 11 66 BRGUEIL fXS 0.040 CI 3 27 57 VIII-14b

Figure VIII-19b 237 SPECTR:L REFLECTIVITT ISCRLED TO 1.0 PT 0.56 MICRONS1

Sc a

ITI'tb

- 0

- 1

Iino H-

1 1 I I I I I I I ,I I I I I I I I I I COLD BCKKEVELT 132-100M) (XS) 0.033 C2 q 3 1 2 COLD BBKKEVELT 1100-2007) IXS) 0.035 C2 [4 5 1 3 COLO BBKKEVELT GOT 200M) fXS) 0.038 C2 q 7 1 1 MEGHEI - IGT 200M) IXSZR) 0.OO4 C2 5 214 5 5 MURCHISBN - IXS) 0.0 C2 5 37 1q 6 MURRRY - RSU (XSI 0.Ot5 C2 2 9 26 7 NCM~PY - IXSZR) 0.04, C2 5 28 60 VIII-14c

Figure VIII-19c 238 SPECTRAL REFLECTIVITY (SCALED TO 1.0 AT 0.56 MJCRONSI F-- U

r rt" ul

F- C) (D 0 rt-

D rt-

III ilII 1 1111 1 111 -30tU2 ul n

0 rU

0S= rz1 D 0

0 FELIX IXS) 0,086 C30 1 13 30 KAJNSRF - IXSI 0.073 C30 5 26 51 CRNRNS - Ng RLT IXS) 0.120 C30 2 23 66 WARRENTON - INS) ICT 200M) TVJ 0.172 C30 5 13 3 WPRRENTON - INS) 1100-200M) TV 0.106 C38 5 11 3 WPRRENTON - IXS) 1100-200M) TV 0. C31 5 15 3 VIII-14d

Figure VIII-19d 131 7_ SPECTRAL REFLECTIVITY [SCALED TB 1.0 IT 0.56 MICRONS1

S1111 11 ~ii i i I ii

0 Q ul

rt rt

M 0-

zln Ll

0 o C 0

0 U P1trl .FW r=u Gl 0 !)

I I I I I I I ,I I , I 1 I I I I 1 I w RLLENDE - Nt PlLT IXS) 0.091 C3V 5 18 32 GRBSNRJ: - IXS ILT 200M) TVJ 0.1107 C3V 5 14 2 GRBSNRAJ - IXS) 1100-200M) TVJ 0.045 C3V 5 6 2 GRBSNAJA - IXSI 132-100M) TVJ 0.052 C3V 5 8 2 LEBVILLE - ASU IXS) 0.089 C3V 2 3 2?1 MBKiOA - ASU IXSI 0.069 C3V 2 6 25 VIGRRANB IXSI 0.071 C3V 1. 7 82 240 VIII-15

to terrestrial absorbed water. This spectrum was included to show the effects weathering can have on the spectral reflectivity of even such an 'earth-surface' (oxidized and hence not subject

to easy weathering) type material. The spectrum of Orgueil is

probably typical of these primitive types, with a reddish visible and a flat infrared and only a hint of a feature near 0.9 microns.

Tyve C2 are less primitive than type Cl, containing chondrules as well as olivine and pyroxene grains. Spectrally these meteorites are nearly featureless with a reddened visible and a flat or increasing reflectivity curve into the infrared. The variation of the infrared reflectivity with size has been attributed to the diffusion of the crushed olivine and pyroxene minerals at smaller particle sizes by Johnson and Fanale. The basic pattern is the same over the size range measured, the spectra are consistently featureless. Type C30 shown on figure VIII-19c, are characterized by the same general pattern of spectral reflectivity, except that the first and second bands are somewhat better defined in these spectra. The flat infrared spectra are consistently found in this meteorite type. Tvye CSV shown on figure VIII-19d, show the same dispersion of infrared reflectivities with size that type C2 shows, and probably for the same reasons. The spectral features weakly present in the C30 type are completely absent in these 241

meteorites.

r) Alternate meteorite comDosition - Anorthosite As discussed in previous chapters, anorthite feldspar is a reasonable extension of known meteoritic material for comparison to asteroids. Since no meteorite of this mineraloy exists in terrestrial collections (Serra de Mare is the closest), a reasonable substitute must be chosen for the spectral measurements. The mineralogical relationships and degree of reduction of the feldspar-rich eucrites is more similar to that of the lunar anorthosites than that of terrestrial anorthisites. The reduced state of the lunar and meteoritic feldspars allows a significant trace amount of Fe to substitute into the feldspar structure. Bell and Mao (1973) verified that it is the divalent iron in the feldspar which produces the 1.25 micron absorption feature in anorthite. Adams and VcCord (1971) have measured the spectrum of a plagioclase separate from a lunar anorthosite, and the curve they obtained is shown on figure VIII-20. From this curve, the CSC of an anorthite assemblage is seen to be: a sharply reddened visible spectrum, a sharp maximum near 0.8 microns, a deep absorption feature centered at 1.25 microns, and a rising featureless reflectivity curve toward longer wavelengths. There also appears to be a weak inflection or absorption feature in this spectrum near 0.65 microns. Although Bell and Mao (1973) did not see the 0.65 micron feature in their transmission spectra of the lunar VIII-16a 242

Figure VIII-20

SPECTRAL REFLECTIVITY (SCALED TO 1.0 iAT0.56 MICRONS)

a -C I I I I Ii I i i i i i0i i I i1_

cn 0 (D 0 cn ii aH

rMi C)

0 -r H0

o Z: CDr -

ni rna z n Cm- FJ-

z a(D Or p *

I I I I I I 1 I I I I ! 1 ANORTHITE FELD (LUNAR 12063 - 0.351 AN 0 0 0 243 VIII-17

plagioclases, it should be remembered that they measured only

the o4- spectrum of these crystals and, as was discussed in

chapter III, the 0.65 micron feature may not exist in the

$.-spectrum but may be quite strong in the / or ~ -spectra. Since this feature also seems to be strongest in those meteorites

with the highest anorthite content, .this feature should be considered to be diognostic of feldspar. However, the 1.25 micron

absorption feature must be considered most diognostic of anorthite.

s) Conclusions As has been seen in this chapter, each meteorite type representing a particular mineral assemblage and metamorphic grade, has a characteristic spectral reflectivity curve. There

are three general types of spectral reflectivity curves, those with strong spectral features, those with weak features and

those which are featureless. The first group includes the ordinary chondrites, the basaltic achondrites, diogentites, nakhlites, angrites, chassignites and anorthites. The weak featured group includes the ureilites, black chondrites, the

stony-irons and some of the carbonaceous chondrites. Those with featureless spectra include the iron meteorites, enstatite

chondrites and achondrites, and some of the carbonaceous chondrites.

The presence and location of absorption features as well as the overall spectral curve shape are consistent within 244 VIII-18

each mineralogical and petrological meteorite class. In general, there is quite sufficient differences between the

spectra of different types to permit discrimination and identification of surface type by remote spectral measurements. The previous spectral reflectivity work on the meteoritic materials, primarily by Hunt and Salisbury (1973) provides good agreement with this work in the areas where there is overlap. Perhaps more importantly, this thesis provides a detailed examination of the parameters, both physical and chemical, of a material

surface which affect the spectral reflectivity of the surface.

Without a clear definition of such criteria and their effects, the credibility of remote mineralogical interpretations is poor.

References

Adams, J.B. and A.L. Filice (1967) Spectral reflectence 0.4 to 2.0 microns of silicate rock powders. J. Geonhys. Res. 72 5705

and T.B. TMcCord (1971) Cptical properties of mineral separates, glass and anorthositic fragments from Apollo mare samples. Proc. Scon1 Lunar Sci. Conf., Geochim. Cosmochim, Acta SunD1. 2 2183-2195 (1972) Electronic spectra of pyroxenes and interpretation of telescopic spectral reflectivity curves of the moon. Proc. Third Lunar Sci. Conf. , Geochim. Acta SuDD1. 2 3021-3034

., P.M, Bell, J.E. Conel, H.K. Mao, T.B. McCord and D.B. Nash (1973) Visible and near-infrared transmission and reflectance measurements of the Luna 20 soil. Geochim. Cosmochim. Acta 37 731-743 Bell, P.M. and H.K. Mao (1973) Optical and chemical analysis of iron in Luna 20 plagioclase, Geochim. Cosmochim, Acta 37 755-759 245 VIII-19

Burns, R.G. (1970) Mineraloical Applications of Crystal Field Theory (Cambridge University Press) Duke, ?M.B. and L.T. Silver (1967) Petrology of eucrites, howardites and mesosiderites. Geochim. Cosmochim. Acta 31 921 .Hunt, G.R. and J.W. Salisbury (1973) Visible and near-infrared spectra of minerals and rockss VIII Meteorites. In preparation. Johnson, T.V. and F.P. Fanale (1973) Optical properties of carbonaceous chondrites and their relationship to asteroids. J. Geohyvs. Res. in press Keil, Klaus (1968) .ineralogical and chemical relationships among enstatite chondrites. J. Geonhvs. Res. 2 6945 Mason, Brian (1962) Meteoritics (John Wiley and Sons Inc., New York)

(1967) The Bununu meteorite, and a discussion of the pyroxene plagioclase achondrites. Geochim. Cosmochim. Acta 31 107

McCord, T.B., J.B. Adams and T.V. Johnson (1970) Asteroid Vestas Spectral reflectivity and compositional implications. Science 168 1445-1447 VOLUME 3

A SYSTEMATIC STUDY OF THE SPECTRAL REFLECTIVITY

CHARACTERISTICS OF THE METEORITE CLASSES

WITH APPLICATIONS TO THE INTERPRETATION OF ASTEROID SPECTRA

FOR MINERALOGICAL AND PETROLOGICAL INFORMATION 246 Chapter IX Interrretive ADDlications a) Introduction

The work of the preceeding chapter and the vast bulk of the work on this thesis has been concerned with establishing the validity of mineralogical and petrological interpretations on the basis of the spectral reflectivity characteristics of a material surface. The physics involved in the imposition of litholoaical information onto the reflection spectra, as well as the physical characteristics of the material which modify this information, has been considered in some detail.

The conclusion reached was that the spectral reflectivity of a material carries a significant amount of non-redundent compositional information which for the case of the meteoritic material could be used for interpretation within the limits of the observational data. The purpose of this final chapter is to briefly discuss such an interpretive procedure based on the application of the class spectral characteristics defined in the previous chapter to the measured spectral reflectivity curves of asteroidal bodies. This work is included for reasons of illustration and should not be viewed as constituting a major portion of this thesis nor as a complete study. The interpretation of the measured asteroid spectra will be undertaken as an extensive program in the immediate future. b) PhilosoDhical ADrroach

It requires no subtlety to realize that what is sought 247 IX-2 is a known material whose CSC matches those of the reflectivity curve measured for a particular asteroid. Sophistication is required in determining how much discrepancy can be accepted, and more importantly, what any discrepancy signifies, determining as far as possible, what process is responsible for the difference. Thus as shall be seen, a 20% different in the reflectivity curves at one wavelength is no reason for concern, while a 2% difference at another wavelength will be sufficient reason to reject that composition. This involves the use of the class spectral characteristics as a comparison base and an understanding of the processes which cause variation in the spectral reflectivity. These subjects have been discussed in the first eight chapters of this thesis. In many cases, un- ambiguous identification will not be possible and one will only be able to assign a set of probabilities to various comparison materials. However, even at this level of identification, one obtains a valuable piece of information concerning a body which heretofore had been a complete unknown.

And finally, but perhaps foremost since it cannot be emphasized too strongly, if the curves don't match anything, then they don't match. One mustn't be determined to shove an observed curve into some category even if it doesn't fit.

(A square peg can be put in a round hole with the application of enough pressure or the use of a jacknife, but this defeats the whole purpose of the effort.) Moderation of dogmatic application is the watchword. IX-3 248

c) Validity of comDarina asteroid and laboratory data formats

The asteroid data used in this study (Chapman et al.,

1973) has been measured through approximately 24 relatively

narrow bandpass interference filters. However, there is also

a large body of data (see chapter II) which has been measured

through a variety of other filters with different wavelength

and bandpass ranges. To be useful, the work of this thesis

should be applicable to the interpretation of data obtained

by any system of valid spectral reflectivity measurements,

within the inherent limitations of the measurement system.

With regard to the 24 filter system considered here, the

question can be raised as to the validity of the direct com-

parison of data measured through filters with bandpasses of

200-400 to data measured at a spectroreflectometer bandpass

of 5-202. To resolve this problem, the measured transmission

functions of the 24 filters (Elias, 1972) were convolved with

the high resolution data of the laboratory spectra to obtain

the relative reflectivity of the laboratory samples as seen

through the filters. Since the filters are narrow with respect

to the features in the laboratory reflectivity curves, the

'filtered' laboratory spectra were found to lie exactly on the

laboratory curve. Only if the width of features in the laboratory spectra are approaching the width of the filters, does the averaging effect of the filters remove them from the part of the curve corresponding to their central transmission wavelengths. Thus for this filter system, no problem is encountered IX-4 249

in direct overlay comparison of the two sets of data. The 'filtered' reflectivity of the meteorite specimens used to determine the class spectral characteristics are shown in Appendix III.

However for other systems in use, the picture is not nearly so rosy. The commonly employed UBV(RI) system (John-

son and Mitchell, 1962) employs very broadband (=' 2000) filters in which case there can occur a significant degree of averaging of spectra features across a filter. For this case

the direct overlay comparison becomes a questionable procedure.

Thus for comparison laboratory spectra should be convolved

through the filter transmission functions and these values compared with the observational data. These 'filter..' reflectivities in the UBV(RI) system for the specimens used in Chapter VIII are listed in Appendix III. It should be emphasized that the lowered resolttion of this system results in a correspondingly lowered lithological resolution. To maximize the compositional resolution, one must obtain the highest reasonable resolution in the measured asteroid spectra. d) Comparison nrocedures

This section will cover the procedure followed in determining the surface material of several asteroidal bodies from comparison of their spectral reflectivity curves both to the curves and to the class spectral characteristics for the meteorites. The meteorite classes, grouped into logical assemblages, are shown on Figures IX-l, plotted for the 250 IX-4a

-e Figure IX-la

-,...

EUCR T-T - TYPE I

Figure IX-lb

- EUCRITES - TYPE 2

1 1 11 1 I 1 1 1I 1 I I 1 0,3 0.5 0.7 0.9 1.1 251 .IX-4b

Figure IX-lc

HOWAqRO1TES

Figure IX-ld I

L U]BGENITES - 1IGP-OL ACHON

1 1 1 I 1 1 1 1 I I I I I I 1 I I i I 0,5 0,7 0.9 1.1 252 IX-4c

Figure IX-le

H-TYPE CH0NORITES

Figure IX-1f

LL-TYF'E CHONORITES

0.3 0,5 0.7 0.9 1.1 IX-4d

Figure IX-1a

-4

CHiN - LU

Figure IX-1h

CHGN- - L5

S1111111111111.7 0.9 0.3 0.5 0.7 0.9 1) 254 IX-4e

Figure IX-li

CHiN - L6

Figure IX-li

BLACK CHONDRITES

0,3 0.5 0.7 0.9 1.1 255 IX-4f

Figure IX-lk

CRRBNACEOUS CHiN - TYPE Cl

Figure IX-11

C:RBONACEOUS CHON - TYPE C2

0.3 0.5 0.7 0.9 256 IX-4g

Figure IX-lm

CARBONACEUS CHBN - TYPE C30

Figure IX-ln

CARBBNACEUS 0.5CNON - TYPE C3V

0.3 0,5 0.7 0.9 1.1 257 IX-4h

Figure IX-lo

ENSTEITE CHON.

Figure IX-1p

RUBRIES ENST ACIHGN

T 0.3 0.5 0.7 0.9 1.1 258 IX-4i

Figure IX-lq

_ CHRSS]J N7TE - OL ACHUN

Figure IX-lr

_- i I I

SNAIKHLIE - DIP-OL ACHON I

I IIII I I I I I-II I

0,.3 01.5 0.7 0.19 1.1 259 IX-4j

Figure IX-ls

UREILITES : OLIVINE PIGEONITE RCHUN

I I I I I I I I I I I I I I I I I I I 0.3 0.5 0.7 0.9 1.1 260 IX-4k

Figure IX-1t

RNGR1TES - AUGITE CHION

0.3 0,5 0.7 0.9 1.1 261 IX-41

Figure IX-lu

HESOSiOERiTE - PLRG-PF STONY IRON

Figure IX-1v

IRON METEORITES : CST OCT - NR ATRX

I 1 I I I l I l I l l I ! Il 0,3 0,5 0.7 0.9 1.1 262 IX-5

spectral region 0.35 to 1.2 microns and on the same scale as the asteroid data presented on Figures IX-2. The interpretive process for each asteroid will be described below. The specific asteroid spectra chosen for interpreta-. tion were selected from the first twenty numbered asteroids for which there is good statistics on the observed spectra. The only direct characteristic which these particular bodies share is that they are among the brightest, and hence by inference, the largest asteroids. Matson (1971) and Cruikshank and Morrison (1973) have measured the albedos and calculated the sizes for several of these asteroids. Their data and other pertinent information concerning the six asteroids, for whom interpretations are made, are included in table IX-1 below.

Table IX-1 Characteristics of Six Asteroids Used in this Study

Name a(AU) e i Albedo* Radius*(km) 1 Ceres 2.77 0.08 10.60 0.07 500+ 40 2 Pallas 2.77 0.23 34.80 0.09 270+ 20 3 Juno 2.67 0.25 13.00 0.15. 115+ 7 4 Vesta 2.36 0.09 7.10 0.22 260+ 20

12 Victoria 2.34 0.22 12.60 ( -50-100) 16 Psyche 2.92 0.14 7.80 (:- 75-125) * Albedo and radius from Cruikshank and Morrison (1973) ** Estimated radius based on size/magnitude relationship 263

These six asteroids exhibit a range of orbital parameters and physical characteristics (albedo and size), such that as

far as possible, these can be considered to be a reasonably unbiased selection. 1 Ceres: This spectrum(figure IX-2a) has no distinct absorption

features and the reflectivity is constant or slowly decreasing longwards of 0.45 microns with a sharp decrease toward the blue and ultraviolet. This pattern eliminates those meteorite types with strong features (ordinary chondrites, basaltic achondrites, diogenites, chassignites, nakhlites, angrites and anorthisites). Among the weak featured or unfeatured spectra, the irons and stony-irons are eliminated on the basis of slope. Among the remain ng- petrologies, only the aubrites and carbonaceous chondrites can match the flat or decreasing infrared reflectivity. The sharp decrease in the reflectivity shortwards of 0.4 microns is not characteristic of the aubrites but is found among the carbonaceous chondrites. The sharply decreasing blueward reflectivity is a typical effect of the dielectric opaque, carbon, as described by Johnson and Fanale

(1973). Therefore the best match to the spectrum of Ceres is some type of carbonaceous chondrite. This material would be quite consistent with the low measured albedo of Ceres. John- son and Fanale (1973) have reached this same conclusion based on their work.

2 Pallas: This spectrum(fig IX-2b) is quite flat with a weak ultraviolet dropoff and hints of features near 0.65 and 0.95 264 IX-6a

Figure IX-2a

B M .E

I I i I I I I I I I I I I I I I I I I 0.3 0.5 0.7 0.9 1.1 1 CERES 265 IX-6b

Figure IX-2b

p~llap~qlI "iI I

0.3 0.5 0.7 0.9 1.1 2 PP LLA S 266 IX-7

microns (feldspar and pyroxene ?). The spectra of an aubrite and a carbonaceous chondrite are the first and second best matches, the flat reflectivity through the blue being the

hardest to correlate in the second case. However, the low

measured albedo (0.09) favors the carbonaceous material. Thus

the spectra of the asteroid can be matched either by an en-

statite achondrite or a carbonaceous chondrite, the latter case being slightly more preferred.

3 Juno: This spectrum (fig IX-2c) has a moderately positive

slope through the visible with a reasonably strong (clearly

indicated) feature near 0.95 microns. The albedo of this ob-

ject is about 0.15. This pattern is not well matched by any single meteorite type although a C30 carbonaceous chondrite or a black chondrite come closest to the curve shape but are inconsistent with the albedo. A better match is obtained by

a mixture of pyroxene or pyroxene dominated material with metal or metal-rich enstatite (E-type). This mixture would provide the best match to curve shape and features and to the albedo of the body. This should be considered simply the most likely material of a group.

4 Vesta: This spectrum (fig IX-2d) is characterized by a strong absorption feature centered near 0.9 microns, an indication of a weak feature near 0.65 microns and a steeply rising reflectivity past the center of the strong feature. This eliminates those meteoritic materials with weak featured or featureless spectra from consideration. Among the strong 267 IX-7a

Figure IX-2c

. MBO M loci Im

I I I I I I I I I 0.3 0.5 0.7 0.9 1.1 3 JUNO 268 IX-7b

Figure IX-2d

g50590gB I]a as a m I,

I1I I 1 I I I I I I.I I I I I 0.3 0.5 0.7 0.9 1.1 q VE STI IX-8 269

featured spectra only the basaltic achondrites, the diogentites

and the ordinary chondrites are grossly similar. This indicates that spectrally the dominant mineral on the surface of

Vesta is pyroxene. The ordinary chondrites have a slow in-

crease past the strong band center as a result of their olivine

content which contrasts with the sharp rise at these wavelengths for Vesta. The diogenites can provide at least one match to the curve in the visible, but the edge of the absorption

feature (and the center) are 500-10002 shortward of those in Vesta. Of the basaltic achondrites, the type 1 eucrites give

the best overall match (position of changes in slope due to features and the positions of the features, as well as the overall pattern of the curve). The identification of Vesta

spectrally with a basaltic achondrite agrees with the work of McCord et. al. (1970). The type 1 eucrite would be consistent with a shocked surface material.

12 Victoria: This spectrum (fig IX-2e) is quite steeply reddened with a strong (relatively) feature at 0.65 microns and indications of a very strong feature longwards of 1.1 microns, Of the strongly featured meteoritic spectra, none is any possible match. However the pattern is quite similar to that of an anorthite discussed in section VIII-r, with a 0.65 micron absorption and a very strongl.25 micron absorption.

While no anorthite meteorite has been measured, or even a meteorite of the required petrology reported, the trend of the eucrites and especially the existence of the very plagio- 270 IX-8a

Figure IX-2e

a I 1 13

1U33 C

I L I - I IIl I I . .. I l I I II 0.3 0.5 0.7 0.9 1.1 12 VJ CTORJI 271 IX-9

clase tich 'Serra de Mage' points out that plagioclase can be concentrated in the meteoritic parent bodies (and therefore presumably in the asteroids). Two points about the composition required must be made, a) it must be a very pure plagioclase because even 5% pyroxene would probably modify the spectrum significantly, and b) there must be some small percentage of Fe 2 + (or an equivalent transition metal) present in the feldspar. This second condition is indicative of an intermediate oxidation state (slightly reducing) in the parent body, rather than the extremely reducing conditions of the enstatite chondrites and achondrites or the highly oxidizing conditions of the carbonaceous chondrites. 16 Psyche: This spectrum (fig X-2f) is featureless (except for a possible very weak absorption near 0.65 microns) and is steadily more reflective toward longer wavelengths. Of the featureless spectra, only the iron meteorites and the more metal-rich enstatite chondrites (Abee) show this general pattern, with about this slope. The hints of silicate (plagioclase) absorptions indicate that these phases might be present, which suggests a stony-iron with anorthite as the silicate. No albedo data is available, but this type of composition would imply an albedo in the range 0.2-0.3. Chapman and Salisbury (1973) with their limited set of meteorites (no irons) concluded that Psyche looked like Abee (metal-rich E-type).

Johnson and Fanale (1973) suggested that Psyche could be matched with an iron meteorite. 272 IX-9a

Figure IX-2f

I L

1. I-

I-

. 1 - I Il L1 LI...I I II I I [1 I 1 • I I- L L I 0.3 I--0.5 t I 0.7 0.9 1.1 16 PSYCHE 273 IX-10 e) Conclusions and imDlications While the type of correlation discussed in the previous section could be extended to many other asteroids, this is not the purpose of this work. However, a discussion of the results and implications of the previous section is in order. The correlation of a mineralogical or petrological assemblage with the spectral reflectivity curve for a given asteroid can be accomplished well if the spectrum indicates a surface with a single known meteorite type petrology. Sur- faces which are mixtures of meteorite types provide less definite identifications, but still offer a usefully limited range of possible materials. Surfaces of unknown materials can be interpreted with luck and the ability to recognize the patterns involved. "he compositions predicted by the interpretation of the spectral reflectivity curves of asteroids give albedos in general agreement with the measured albedos for the few available measured asteroids.

References

Chapman, C.R.; T.V. Johnson and T.B. McCord (1973) Asteroid spectral reflectivities. Astron. Jour. 7? 126

and J.W. Salisbury (1973) Comparisons of meteorite and asteroid spectral reflectivities. Icarus 19 507 ! T.B. McCord and C. Pieters (1973) Minor planets and related objects. X. Spectrophotometric study of the composition of (1685) Toro. Astron. Jour. 78 502

Cruikshank, D.P. and D. Morrison (1973) Radii and albe.dos of nine asteroids. Abs. of Third Annual DPS-AAS Meetinz Tucson, p. 95 274 IX-11

Elias, J. (1972) Calibration of standard stars for painetary reflectivity studies. Masters dissertation, MIT

Johnson, H.L. and R.I. Mitchell (1962) A completely digitized multi-color photometer. Comm. of Lunar and Planetary Lab. Univ. of Arizona, 1 73 Johnson, T.V. and F.P. Fanale (1973) Optical properties of carbonaceous chondrites and their relationship to asteroids. J. Geophys. Res. in press Matson, D.L. (1971) Infrared emission from asteroids at wavelengths of 8.5, 10.5, and 11.6 microns. Doctoral dissertation, Calif. Inst. of Tech.

McCord, T.B.; J.B. Adams and T.V. Johnson (1970) Asteroid Vestas Spectral reflectivity and compositional implications. Science 168 1445 275 Ap I-I

Appendix 1 Meteorites Measured in This Study Keys 'Source' Collection from which specimen was obtained. (HAV = Harvard; CHI = Field Museum, Chicagos YAL - Yale University; ASU = Arizona State University; IA = University of Iowa) 'Condition' Quality of specimen (see Chapter VII). O = Perfects 1 = no rust or blemishess 2 = very minor rust (insignificant); 3 = minor rust, minor effect; 4 = some rust, some effect; 5 = rust dominates color of sample; 6 to 10 = increasing evidence of alteration, no acceptable 0 thru 2: used with no problems in spectrum 3 thru 5: used conditionally 6 thru 10s generally not used at all

Name Type Source Condition Bereba EUC CHI 1 Haraiya EUC ASU Jonzac EUC CHI 0 Juvinas EUC HAV 1 Nobleboro EUC ? CHI 0 Padvarninkai EUC CHI 0 Pasamonte EUC ASU Sioux County EUC ASU Stannern EUC HAV 0 Frankfort HOW CHI 1 LeTeilleul HOW CHI 0 Pavlovka HOW CHI O Petersburg HOW HAV 1 Johnstown DIO HAV 0 Roda DIO CHI 0 Shalka DIO CHI 0 Tatahouine DIO CHI 1 Bishopville AUB CHI,YAL 3 Cumberland Falls AUB HAV 1 Norton County AUB ASU 3 Nakhla NAK HAV 1 Angra do Reis ANG CHI 276 Ap 1-2 Name Source Condition Chassigny CHAS CHI 1 Novo Urei UREIL CHI 2 Abee E4 CHI 1 Indarch E4 HAV 1 Atlanta E5 ASU St. Mark's E5 CHI,ASU Danial's Kuil E6 CHI Hvittis E6 CHI Khairpur E6 CHI Pillistfer E6 CHI Bremervorde H3 CHI Tieschitz H3 CHI Dimmitt H3,4 HAV Ochansk H4 CHI ,YAL Beaver Creek H4 HAV Kesen H4 HAV Weston H4 HAV Tysnes Island H4 CHI Misshof H4,5 HAV Allezan H5 HAV Beardsley H5 ASU Bur Gheluai H5 HAV Cangas de Onis H5 HAV Castalia H5 HAV Collescipoli H5 CHI Forest City H5 HAV Hessle H5 HAV Pantar H5 ASU Pultusk H5 HAV Richardton H5 ASU Saline H5 ASU Cape Girardeau H6 YAL Djati Pengilon H6 HAV Kernouve H6 HAV Lancon H6 CHI Nanjemoy H6 HAV Queen's Mercy H6 CHI

Gruneberg CHI Quenzgouk CHI Rose City CHI Vernon County HAV Ap 1-3 277

Name Type Source Condition Zhoutnevyi H ASU

Mezo Madaras L3 HAV 1 Bald ?rountain L4 CHI 1 Bjurbole L4 HAV 5 Cynthiana L4 HAV 1 Saratov L4 ASU Tennasilm L4 YAL Ausson L5 HAV 4 Borkut L5 HAV 1 Homestead L5 HAV 2 Honolulu L5 YAI Knyahinya L5 HAV 0 Molina LS5 HAV 8 Shelburne L5 HAV 2 Farmington L5B HAV 0 Parazould L5B HAV 0 Tadjera L5B CHI 6

Alfianello L6 HAV 3 Aumale L6 CHI 1 Andover L6 CHI 4 Apt L6 HAV 8 Buschof L6 CHI 2 Bruderheim L6 ASU Cabezo de Mayo L6 CHI 0 Colby (Wis) L6 CHI 1 Chantonnay L6 ASU Dandaour L6 HAV 4 Drake Creek L6 HAV 1 Fisher L6 HAV 5 Forsyth L6 YAL Girgenti L6 CHI 1 Holbrook L6 HAV 3 Kunashak L6 ASU Leedey L6 ASU L'Aigle L6 HAV 8 Marion L6 HAV 3 Mocs L6 HAV 2 Modoc L6 HAV 4 Ness County L6 HAV 9 New Concord L6 HAV 5 Nerft L6 CHI 2 Pavlograd L6 HAV 5 St Michel L6 HAV 2 Segowlie L6 HAV 7 Tourinnes La Grosse L6 CHI 2 Ap I-4 278

Name e Source Condition Utrecht L6 CHI 3 Zavid L6 CHI 1 Bursa L ASU Elenovka L ASU Olmedilla L HAV 0 Pervomaisky L ASU 5 Sevrukovo LB CHI 1

Chainpur LL3 ASU Parnallee LL3 HAV 1 Hamlet #1 LL3,4 CHI 3 Kelly LL4 ASU Soko Banja LL4 CHI, HAV 1,1 Olivenza LL5 HAV 0

Benares LL6 CHI 3 Dhurmsala LL6 HAV 3 Jelica LL6 CHI 1 Lake Labyrinth LL6 ASU Manbhoom LL6 CHI 1 St Severin LL6 CHI 2 Vavilovka LL CHI 1 Alais C1 CHI 0 Orgueil Cl HAV 0 Cold Bokkevelt C2 CHI Me7hei C2 CHI Murchison C2 CHI 0 Murray C2 ASU Nogoya C2 CHI O Felix C30 CHI 0 Kainsaz C30 CHI 1 Karoonda C30 CHI 0 Lance C30 CHI 3 Ornans C30 CHI 0 Warrenton C30 HAV, CHI 0

Allende C3V CHI 0 Coolidge C3V ASU Grosnaja C3V CHI Leoville C3V ASU Mokoia C3V ASU Vigarano C3V CHI 1 279 Ap I - 5

Name Source Condition Casey County CSE OCT CHI Chulafinnee MED OCT CHI Coopertown MED OCT CHI Juncal MED OCT CHI Butler FST OCT CHI

Babb's Mill NRA CHI Veramin MES HAV

Barwise CHI Kap Oswald IA Vouille HAV

Note: The Yale specimens were received too late for measurement but will be measured in the next set. 280 Ap II-1 Appendix II

Spectral Reflectivity Curves of Meteorites Measured in this Study.

This appendix contains all of the spectral reflectivity curves measured for all of the meteorite specimens used in this study. (It should be noted that several specimens listed in Appendix I were not measured either because they arrived too late ar because the quality was so low as to preclude useful data.) For ease of comparison, all spectra are normalized to 1.0 at 0.56 microns. The reflectance at this wavelength is specified for each curve. The information given along the right side of each plot contains the following data:(example)

1 SMITH VALLEY - NO RUST 0.154 LL6 27 27 52 (a) (b) (c) (d) (e) (f) (g)

a) Caption to the curve with the corresponding number b) Name of specimen and any comments c) Reflectivity (albedo) at 0.56 microns d) Meteorite type e) Day of measurement f) Run number for that day g) Specimen identification number

* It should be noted that for several very dark specimens which were run with the 'expanded scale' option on the spectroreflectometer, designated 'XS', the 0.56 micron reflectivity may be incorrectly specified, generally about ten times too high. In this case the actual reflectivity is that of the preceeding normal scale run. WIVELENGTh(MICMON)I wnv.LarNTH MICftiW3 L i....i JIllI I I I I I I I i I I I I I I I i I 1 , aC3U)U i i i i I 1 T I I V I I 1ir11 a ,1 !i 40 ('6N I.K C3 U) 40 40 a Ii (24040 4'. a - r Cl I c I- a i 4iIa

4. tia CUC 22a1 a

r vll~ l l ll l l l i l l w w III

H fu P.n ?.A H n I r I I t In L L E n vi u I:S1. r EVcLEGtt I IClN~I HAV.fNCTn -(ICMS a a 0. i I l iI l l l If l l l f i l l I I I I I l l - I . I I I I I I I l I i I fI I l I l -iI fl l

U)U) I) U)

I a a hiti o wC3 goto ca 4 o. 21 d I

E II I I I I I I VI I I i i i i g

t-I -

SIN 43a?

Cm 0.5 1.0 1.5 2.0 2., . I L.1. I 2.0 2.5 cm on 0.5 I.n t.5 2.0 2.5 -

------~- Ap II-3 282 * WSP~CR REFLECTIVITY ISCALEDTO 1.0 NT0.56 n]CRONS1 SPECTRAREFLECTIVITY ISCALEDOTO1.0 AT 0.56 NICRONS a - I.- 1il- li t I I -

I III:.I l il i il l

- P

I I I I I l I I i l i I I i i ! 1 1 1 1 i t! t i i ,l,

t SIOUX COUNTY- RSU 0.3511 EUC 27 26 t PAOVARNINK(I - NO RUST IShER) 0.264 EUC 3 l 0 2 SIOUX COUNTY- ARSL 0.343 EUC 27 29 7 2 PADVARNINKAI - NORUST ISIER) 0.253 EUC 3 5 0

SPECTRALREFLECTIVITY ISCALEOTO 1.0 AT0.56 "ItCRONSI SPECTRALREFLECTIVITY ISCALEDTO 1.0 RT 0.56 MICRONS) :- r, p - -- A Ur i _ id[) _ ± . _ . rl .. ... jI I I I I i i I I ll I i It it I I I-t I i i

0 V1

Ln It '"

Ir U'

. 3

it 1 11 11 tit il

I I I I I I I 1 I I SI I,, ,i ...... ' ' I STANNEN WIOLE ROC 0.261 EU: IS 32 53 L PASAMONTE- RSU 0.334 EUC 26 9 13 2 STANNERN PO)ER 0.216 EU: 16 6 53 2 PASANONTE- ASU 0.327 EUC 26 10 13 3 STANNERN PONDER 0.206 EL: 16 7 53 4 STANNERN PONDER 0.219 EU: 16 8 50 Ap II-4 283

SPECTRAL REFLECTIVITY ISCALED TO 1.0 /IT0.56 MICRBNSI o - (V

fI- P

-4

I NOBLEBORO - NO RUST 0.303 HOW 27 2 70 2 NIBLEIRO - NO RUST 0.298 H1HW 27 3 70 IRVELENGT"(rMICfONtSI ,u

00 5 CM

ta a

a 2. c- (* sa to

tI gig

ga t t I I T J I I I I I I c - n

=E K-..I3 I p p

, IC A Ai

C3 CID cn C3h"'3 C-) E)i t wN

cm on

a a c2EE- I'W.(~t(C K. w w

a. ACu 9N

I- -- I Ap II-6 285

SPEC~fA. EFLECTIVITY ISCA.EO TO 1.0 AT 0.56 "ICRONS1 SPECTRL REFLECTIVITY SLED TO 1.0 NT 0.56 MICRONS1 0 - p I'. I I I I 1 1 1 I 1 1 I 1 I I-I p li F IIII AD I I II I I I I11111

An Ii

lI I I I IIIIIII III I I I i l I I I I I I I I I .I .-I II I 1 SiLAR - NO ALT 0.231 010 28 43 '72 I JOHNSTONN t4OLE R3: 1 0.370 010 15 9 29 2 SHtLA - NOALT 0.249 010 28 44 72 2 JOHNSTOWN e LE RICK I 0.339 01o 15 10 29 3 JO NSTONN COARSERPOHOER I 0.296 018 15 13 29 4 JOHNSTONN FINER PCMDER I 0.353 010 14 29

SPECTRL REFLECTIVITY ISCALEDTO 1.0 iT 0.56 MICRONS) SPECTRAL REFLECTIVITY ISCrED TO 1.0 AT 0.56 nICRONSI 71 rj M ..... _ I 1 I I I I I I I. I I I I I

1 /I B

- .'

I !I ! I ! I I I i I 1 I I i IiI 1 1 i . i iii I 0.332 010 27 14 76 1 TATAMJINE - WILEPCK M1NALT 0.1655DJ 26 36 5: 21 RPA - NO RUS 0.32 DJ0 27 15 76 2 TfTACUJINE - PN2ER NO ALT 0.423 01i 28 41 51 2 RDAO - MSRUST 3 TITA UINE - PO NER NO ALT O.45 013 26 42 51 14VEIINCTtt (MICOOMN3I ~1VT1~'T-rrrIFrrrvrr-rrr- I -Tr - I -i T- -T7-F FrIrI'r 1f I iV I I I t I 1

0 w cm mU

.T I& It .(

Lol

It f

d. ?1 11

I~j

OJ nLL.LL 1n.LI±JULL% ?(IL p~I~LL nl i n I C fl P.C WAVELENGT"((1ICM3J

~ItC

r) II' -- AA-L ILLI LLLIII I LILIII LL.IILILI . n 11; n I C, ?.fn P.l

I I Ap I-8 287

SPECTRRL REFLECTIVITY ISCALED TO 1.0 AT 0.56 MICRONSI

* aZ

:1 n I- z c)

1 - z

I NAKHLR FINER P OER 0.132 NRK 15 15 39 2 NAKHILA COARSER PODER 0.124 NAK 15 16 39 .AV%1- 288

SKfCVTM8tEF.Erl-d1Y JSOE.- TO 1.0 AT 0.56 "ICA01S)

"3 -

III I I 111111111I~ lit 1 2 55 I R~4~ ~OS - 2.2ss &qc 1 2 se 2 RNCRRCO. I'S - 41. KT a. :51 filic 1 3 656 Ap II-l10 289

SPECTRRL REFLECTIVIT ISCRLED TO 1.0 lT 0.56 M]CRNSI I ~o o

K i -

LT t -

I I I I I I I I I I I III I

CHRSSIGNI - NO ALTERATION 0.353 CHAS 2" 12 31 CHRSSIGNY - N ARLTERATIN 0.359 CHAS 27 13 31 Ap II-11 290

SPECTRRL REFLECTIVITY ISCALED TO 1.0 AT 0.56 MICRONSI

S 1111 11111 1111111 I

Lno

i I i I , I I I I I I I I 1 NIGV-UREL - RRRE ALT o,098 UREI 1 4 82 NGVG-UREl - RRRE PLT 0.088 UREI 1 5 82 NGVC-UREI - MINRB RUST INSI 0.078 URE 2 16 82 NOVO-UREI - MINOR RUST XS1 0.707 UREJ 2 17 82 NOVO-UREI - MINCB RUST IXSl 0.707 UREJ 2 18 62 W.VELNCrZH -IM. ,ONS i II 'l l IT I l I l lfIl lI I l 1 l I ll l I l MRi 7w t- iw

UP V! a it) a Ca C;c;C; 9; 4a N ZI d 9 a C I-

a a m W zggzgII,, n1. tz '" 15

1 0 # I

.L L. .L I I LL..L.J. LL J..L I.LL L IL. L..LL. 5 1,0 1.5 2.01 2.5 0.5 1.0 1.5 2.0 2.5 WiAVELENG7tI(ICIONS) IVELLNCT" mtrnIftiN3I T" rrr"Trrl-r-T-r-lr-rT -- -- ili I - Ir I i ir 11r- i T T 1"1'1 1 1 11 1-1I I I a

a0 C;Gi; In in 1S4in

Slin 0 W11i iiiii~III1331 1.

1 - i

a IM C~ 2. C

LL .I -L.L.I.LL.L±.L. I I LL.LI I I I I I L L.L.L.LL.Ii..LL.L _.L.oL. 1 i- LI iLlLIL.L. iLiL r,5 1.0 1.5 2.0 2.5 0.5 1.0 1.5 2.0 2.5 IIVELENGCTn(IICIIlS - W-VLrtl.rIr rMlrrN'rT r T T -T ''F T v (V - T'TT CTr I T T I T Irf- rf i IC1 u , a rC4 D t C, cuu c;

V 6-)1, 0'e, 1.-C, a a I- .1 21.c

I? T.Ic

cm I |. aa

(3,**113,1 I, i, (I*lii 8.C I II 1,i (IIIii 5f (III

u za toI. 41. LI a ~ I I I I i i I I I I I I I 1 I I II I I 1 I- S I I L I I I I I I LI L1 LI I L 1 1 1. L ~i- L.A L L L 0.5 1.0 1.5 2.0 2.5 In(1 .P ..Cl SI ,5 niONf. a I( hi h 0A. L0 TH MIcfloN3 0 iVELENGTH IMICitNSI I-.- I I I I I III I I I I II I I I I I I I I-1 1 I I I I 1-I i - r I I I I I I I I I I I I I I I, - N

I..

I ma I.

- 3(

I I I I I I I I I I I I LI I11 1iJL.L I...L.I..J.... IYli I I.. I II-i-i. .. I I I I I I I I I t t i l i t 0.5 1.0 1.5 2.0 2.5 0.5 1.0 1.5 2.0 2.5 IVLNGTh (MICIN3SI IVELENCGTHMICRONSI I' 1I'1 I rl1 r1IIIIIIIII 17 1- at

a Sec 3v. I WV)e i .

a a

.c ii I I

o.C a I a I i i ii i i i I I a I I i I Ia I I 0.5 1.0 1.5 2.0 2.5 0.5 1.0 1.5 2.0 2.5 MVELENCGTNIICMNSI MVELENGTH9mIC4lNSI

m, rri a a a ft !;

U) 4 ccQ a

I I I.II I I I I IIj l i _l1I I lI I I I - 2." iiccp

0.5 1.0 1.5 2.0 2.5 0.5 1.0 1.5 2.0 2.5 IVELENGThINICfMSl I dIVELENCGT(MICONSI IIIIIIIIIILUl lll!ll!illllII r- ''

dd a I r-2

I a l i l l iTI I I I I I I ITlTl ! ! ! 4arr I 0.5 1.0 1.5 2.0 2.5 0.5 10 1.5 . 2.5 tIWVELECTrNhICNM I WAVELENrTHINICAN

a-3 C0 I I a 6 Li I1.1 44" I ;~I-

CJ-; U, V a Ire9 iif

il 111 III Il I l ll I1 I1 I I II IRIS III 111111 111I II III II II -v 0.5 1.0 1.5 2.0 2.5 0.5 1.0 1.5 2.0 2.5 MIVELENOTH(IICrIMI NAsVELENTHIICRON3I in II IIIIIIII IIIIIlIIIIllI II III "IId 11111i 1 "'1 i Il l I aa iiI

f~a 2.t It i& II a I - rr,Ifm ac -L129

I a. c; C IIt;U) aa i a a ai a a a a a a a a a a_ a a -I

ma.la sl a a ,l ai a i alalala lamlalla 0.5 1.0 1.5 2.0 2.5 0.5 1.0 1.5 2.0 2.5 IAVELENCThm(ClHa AVELENCTha(NICrSl rr- I I I I I I I I I I I I I I I I I I I I I II I I

N N I"r a as

2. - 2. iil aPa,

•,,,,,a a . 0 1. 1

eii - N - ,,

a. 0.1 0.5 1.0 1.5 2.0 2.5 0.5 1.0 1.5 2.0 2.5 ACVELENGTH(HICROt4 1WAVELENCTH(1IC1NSI I I I lI l 'l l l l I'l .l.Jllil l l l I lIIl I N(a co N g (NN C-4w40cu a cm C l. zn C;t ' i"

PRCi a i

-NSI 0 il

c I I I II LLL

0.5 1.0 1.5 2.0 2.5 0.5 1.0 1.5 2.0 2.5 WVELtNC'tHI IChiNI WIVfLENCGTIlIEIMtNI

r a a cm

I 2. 0.L

1.'

aE -1

III - N S I I I I I I I I I I I I I I I I I I I I 0.5 1.0 1.5 2.0 2.5 0.S 1.0 1.5 2.0 2.5

------~ Ap II-1l3 297 SPECTRL REFLECTIVITY ISCALEO To 1.A AT 0.56, InaNS SPECTRA EFLErTIVlTY ISCLm 1.0 AT 0.56 NiaiS) 0" A) a AS C-- B

I...ii1~il i 1 11 11 1 I~

I') 0

ln) U'

liii 11111 11111 1 ...... I I , I • I ERNOUVE COARSERPONDER 0.186 6 15 1t 30 I SALINE - ASU IRUST) 0.194 HS 27 26 5 2 KERNMUVE FINER PONDER 0.171"6 15 12 30

PETLr REFLECTIVITY SCLED TO 1.0 AT 0.56 "IJCRONS) PECTAf. IIELEICTIVIT1Y ISC.D TO 1.0 T 0.5 -IICRONS

Ln I 1J tit it I I I I i

a .. O

S, 1 1 l i l t I l tt I lI t , t LANCON - M lX Vvf RUST 0.242 hS6 32 26 I 0.JTI-PENGILON FINER PNER., 0.147 16 171 22 2 LIACON - VV RUST 0.295 "6 4 35 26 2 OlATI-PEIILON COARSE PONDER 0.122 "6 172 22 3 LANCON - VWVtfST 0.317 "6 4 36 26 WAVCLfl4GT"(?4ICRON3J WAVCLENG7"(MrI T'r7IONTI 4n MI Cs

g t~. (4 Cs

C2 a C* ca Cs Cs

Cs 5- I-

I I IILIP111111 LI i4 I. 'I, .1~* "4u t I II) tI9 ; 0. L11LLLLAl1W I 11 11-1111LLL .5 .0 1.S 2. 2.5 U.S 1.0 1.5 ?. 0 2.5 C3C UMMttIW*IIN' rT-rrrTT FurrriFo r-rrrrri i -vur ri ri-v rri ri vi i i. iFi r r v I-A

C III -r II) vs a *1 44 Cs 4, 5- IL IC C a

Cs 2. w r LI :m w d vi6.I vi

5- w4 qJ) I- 'Iv ~i 1 4L 1. I I.(

W4I I.s4-2 I - Cs n a1Ilaa ii a4J W-t IWJi llL llLL± 0.5 1.0 1.5 2.0 2.5 0.5 1.0 Ls5 2.0 ?.S Ap II-20 299

SPECTRAL REFLECTIVITY ISCALEO TO 1.0 AT 0.56 MICRONSI c - ro i I I i I i i I It I I 1 I

-1

Lu z I)

-1,, , m,,,,

ZHOUTNEVYI - PSU 0.358 H 28 2 9 2HIUTNEVYI - PSU 0.3468 n 28 3 9 WRVCLCNC~h(I4ICflONSI WAVELECG" fMJCAONSI SSE SES IEIISSSS11 111 1 lITi 42 ca a g

InMnE 3 C; rmr-

E 9- En u U. ~ 2.C

21.(

0.5 1.0 L.5 2.0 2.S NAVELENCYQ IN lilfN HIVELeNG7" (NICRONSI SEE Eli 5115115 EEESEESIS ISlES IS

XI 4

WI ~ 2.c I aa

I a

v 9 I.C I

gew a as a a a a a a a a asa LIllilIl is asass a. U. 0.5 1.0 1.5 2.0 2.5 0.5 1.0 1.5 2.0 2.5 WIVELENGTNHCICN Il WAVELENGTh(MIGCONSI W

cmc *aS (a VJ _-r .7 0~c

hr) k"t

I - c wz t aI aI a1

H04 H cc-E f a~

I I I I I TT I TI I I I I I I I I I I I I I I t 1I .I-. I I .L L. LL. LIL L-L.L.J. L.L. LL. I-.-L L L..LI L. U.S 1.0 1.5 P.0 2.5 SVE"' NG (Mcr II r rCTN N ~NN -rT rrrTr' r- r, rrr r r rrrr r r, r i cm CYc rQ rr m 1i oo!-INff . I-" i-;-4a a a-4 ar I. 'C a aam a*CM~~ a-.I te r ccdd a ha

0 . ie U.

I .C - N ~

-WI4 ha

Ytr A" Ytr

CYC

LLLLLL..L.LLI 1.0 I 1.5 2.5i JLII I -- 0.5 2.0 0 . L-L.LL.1 IL LLL 1LL.L.-_.L LL 0.5 1.0 1.S 2.0 2.5 - I II L~ I I

1,IEG H1" InCtM1F31 1 1 I 1 1 1 IIRVELENGTH 1 1 (MIClONSI 1 1 1 I I I I -rr -t- -r--rr i T -rT-r, , , , i 4r-

K GCZ N N r,C3 Jo

a N N i .C

ccc / -1

i.C

a C

I I in

L L _l .l. LI. _.L .L--L-..LL L.L AAL.. J1 -.L l i I 1 lIi I 1 11 1l m !i I 1 I i C.5 I.C 1.5 2.0 2.5 0.5 1.0 1.5 2.0 2.5 I VELENGTH(#MICflN3I

c; ~i1-.4 c& c; -

a a Ca I' * cc-2

emr tj~ I 2. fiI. a

0. C 0.5 1.0 1.5 2.0 2.5 Ap II-24 303

SPECTRL REFLErIVITYT ISCrLED T 1.0 AT 0.56 ICRCNSiS SPECTRA.REFLECTIVITY ISCfILEDTO 1.0 AT 0.56 "JCR0NSI 1 7* !D

I I I II I I I I i1 1 1 1 1

L - Illsllillllllllll SII I!II I II I ItI I Ill 1 E a i e l s t s r I SH.ELUltE RRRERUST o.2q0LS i18 11 tI KNYtIUNA WOtLLEROCK .1"72LS 15 22 32 2 SHLaRNE RARERUST 0.221 L5 18 12t4

SPECTRIL REFLETIVITY ISCRED TO 1.8 iT 0.58 MICRONS1

l i I I I. I I1 1 I I I I I I I l I l ~ I ~ l l ~ l l I I I I fMOLLA POCER. V. oDRi. R 8.131 L5 17 32 36 Ap II-25 304 'PErCTf. RtFLETIVITY ISCRLEO1 1.0 AT 0.56 MICRONS) SPECTRRLREFLECTIVITY ISCALEOd t1.0 AT 0.56 MICRONS) a I

P F

JI 1 I lI'1i i I I I iI i'1 - I I I 1 I | ! | n I I n ! I SEVRUJKOV- (I RUST INS) 0.081L 8 4 222 I FARINCTON V. DARK NO IST 0.106 L 15 1 24 2 SEVFiUKOVO- N RUST IXS1 0. LI8 232 2 FARMINGTONV. DARK NO RUST 0.108 L5 15 2 24 3 FARMINGTON- NO RUST INS 0.100 L56 3 31 24 1 FARMINGTON- O RUSTIXS 0. L56 3 32 24

SPECTRU REFLEICTIVIT ISCLE 10 1.0 ATI0.56 MICRONS1 SPECTRA REFLECTIVITY ISCALEDTO 1.0 AT0.56 MICRONS .- II o _ 1"1111 T I li i w Li I Tm1rr--r1T ~If

in cL

o 0

z u1 f

•n - rFl- M

uliLn cnm

I ,,I III II I I I I I I I I PARAGOULO WHcE RtCK.V. OR 0.096 L5 15 29 44 ROSE C1 - h R X IN) NO V RU 0.0 65 9 5 1196 2 PARR;,ULO PiNOER V. D K. 0.061 L5 15 33 44 2 ROE CITY -RWRIXiXS) NOV RU 0. i 3 5 1295 3 PARACGOULO - NORUST INS) o.Or L58 3 19 44 4 PARAGOULO- NO RUST )1SI 0. L56 3 20 44 5 PARAGCULU - NO RUST IX1 0. L58 3 21 44 Ap II-26 305 SSPETINAL REFLECTIVIT' ISCILED TO 1.0 AT 0.56 NJIOl S1 WPCTA.L fEFLECTIVJTT IS1.i TO tl.0 BT0.56 NJCONSI n II

-d z 1 11111 1I11t. 11 lll'1 jm t I

I APT IIUSTEDI 0.137 L6 14 10 12 I FIANELL0O 0.292 LS6 t 6 10 2 AILFlANELO 0.273 L6 14 9 10

SPECTRAL RFLECTIVITY ISCALEDTO 1.0 AT 0.56 NJM"S) SPECTRL REFLECTVITT ISC.L:E TO 1.0 AT 0.56 mIJCRoS I') J ' . 5' IIII JillI 111111111111

SaLr

n, U' I- - fu rUi

, i ,i I I I l I l l I I l l 111111111 ~~~ ~ liii,,, I MAIKALE- N IoX Nu UST 0.202 4 20 37 I MoNUVER- Se rUUST 0.314 L6 4 16 66 2 MJILE - NORUST 0.32 L6 4 21 37 Ap IZ-27: 306

SPECTRALfEFLECTIVITr SCRLEOTO 1.0 AT 0.56 MICRONSI SPECTRA EF16LCTvITtY SCRED TO 1.0 AT 8.56 IJCROnS) nr --- S n I I I I I II 11 I T I I I - I

2 Sif I

AI I I I I I I I I I I I I I I I t CRASEO BE NAYe - FINER. NOR 0.33 L6 26-31 W 0.328 L6 25 11 2 CRDBE6OE MAYO - COARSER NO RU 0.319 LS 20 31 14 2 BRUDERHEIN - ASU 0.L3 L 286 11 2 RIJOERHEIN- A~iJ

SPECTRAL REFLECTIVITT ISCALEDTO 1.0 AT 0.56 MICRNSI SPECTRAL REFLECTIVITY ISCALO TO 1.0 AT 0.56 nMCROlNS p n,

- , , , , , -, , _ . . . . I II I I 11 I I I

C U'

0 I

fub IN t0 0

I ''liii III lII IIlI 1 I I S I I I I I I I t I S CHNT ONNAf - ASU 0.161 LS 26 22 20 1 USCiw - Ps"4 STl 0.206 L6 4 27 75 2 CiTriNINAY - ASU 0.167 L6 20 23 20 2 IUSCnOF - nfit ST 0.371 L6 4 30 75 3 aUSCnOF - NOq15. 0.366 6 4 31 75 bIRVELENGTtItt41CiMNS1 3%. c.J 1:1q= cJ i ago a) aD - C"' ;j 0 = C !! to to to do

K to to go co to to la cc to in a n 0t r- a ca az az 3 a C* toec

C3 I U I I.- mit I S 6 "00 0. C - tv Sn 0.5 1.0 1.5 2.0 2.5 t.' H &4AVELENGTIIRIMNSI H tIAVELENCThIIMN'l

Fl: to rn 4D to to CY y. cm toI to W)a a 40 a0

ot Sfl - 0 g~ggggj j ggjg gg gg g I 10000 ~ C

a Wi L to g ~~g~ gI I i A A I I I I I I1 I g I I I I I I I I I i 9". p.. I- - (SU

-"111iiillllllllilliilm C3= 11,1,11111111IIIIIIIIIIIIIIIIIIIIIIIIIl oo o

ILh 111 H ~

(J , 0.5 1.0 1.5 2.0 2.5 0.5 1.0 1.5 2.0 2.5 Ap II-2! 308 PECTRIALfEFLECTIVITY ISCRLEDTO 1.0 AT 0.56 $"ICRNS1 SPEC"ThL REFLECTIVITY ISCLED TO1.8 AT 0.56 IJLCONSI Wn) C .- 11 11 i11i I i I I I I i - • .... ";- ...... ':"-.. _ _ _ _ F

En- :LP°kVI IA - i

t- -u I' I I I I I Il l I I I I I I I -,,, I

I KIQ5'RAK - ASU 0.27 LS 25 14 16 I CIRGENTI - VR RUST 0.266 L6 2 13 ,I4 2 lvfIISHAK - RSUJ 0.216 L6 26 15 16 2 CIRCENTI - VR RUIST 0.26l L6 2 147t1

SPECTR AEFLECTIVITY ISCRED TO 1.0 T 0.56 JCAOnRS1 SPECTNLREFLECTIVIT Y fSCL.ED TO 1.0 AT 0.56 MICRONS) :. , -2 c ' "/_i I II I 1 1 i I i I E I-I E

Lnr

til! i ii i I I I I I I I I l I lI I I t HCL ~CK FINER PCisER. RAR 0.269 L6 17 19 2"1 I L 'ICLE POiN40. UST 0.135 LS 17 21 33 2 ttCI.SAK CORASER P"u'tU. 0.?I24 L6 11 20 27 Ap II-30 309 SPECTRIIREFLECTIVITY ISCR D TO 1.0 AT 0.56 MINCRONSI SPECTFL REFLECTVIlYr ISCULEO TO 1.0 AT .56 flCRONSI ..-. , - - -) I III III 11111 1

roF.' - Go - Vn

I Ln _ z

tU L

I I I IfII II ~, ...... I ," ... , , , , , , . I MSCS FINER PONDER.RA 0.329 L6 17 8 36 t LEEDEY- RSU 0.317 LB 27 18 1 2 MCCS COARSERPONDER RA 0.245 L6 179 36 2 LEEOEY- RSU 0.319 LB 27 19 1 3 HCCS MIXEDPONDSER R 0.303 L6 17 10 36

SPECTRNLREFLECTIVITY ISCLEO TO 1.0 AT 0.55 M]RO1NS1 SPEVTR. REFLECTIVITY ISCILD0 TV 1.0 AT0.56 MJCROn .-- ..- . c= .I . .. . , _mm,- .. ... -- i " " It- SI I I I I I I I I I I I I I I I I I I I I I I I I I I I I I tLr

fur

-Ut

U i -i ii

I ! - - ---I i i I t l I1,I III I I 111 I 1 L 1111111111 ' ' P"CE. t 0.341 L6 17 28 34 MOCDC PONDER SE tR 0.264 L6 1131 37 I mtRIN FINER 2 MARION COARSEr P"ER. SO 0.333 LG 11 29 34 IEAVELEWThthIIC1fq") WAVrifWu#T f4ICI~a" ~T~rT-rTTi-T iTi-i i iTr--- r i ( r- rfl

to 0

aj j

-LL. I

0. S 1.0 1.5 2.0 2.5 (1.5 1.0 1.5 2.0 2.5 HWVMENGTH(HICM~:l J .c IAVELENCYh(MIrAI'OS , I I I I II I I I 1 F 1 r1F,1J -rTT~rl- TT-n -rr='-TT-T-r7T-t--r-F

4--

~*LWLLLLW L.~L L4.-LJ ...... J.1..

- I

..LLLJ. LL.LLLJ. -J I I IL 1L L 1 . 1 1 t Ii 1.0 1.5 2.0 2.5 1.5 1.0 1.5 2.U 2.!) Ap II-3: 311 SPECTRL REFLECTIVITY ISCRLE0 T 1.0 RT 0.56 IrrrSNSI SPECTRL REFLECTIVTY ISCALEC TO 1.0 PT 0.56 UICRONS1 I') It IjI I I I I I I I I f ... NP . if FI I I I I I i t 1 I II I1 T

-IF 0 KN I -1 0 I- -1.-! I

i -' 1~ F rJ oi 43I7L

-A)

i I II 111111 l1lt, ...... Illll JWLLLW, I I I I I I I I., I UTRECHT- NIXED MINORRUST 0.73 L6 4 16 0 I ST. MICHELWHOLE RCI 0.335 L6 1l5 9 2 UTRECHT- FINES MINORRUST 0.323 L6 4 17"0 2 ST. NICHEL CSORSERPM"CER 0.301 LS 16 1 9 3 ST. NICHEL FINER PCoC-ER 0.321 L6 16 IS 9

SPECTRALREFLECTIVITY ISCALEDTO 1.0 AT 0.56 rMTiEN SPECTRAL REFLECTIVITY ISCLE2 TO 1.0 AT 0.56 NICRONS! r a :1 !Qb I. - _ - .. .. . ,, I i liii 111111 11111

. 1 wt

4- I - LJ

7-t T'J iii . 1 I-lf t -, I- F - at I I I I .4. I -I I I [ I I I A, *I i ,, , I , tal 1 111 1 I1 S i l l i " = 0.296 1.6 17 72 it AVID - 4 ROXVR RUST 0.230 LS 4 10 6S: 1 T~URINNES LA CROSSE- R4E 9LS 0.31 L6 1 16 72 2 7.AVID - VRRUST 0.29 L6 ' :2 61 2 TCURINNES LA GROSSE- R9;R RS 3 WVID-VR RUST 0.287 LS 4 13 61 Ap II-33 312

SPCTRaL AEFLECTIVITY ISC.ED To 1.0 AT 0.SS NICRONSI SPECTRW REFLECTIVITY ISCLEO TO 1.0 AT 0.56 NJCRONSI

o ,

- l ll 11llil 1 111 11 I lIt I I I l I i It I I I WL.ECILLA O.187 L 15 21 43 I O3SR - ASU 0.305 L 28 11 tI 2 OURSA 0.306 I 28 12 111

SECTRk REFLECTIVITY ISCLOEDTO 1.0 AT 0.58 NICRONSI SPECTR REFLECTIVITY ISCRLEDTO 1.0 AT 0.56 NMCRONSl :- ...... w.... .,- _ - , I . • . ... FTI I I I I I I 1 1.1 II i F II _ 1 I I I111

0, 0, U'1

N U' U)N

lii

I I t I SI I I I1 I I I I I I I Ii I I PERrISKY - ASU. RUST 0.258 L 21 26 22 I ELE8OVKFr - aRs 0.311 L 28 27 23 2! ELENGlVI - ASU 0.30 L 28 28 23 WRELENCTI CICION3C WAVCLEYCT(ICnlONS) --r I i i w i rr-rirrT-r-r ---rrT-r SI r i r'FTT-rrT-T r'rrrrTTrrr-rr

?133 '3 to C)4( cza aC 40 10, Iw

U) it-41.1 aIn,

Inc *-

cua rc'11 r" Wr'O I ,

a~ain mazaf to

L LL-L.L-.LLLLLL.L.LALL.LLL.L.L.LLL x b 11. LL. hirt O.C 1 L.LLL I.LL4 1J1.L I IL.L .LL. L.LLL.JL.._.L. 0.5 1.0 1.5 2.0 2.5 I4 0.5 1.0 1.5 2.0 2.5 hi ti i iFit il11111 MI t t,I; fI1 I l"I l Il I f fI! I f 1(1IN',I1 1 I I i lI I Ii WIIVfl.NTII rMICf" ONf f -T-rTrT 'TrrT rrrFPT'T"'T rl' IT"'rTTr-T-T- i9" or I, m2C

4')4 I 4. ta In .4.2 ~ In..,it ,In w a at 4-) 4 I, 4.4 C 2.C rr

a d E

.p-44r

L .) /S~~I U, - Y. 4c IL

- 4. C! C

-1 -L. 4L LL. LL. LJI- LLL .- L L 1.0-L L .A1._LJ A. .1S .C . 1. 1.5 P.0 , . 0.5 1.0 1.5 2.0 2.5 WAVELENCT H]MICM~l WAVELENGTH(MNICONS) I,l 8lIC2 G i 9 Nu U) V) WflV) in I8 a a In89a aM! a aCD a a -.- - w , S2e.0 1-j 8. = tiI828

84 1 IZ IMIx I Uj IQ I a R (n~~fill. t PCLc ~I.(r l II I I II wSS i 1 1 N~rr '5 * Cdd

g C2~s

I I I I I I I I I . L111 I I I LLI 1 1 I I I! I I I I 0.5 1.0 1.5 2.0 2.5 0.5 1.0 1.5 2.0 2.5 WAVELENGTH(WICrIONS C C VEl T I ' II(MICrON3 S T ''f r Iu'i f t i so " "t t ' luv u u =

8 ago 8 Nv a aRa a r- cc

8 1. ---

-N

- I i-*,*** I

I I I I I I I I I I I I i I IL lI I. a. L L i III I i I I I I i I 0 . l I I I i I I I II I II I 0.5 1.0 1.5 2.0 2.5 0.5 1.0 1.5 2.0 2.5 Ap II-316 315 SPECTR REFLECTIVITY SCRLED TO 1.0 AT 0.56 ICRONSI SPECTRA REFLECTIVITY ISCRLE. TO 1.0 AT 0.56 MICRONS1 p " , .a *- !0

I I I I .1 I I I I II lIi I II I

I III I I 1 I I l1I Ii u 0 0U' E Fn 6

Iu M

(11111111111111111.OIIJ$- OXS1 RS S BENARES- FM PON SOMERUST 0.231 LL6 4 19 16 i SOKO-6RNJA WIHOLEROCK TRSI 0.255 LL4 15 17 S9 2 SOKO-ANJA FINER PONDER 0.183 LL 15 16 49 3 SOKO-6ANJA COARSERPONDER 0.165 LL 15 1 49 4 SOKO-lANJA WICLE ROCK 0.214 LL4 15 20 49 5 SGKO-BANJA - MIXED NO RUST 0.272 LLY 4 33 19 6 SOKO-6ANJA - FINER NO RUST 0.296 LL4 4 34 19

SPECTRALREFLECTIVITY ISCLED TO 1.0 FT 0.56 JHCRONSI SPECTRALREFLECTIVITY ISCIEDOTO 1.0 AT 0.56 MICRONSI

a Ln

a -

:l lxI

1 1T11TI T111 IIII tTTItIl III I I I I I I I II I i ELIVENZA 0.369 LLC 11 6 it q. SILR FINER PODOEA. MI 0.296 LL6 17 3 21 0.366 LL 14 2 6 MI 0.222 LL6 174 21 2 OLIVENZR 2 OIURISALA CORRSEPONDER. 3 OLLVENTI 0.366ULLS 14 3 6 3 OnRFSAR FINER PONDER. MI 0.267 LL6 17 5 21

- T -

Ap II-3 316 SPECTRAt REFLECTIVITYT ISILED TV 1.0 AT 0.56 tilRONSI SPECTML fEFLECTIVITT ISCfLE TO 1.0 AT 0.56 MICRONSI - 0 ,- ,m L C- 0 - , . . . .w .. w , ---, I I - I I I I I I I I I

ill11 I ~liI III 1111 111

-ft

.1 I I I I 1 tillIIlllIll

I MINNBHOO- NORUST ICG Ol 0.227 LL6 28 31463 I .JICR- V RARERLST 0.234 LL 3 2253 2 NRMNBICO- NORUST JCIGST 0.256 LLS 28 35 e3 2 JELICA - V RE RUST 0.257 LL6 3 23 53

SPECTiL REFLECTIVITY ISCOLECTO 1.0 RT 0.56 ItCRONS1 SECTWL REFLL TIVITT fISCL TO1.0 AT 0.56 14CROS1I o :. . = . lIlt1 I I I II 1 I I I I I I

°~ t I I I I I i I - ;,-Li:

Li Li j

, I 1 I I I-- I I l I I I l I t ST SEVEMR - WREX Sr FUST :.ii? LLE 3 to 514 I LAKE LRB'TRINTM- ~S 0.?04 LL6 21 20 2 2 LAKE LPSRINTM - RSU 0.209 LL6 27 21 2 Ap II-38 317

SPECTRRL REFLECTIVITT ISCALED TO 1.0 RT 0.56 MICRONSI o - ro _ , ,

0 Lni

n LTiS ed

11111111111111 1 t VAVILOVKR - VR RUST 0.267 LL 3 2 59 2 VRVILVKR - VR RUST 0.271 LL 3 3 59 &AVE1INIGh4 fMI1CMMtI I4AVCLENCGtI(NICMlONI i rI co 5 1 1 I IIII III1 1 1 ~ ~ III I I I I I , , , F 111 ~1 r-T-

41D C& CA !!4~L 01- - ;I c; i 42

... 9 00C 0 oem;C;C;C

in n1 Icil

L I I uL". a. usjagiqiii~s aaja i f" LO LLI4J VI-ALA L-LLIILAA-J

ILI,

3~7~rr

I0 2.0 i

C -1 Ral n LES

In 40 do ea in

ea - Ui Li 41UL

2.0 2. . 1.5 . 2. 2.5 ~ Ap II-4C) 319 SPECTAL RULECTIVITY ISCIaED T 1.0 AT 0.56 nJCRONSI ETR MEFrLE"TIVITY ISC.E T 1.0 AT 0.56 IlJCRONSI ., _ . . ... -* - - I i I I l .T I I I I I lll l l " - - -I l l -l i - i l l l l l .... --

-M % d

S11-- 111 I I I I i i I I I I I I I i ll ti lil t I IIit I mUCOTA- INS) 0.0146 C2 5 27 80 I UR.ilSCN - i Rex fXSl 0. C2 5 3241 2 NOGCTA- IXSP2)1 0. C2 S 28 80 2 JRCi S.N - tNR9OX1 IT, 0. C2 5 33111 3 99CHISiN - IxS) 0. C2 5 3741

SPECTRA PIEFLECTIVIT fSESCLEO 1.0 AT 0.56 MIWCRS) SPECTiL RELELIVITT 15CArED To 1.8 AT 0.56 nICr"NSI a - 111'' 1ili11111i 111

a - 319

I tA I 1 111 11 111 E l l 1 i t i rt l i l 0.086 r30 1 12 30 L MAWr - ta, fcE 0.04C C2 2 6 26 0. C2 2 9 26 2 FELIX 9Xsl 0. £36 1 13 30 IXsl S .ftPkAr - AS fXS" 0. C2 2 10 26 3 FELIX 0. C30 1 14 30 adIVLtwG~t(MICMW~1 to (aw w

Ct; V) C3 a ar a ati ocq

I3 a a a z~c0 T a in t L I II '1 I I I I 1 II .I i I I i I I I I I I I I

-) 0 i 00 cuz !s1t

- - r-

~atoit t . CL M 3' ttlllllllll ll a aall alaallzalallalal SIIIII11 11111111111111 11111 W. - N 0.5 1.0 1.5 2.0 2.5 0.5 1.0 1.5 2.0 2.5 M VVELENGTi(1I1CION1S IT-Ti Ili Ill T-1V -11 1 - ifTTT Tr-T

*2 4',to ~ -0. to14- 10 ry . ru!0 LOG to S M d L @2I [email protected] nI c; C; C; t; 0000; a a a 2T1- .- I- I.-

s Eia

gggg a a a a NEC

SSfi ~1E1"i1~ s = =3 U-

~I 9 _j j.jI I IsI I I I .L.L.2LL .I L 0.5 1.0 1.5 2.0 2.5 U.5 1.0 1.5 2.0 2.5 WAVLENGTT (ICRONSI VIrELIENCT11IICMrtNSI I I I I I I I I I I I I I I III I I I I II i rITIIIT 'TII I -T-T r rrr - a-rN~V) z U)0 w w w £0to 0in 00

edddEi$$WV a a a a aa

I-- coo

a- a- r s) V a --- aI -I N I h U H a i a a a t a a a t a p4 L_.I1. L LLL.LIL.LLJ LL iLLLi.LL CUCM 0.5 1.0 1.5 2.0 2.5 0.5 1.0 1.5 2.0 2.5 AVELENGTh_WICIMNSI WIVELENCGTN" MICflONsII -TrFT-TrTFr T"TV~"TFTT l T 'r" -I-rT-Ia-r'FT rI~rTl T r IIII II i IIIIII IIIII lII!I

a- - , 1

6 - c~a;; - I-

a a a a a a a

t 2.C a E r. xc ac a I,,,, I;;ii.- N WI ) £0l Il- Cb C3 caEcmCO2 6; s sE I-I ft o w o, a-' --

at a C3 La

- aL _t .. 1 L LL.. LL I I I I IL.LLI 1LI 1.._LL._l C.5 1.0 1.5 2.0 2.5 0.5 1.0 1.5 2.0 2.5 wELCNG7 rnIOtr WA~rLENGT"(MICAMS r, r Vr-r- rrT-T-rrm vrir-rrrriir -r7--TrrrTT-iTrir-rr--r rr r r r rTrr I;;

cfl 6O

0 S~ S J('J 0 ~ I- B- ccc

B-

t.c

G S

n.OL L IA L LLLL.LIA iE .LLL1I.L L. L-I- -Lid. J,4 L L L.LL1 4-J..-LIL-- -iLI -4LLL1 1.1n hb 2.0 2.5 WP~rLENr,7"tMICrMeII

arr

L. I-.

a ~I.C

.la 4:

0 .L- LA-1 L-1-L-1- I. Li LLI1 L- I. 1LLJ.L.I..L 0.5 .0 1h 2.0 2.5 Ap 11-44 323

SPECTRALREFLECTIVITY ISCAED TO 1.0 AT 0.56 MICRIONS SPECTRAL RfEFLCTIVITY I1SCELD T0 1.0 AT 0.56 ICAJiNSi a ," ..

I I I I I I I I I I I I I I I I i I I I I I I lt, i', pt a -

o -

p - _l I- rq ( ju

n UI

111II 1 I IIti I 1! 1I! SI I I I I I I I I I I I I I I I • l " "l i l l .. ... S SBUTLER- CUTSCRTED SU 0.289 FSTO26 6 21 I CASEYCOUNTY - CUTSUR 0.168 CoCT26 15 "76 2 BUTLER - CUTSCRTED 5LP 0.296 FSTO 26 5 21 2 CASEYCOUNTY - CUTSUR 0.161 COET26 16 76

SPECTRA REFLECTIVITT ISCLEO TO 1.0 AT0.56 ICROWS SPECTRAL REFLECTIVITY IsCrED TW 1.0 NT 8.56 MICRONS -IlIIIII II i1 1 111111 LI-- IITI T I I F Til I I T

_1,,, 1,,,,,,,,,, o

lIL A I I I II I I ARB'S MILL - CJT St5R 0.246 NRAX26 13 5 I CNULRFINNEE - CT SU 0.173 ":'T :5 1, A 2 6I8B'S MILL - CUT SUR 0.232 IIRAX26 14 5 2 CLAlINFiEE - CL, SLR 0.191 IMC:T26 12 91 iWAVLENGIT (IIICISI AVLCNTGt 2eI INSI -I rr'T-rT-T T- -r-FFFrT-TT-r-r--r-lT"rT - II-r I-rII- T"r rI--rT r- r 'T- TVrT r rrT-- on it ci AG t r,t; X: to v c" C; di C C)

' 2.a1 2.C C) C3 '4 us a a

IO - N .C k......

CUi " 0O I I li I 1 I I I I I tI I l I I I l I I I I - I - BEuN F l I.. I I .5I L I aO.a--l A I I I LI 0.5 1.0 1.5 2.0 2.5 O.S 1.0 1.5 2.0 2.5 Ap II-4 6 325

SPECTRL REFLECTIVITT ISCALED TO 1.0 AT 0.56 MICRONSI

a tr

'-a c,~

ol 0

I ! I I I I I I ! ! I I I ! I I I I

VERAMIN NHOLE ROCK RA 0.177 MES 18 16 51 VERAMMIN WHOLE RCK Ra 0.153 MES 18 18 51 WVLENCIM ftIc~ONl -r-rrT-ir7-rFTTTT-r- TI111 r-T-T1FT-T-r-'rrrT-~rrT-T r r ('" t" ("~ ro

*' I ,

IM 7

W R,

/ I.LL~±L-LLL~.LLLL.LA~

"LA-1L-LA L I II I -L I II I i L"- L. O.S 1.0 1.5 2.0 2. 5 WFIVELENGT"(141CMNSI

mf U a;

:I CI2.C - C3i

0.5 1.0 1.5 2.0 2.5 (#4IC~ONSI

bI~V!LENC7H

WYCVLENOT$ItICMNI mr~~r~rT1 Tmr-r-r-I-

S 92c 24 r r. r. !; 42 I~ ~ -

I.)

II Wi I I I I I A I I 1 1 1.-1 1 1 1 I l A I I II I -y C" 'V tA t i A kI. I ~ L L.-LALLL ILL I.-1 1 .1- L"

1.v0 7E I rlq .0N2. C* IdAdCLf~c~tt fI4JCNON2 rrrr.,-r-r -1tm.r -r-r Trr-T- r-rrrTrrrr-r I- ~T-nT-rlrT---r Fr- -rrr-r-rur-r rrt a.

a. a a ~2. a a" a,

a. At. L--L LL.L.L L L.L.LL L-L L.LL.L.LI AA. IL. .4LJ.LLLLL..L.L....L.LAA.I .1..L-LL L.L CA-i 0.5 1.0 2.5 ?.0 2.5 Ap III-1 328

Appendix III

Spectral Reflectivity of the meteorites measured in this Study in the McCord 24 filter system and in the UBVRI System.

This appendix contains the spectral reflectivity data for the selected (see Chapter VIII) specimens measured in this study as seen through the 24 filters of the McCord system and the five filters of the UBVRI system. The 24 filter system data is normalized to the 0.56 micron filter. The albedo of the material as seen through this filter is specified. The UBVRI data is normalized to the 'V' filter. The data contained in this table is in the following format:

a) Type of meteorites b) Number refers to vertical column below. h c) Center wavelengths of the 24 filters. d) Each vertical column is the p reflectivity of the specimen in each of the filters. e) Reflectivity of each of the specimens at 0.56 microns. Lull f) Designates UBVRI filters. g) Each vertical column is the reflectivity of the specimen in each of the filters of the UBVRI system. h) Key to the numbers of 'b', name of the specimen and type. 329

A(II)- 2

EUCRITES - TYPE 1 (PY-PLAG ACHNCJ)

FiLT-R, 2 3 4 5 6 7 8

0,360 . 530 0.502 0.528 0.C5633 0.547 0.666 0.556 0.039 0.383 3.659 0.654 0.657 0.676 0.657 0a785 0.681 0.035 S4.2 C.748 0.737 3,738 -.747 i.0725 S846 0.757 0,030 0.434 3650 0.851 00825 G.835 0.815 0.911 0.844 C=a884 Ce468 C0.904 C03 893 0.876 00949 0.898 0.019 0.500 03930 0.922 0.907 0.914 0.957 0.925 0.013 02533 '2969 , 968 3.965 .'962 .0989 2.969 o J:6 02566 . 000 1.000 1.033 .000 1.000 1.000 0.0 :- 599 *. 33 1 0329 1.33 1b 32 1.-14 S"29 l 57 t..35 C a632 .. 059 1.062 -..374 1.033 .0C61 0. 010 C 665 0 109 1v092 1.100 1.1:2 1.C63 1.099 0.699 1.125 1.137 1.149 .. 160 a097T ..136 I.158 0.729 13124 .172 1.122 L.147 .,359 1.103 .972 0,837 C8854 Z.886 J.990 £: 997 3.943 3,047 8.55 C. 782 0. 63C 0.678 0.7'2 0.822 0.747 0, 362 O.906 0.697 0.592 V.673 G.725 0X726 0.658 O, 363 ;,947 :. 722 .. 698 V.741 ,0735 0.676 o0 063 %.853 0.9370.721 C.717 L,635 ~.851 0.859 0.806 ".053 -. 025 0 .911 20930 C.978 0.978 0.036 1.101 i.153 1.093 1.085 1 074 i.075 I1.86 1.1 I 0.039

R ( .56) .25, t.28u Z.i73 2.3_2 t.333 .;205 0.257 0.048

2.59 L.57 0.59 W.60 0.72 0.61 0.04 ..86 ,.83 G91 0.86 00 42 ..00 1.00 _.00 oCO 1.00 ui 0 L 0740 1.C9 C-a 8 1.11 :.j61.039 1.08 0,74 ,.78 CoE5 4.89 0C.54

HARAIYA ASU EUC JONZAC - NO RUST EUC 3 JUVI NAS EUC 4 NO3LEBORO - NO RUST HOW 5 PASAMONTE - ASU EUC 6 STANNERN POWDER EUC 7 AVERAGE FILTER RESPONSES a) AVERAG: VARIATION 330

AII1- 3

EUCRITES TYPE 2 (PY-PLAG ACHON)

FILTER 1 2 3 4 5

C-,360 D.615 0.479 0.736 Co.60 n.081 0.383 0.745 0.657 0.8:1 0.738 0.054 0. 42 .8:3 0. 752 &.864 3o813 .,, 38 C.434 .8682 0.840 i*919 0.830 o.027 G 458 0.916 0.893 0,952 04940 0.021 ;.5o 3;.935 0.923 0.963 0o943 .15 C,533 3,975 0.966 L,988 ,976 u.D08 ;*566 -. 03 L 033 i-.3 ;. S 1w1 *599 .0A5 .022 i4 7 00=3 *9632 ,.028 .036' !.026 13GC .004 .665 ,.,42 . 56 ~.4. .0- " 0.07 -.- 699 . 5- -. ;55 e, 6 o*-58o05 3.729 ..05 1.053 063 .0,5 0.005 3.753 If.06 0.985 .44 L.C.5 . 020 48.7 .6884 .793 .. 947 ,3Z75 L.054 C,85565077i , L.597 a814 .7L3 ... 77 0.906 3,650 0.524 ,0735 o.6Z6 *.075 0. 947 0.664 0.555 74C C.6 3 . 065 13003 30777 0.712 .8a22 0,770 039 .053 3.924 0.903 0.94Z C0.93 0.013 i."31 .033 L.034 1.C37 .o35 0.002 R(3,56) 3.257 0.264 W.355 .Z9Z 3.042

U 0 68 0.56 0.75 C.66 0.07 B 0.89 0.86 0.92 .09 .' 2 V .. 3 _.C0 .0: _.. , 0.o0 - R 1.03 1.02 .033 0o.0 I L*80 o.71 0.86 0.79 6.05

S6EEBA - NO RUST FINER (NICE) EUC 2 PADVARNINKAI - NO RUST (SHER) EUC 3 SIOUX COUNTY ASU EUC 4 AVERASE FILTER RESPD0NSS 5 AVERAGE VARIATION 331

A(IIII- 4

HOWARDITES - PY-PLAG ACHOok"DRITES

FILTER L 2 3 4 5 6

0.360 0 "497 0,567 G.589 .616Si 3.567 0.035 0.383 .62~ 0.682 '.705 0.711 J,680 0 029 0.4C2 3 .691 0.745 '.766 C.775 0.744 C=327 0.434 m77 0.831 G.840 0.852 ,.823 c026 :,468 .3842 Z.895 2.896 /:.91 2 t.884 CSCO OI.E96 0, 929 0.932 0.927 C.921 0.013 0.533 o,.975 C,977 5.982 0.972 0.976 0.003 0.566 - .300 0O0 1CCl0.C30 :.599 1 .017 .. 020 Co 002 .632 .39 D355 .; 43 0.665 Al ,C56 .075 i.055 ..070 0c010 0.699 :.069 -- _.391 .729 ,042 i.091 1.049 i.078 C,763 0 ,916 0.971 1.003 2 8 7 .. ,66 ,.835 :.773 C855 3 0.629 0.582 Cs745 L.603 0.906 j c.55.: 0.52 ".532 0.372 : 947 G.a 595 3.555 C. 677 Je569 C 3G66 637 0.785 v.742' S.742 Co53 887 '.992 .932 :0993 '- 42 1.038 -. 951 1.13 _ '347 :.s 11- 0.045

R( .56) 0.273 0.299 C.265 G.230 0.267 0,319

0.62 C.64 O.6 b '.62 C.03 0.85 C.86 C.sE7 0.84 L.CO 1.00 0.02 1.00 1.03 3, 74 2.73 083 0. 72 'xci0.07

' FRANKFORT - V. MINOR RUST HOW LE TEILLEUL - KO RUST HOW PAVLOVKA - N0 RUST HOW PETERSBURG PO*DER V.V; RARE RUST HOW AVERAGE FILTER RESPONSES AVERAGE VARIATION 332

A(III)- 5

DIOGEITE - HYPERSTHENE ACHONDRITES

FILTER 1 2 3 4 5 6 0.360 J0428 0.539 0.696 3055 J.544 3=.76 3z383 L. 542 0,65a C.794 0630 ,.654 3,070 304C2 "2592 0.7:C 0.831 n 671 C.701 0369 D434 ,637 £C782 .866- .7 *e748 3V076 0.468 0O702 0.850 0.899 C0768 0.805 0.070 0.5C 3.779 0.908 .934 0.345 .o366 0o.54 0.533 0.948 0.977 C,9b0 C.965 u.967 C.O11 0.566 i,003 4.000 1.03 .i.o 0 3,:599 .324 1. 15 1 01 5 o'17 .317 0.632 .. 048 1.035 1.036 .0.34 0.005005 0.665 1.083 1.354 1.054 1.C.l ,.052 0.016 9 0.699 1.100 1.065 .C52 . c' .049 3 334 0.729 1.055 .1039 1,327 0,899 a.005 :0.353 2.763 -v879 C.938 393 ,.75 ,. .37i 369' 0.807 0.582 0.719 0.707 0.a33 *.623 C C090 0.855 3375 0.51 0.532 C.3. ,.,425 0=906 0.311 .44-3 0.442 ;.25E ..365 C1085 04947 30336 0.484 0.50ki 0 39 J.407 =.>33 .492 0C675 3.736 !%.55 .a053 .,760 0.899 0Q97. 0.78 ).854

1..01 .0OC; ~.029 .0C85 0.982T i.024 033-. -33 R(OoF6) Z.353 0.332 0.231 C.423 ,.335 0G053

0:49 0.59 C*74 .60 0.70 0.81 0.89 .9o7 :. 79 -f6 -. 09 -. 00 '.w3 L.53 C.99 0 99 Z"52 L.b4 Ce64 G.45 0.56 0.08i

JOiNSTOWN FI'ER POWDER (HY ACHON) DID RODA - NO RUST DID SHALKA - NO ALT DIO 4 TATAHOUINE - POnDER NO ALT DIO 5 AVERAGE FILTER RESPOSES 6 AVERAGE VARIATION 333

A(III)- 6

AUBRITES - ENSTATITE ACHON

FILTER I 2 3 4

0,360 0.885 0.890 C.887 0.002 3.383 3,915 0,915 0.9i5 b ,432 ,928 0.931 3.929 0.434 3a949 0.950 0.949 0,0010.0 ,0468 3.966 0,968 V.967 O.0Ld 33977 0.978 0.977 0.o31 ,0533 5,991 3.990 0.990

0, 599 .037 1.302. 1.004 0J003 0.632 0.02 0 665 r,,-c7 1035 0,C32 0,699 .334 r.02 .2729 .,037.. 007 3.763 1.304 0.807 3.999 3.994 0.996 3, 855 W,9E- ; .935 0.936 C,936 .,987 3,984 0,985 .947 :.992 0,985 C 988 1.003 ,.994 .G002 0.985 C.959 .,994 3.989 C,.085.3C54 0.993 0.985 ,0989 C,. -4

R(0.56) 0.232 3.208 .220 0,3i2

3.93 0.96 0.96 C.90 C,GG ,Ol 1.01 030 iGl 0.0.l r 2 c .. >00 0.99 C.99

CU.BERLAtND FALLS (EN ACHONo4 AUB CUMBERLAND FALLS (E-N, ACE-Or4.) AUB AVERAGE FILTER RESPO'.SES AVERASE VARIATION 334

A(I)ll- 7

NtKHLITE - DIOPSIDE-OLIVINIE ACHOJ

FILTER I. 2 3 4

S,360 ,.279 0~289 0.284 0. 0: 0.383 0.357 0.351 0.344 0.3CC 0.402 0.40 0.417 0 4C8 0.009 0.434 0.510 0.524 0.517 0,468 C.623 0.616 0oC37 c:,50 2.733 *723 0,533 0,882 0,892 0.887 0,005 C.566 1.000 0.0 00599 1.037 1.025 0 2 6 0.632 1.035 .. 017 1. ,o 0..59 .65 S 35 1 -24 0.699 .032 1.011 1.321 0.729 .020 1.010 0.,763 C1 994 ,.983 o.988 0.8:7 0.953 0,944 C.947 80355 .., 44 3.843 3.843 0.936 0.654 0.661 u.657 C.947 0.544 C0555 0.549 0.032C,CO3 3.511 0.513 0.512 .C 5 1,353 .,555 3.562 0.558 C '%3# ,666 0.669 3.657

R(0.56) 0.132 0.124 0.128 0.0 4

0.34 :.33 -' 3.58 0.60 0.59 1.00 .3 I., r " 54 :.83 0.83 C. 32

NAKLA FINER POWDER (DI-L ACHO%'J NAK COARSER POWDER (UI-OL ACHON) NAK AVERAGE FILTER RESPO 1SES AV.AGE VARIATION 335

A(III)- 8

CHASSIGNITE - OLIVI.E ACHOGN

FILTER 13,2 2 3 4 0:360 3.223 0.222 3,222 0C383 0a359 0.356 3.357 O.032 0.002 0.402 0.451 0.451 0,6451 3,434 0,575 0.575 0.575 0.0 0.468 3.655 0.653 0.654 0.oC01 0.775 3,777 0,533 0.919 0.917 0.918 0.001 0.566 1.313 4.000,.OCC 1.300 0.0 0.599 ;.009 C.300 .. 028 0o632 .27 io27 0,003 Z,699 1.303 1.001 0.729 3.91k 0.910 O,a %d 3.763 %3787 0.786 i532 ,, 5. 5,633 0.632 0.855 0.5i13 3.513 t.513 936 .439

0.947 0.387 0.390 ~.388 -0-C-i 0.331 0.33: 0,6. 1.053 0.319 0.320 0o0f: O, CC2 1101 0.3642 .. 341

R(3.56) 3.353 0.359 30356 C3

3029 0.29 0.29 .63 0.63 0.63 1.00 30 "o C 0.99 0.99 .,99 2.55 3.55 :,a 95 5

CHASSIGNY - NO ALTERATIO'N CHAS CHASSIGNY - 143 ALTERATION CHAS AVERAGE FILTER RESPONSES AVEAE VARIATI3N 336

A(III)- 9

ANGRITE - AUGITE ACHONDRITE

FILTER 3 4

0i36C 3.83 0.848 0.843 0 .005 J,383 3.,63 0.866 Co864 . 883 0.885 C.434 0.885 0.884.864 0.468 3.866 0.880 0.873 .023 2.882 O.888 0.885 C,533 i. 946 :.949 0. 556 .00 0.599 1.074 1.C52. 1.063 0.026 0.632 1.176 0.033 0665 1.240 1.275 0,699 ,473 1.380 1.426 o 547 0.729 1.623 i.50 3 1.561 0C,51 2.71 i,566 6i.39 0.373 ..710 o1552 0,G 7 9 .51 . 1o585 0.275 C. 9065 &.572 1.429 0 947 13404 .472 1.003 1.578 1.436 1,507 1.053 1.693 1.542 1.617 0.0 76 1.101 1.847 1.680 1,763 0.0G 3 *v*36 R( 256) ;3355 J3a53 3a054 ,021

0.86 0.87 0.86 3.9C Ce 69 1.033 1.00 &.42 1.33 1.37 i.67 1.59 0.07

ANSRA DOS REIS - NO ALT ANG ANGRA D0S REIS - NO ALT ANG AVERAE FILTER RESPONSES 4 AVRAS3 VARIATION 337

A OVINIITE-

UR=ILITE - OLIVINE-PIGEO\ITE ACHCN

FZLT-ER 'I 2

, 360 ;.594 0.592 0.593 0.01 C.663 0.659 0.661 0.00? ,.734 3.783 o 434 0.781 0.782 0.00 0.468 0.829 3.,830 0.829 0°530 0.376 0.871 0.873 3.533 11C. %-I 0.954 C.955

0.599 1.021 1.022 1.026 C.C30 (.632 1.027 ,1.326 Co665 io.31 1.030 1.03031°031 0,000 0,699 1.033 0.0;2 0.02 b,729 iD29 1.326 1.027 %,763 S*. 017 1 012 "=:i4 0.002 .807 C.969 0.969 0.969 0.55 C~,929 0.928 0,928 0.02 =936 CD905 ,-947 0.023 C 929 0.93i 1.003 0.929 1.053 0.95 C- 954 0.952 C.C,5 1,101 (.975 90.94 C.979 Ca C4-

R(0.56) ;,.88 0.078 0.383 .015

5.64 0.64 %,a6

.305 1.05 3.97 0°97 0.97

NOVO-UREI - RARE ALT UREI NOVO-UREI - MINOR RUST (NS) UREI AVERAGE FILTER RESPONSES AVEASE VARIATION 338

A(III)-I

NI.KEL-IRON METEORITES

FILTER 4 5 6

C,360 1.722 0.734 C.658 0.765 ,.720 8,,031 0383 0 769 0.773 0.735 0.799 j.761 0.028 " e74.! _,791 0,434 0.862 0.860 0.,08 0.87b 0.851 0.022 0.468 0.905 0.875 0.903 0.500 C0.940 0.934 CG919 C 942 0.934 00307 0,533 , 970 0.967 0.962 ,,977 0.969 C0 C05 ,.566 0.599 4.025 1.026 1.034 A.G26 0o004 3.632 1.052 1.348 . 0362 L.037 ,.053 C. C07 0.665 1.C74 1.068 1.090 .= 57 1,072 0o10 3 699 i.084 .092 0,.015 :.729 "119 1.1.39 ..- 82 1.111 0.763 1..47 1.127 ;°.17 V.093 .133 36 .*807 .172 i.143 e195 i°153 C,855 " '199 1,166 1.2r3 .. l=O6. .. 189 i.255 1.154 i!99 v, 044 ,906 1,Z59 .214 1.286 .. 223 0.053 ;z305 1.249 i.327 _,152 19.253 0-047 :*053 :.337 :.278 1.358 -,*,° , ".283 Co.074 0 0B °. 71 1.01 e.375 1.385 10308 E 57

R(0.56) .16~8 0,1 73 C.289 3.232 J.215 C.45

,.75 &.76 0.69 C.79 C.75 0.3 0.88 0,89 3.87

h.23 iA.0 1.12 .07 >.10 4, J.I 1.20 1.26 io_3 A. 20 1.34

CASEY COUNTY - CUT SUR CO T CHULAFINNE - CUT SUR MOCT BUTLER - CUT SCRTED SUR FSTO BABB'S MILL - CJT SUR NRAX AVERAGE FILTER RESPONSES AVERA$E VARIATION 339

AESPAG STONYIII)-I

MESOSIDERITE - PY-PLAG STONY-IRON

FILT-.R 2

0.360 C,825 3.721 0.773 0.352 ., 383 802 :.729 :.765 0,402 0.746 0.774 0.028 3 434 3. 843 30810 3.826 0468 is 876 0.857 0.866 CO;70.39 j0.500 3.909 0,894 0.93".

.*533 S1 95.. .946 -. 948 *, .J ,- .566 1.000 1.000 coo j.05Z 1.047 00035 3,632 1.102 1.090 Cl 655 1.146 1.130 L . ,,. = 699 .194 .172 'o 729 1.231. 1.204 Z. 763 -''59 1.23- .. " OGB37 2.223 1.245 282 c0.0 .27: 35 1.294 1.1793 4 1.364 1.334 35553 1.440 1.396 O. -43 12053 J.U3.,'T3 1,4181,,393 1.448 lI 1,546 o1.482 S. C 64

R(,55) ..177 ,.153 3.i55 .'0L2

o74 3.79 .87 0.34 C.85 C.Oi

.. 15 1.17 C.9 :,29 1.,35 1.32

VERAMIN WHOLE RGCK RARE RUST MES VERAmI N WHCLE ROCK RARE RUST MES AVE~RASE FILTER RESPOSES AVERAG5E VARIATiON 340

A(III -13

CHONDRI TES TYPE E (ENSTATITE)

FILTER 2 3 4 5 6

C,360 0 .83 0.692 0.676 0.730 0.724 0oO40 0,383 , o 846 0.722 3.705 0.752 . 756 0X245 C402 0 0.755 0.738 C.733 0.785 0434 .935 o,817 0.799 1°83 °.838 0.034 C3468 v S942 0.867 0.837 !0870 %o879 .96Z 3.031 2.938 5.885 : 95 0o915 :-0 23 .:533 . ,985 .,959 2.953 ,.964 tX566 1.000 1.333 1.000 1.000 ~~599 i.;23 l.026 0007 3.632 1 i.047 i.36i 1.043 .387 4 6 1.060 "oJ7 .062 3,012 0 699 .374 ",082 -. 3+'79 ..017 ,114 ia3 89 -. "98 0.837 . .435 .393 . 25 i.075 C 025 03.853 -,039 .,12O8 "074 105 ,17. _.090 1..3 3.379 1.117 033 v947 . L,102 -. 45 ..133 . 39 '9148 0.047 I -67 ... 53 CI fo116 .. 177 ..159 1.1$1 1 1.185 C. -'C *262, 1.116 '.171 2 6 5 R(C56) C.376 0A179 30.18 0.163 0.i34 0G.37

.a3 ,,372 0.70 0,75 j075 0.92 .84 0.82 3,35 J. 36 C.03 -o 11 &.1 y . 09 .39 G.1S 0.03 1.11 io16 J'0 014

ABEE - MINOR ALT (NS) E4 HVITTIS E6 KHAIRPUR - RUST ST E6 PILLISTFER - NO RUST E6 AVERAGE FILTER RESPONSES AVERAGE VARIATION 341

AtIII)-14

CHNDRITES - TYPE H

FILTER 3 4 5 6

0.360 0.640 0.7o49 0.599 0.648 0.635 0.601 0,654 C.383 0.7:3 0.81 C.668 0.774 %.708 0a71 T 0.673 0.402 0.7653 0~842 0.708 0.750 C,767 0.7.1 0.764 0,434 03.832 .786 .368 0.820 c832 0.782 0.830 G3468 3,926 .547 0.877 C.879 0.839 o,883 0.924 o 952 0.896 0.944 0.919 0.920 0.886 0.920 0.533 3.974 0.986 30955 c.983 0.970 0,970 0.955 0.970 0.566 i.000 1.000 1.000 1,000 1.0X9 0 599 1.009 l.024 1.022 1., "5 1.331 1.018 7,632 .3233 1' J4 1,335 26 1.355 035 3.665 1.049 1,340 1.043 1.047 .O"35 1.074 0.699 ..062 1.350 1.072 1.055 1.C37 1058

0.729 3,.62 1,359 1-49 1.*055 1o32 1.078, 1,057 233 1.349 .093 .,342 .C!6. o.038 0,96 ,.954 3.995 0.977 0.973 0. 853 0,392 3.93.5 0.929 0.881 0;, 994 -"906 J*.398 0,772 0.827 0;858 -3 947 3,374 0.873 L.910 0.783 0.938 0.869 - -. 3 0 ,913 6093 0.944 0.833 :O52-.: 979 2955 0.965 -.873 0.921 0,938 1,010L L0 933 0.980 0.984 0,912 0.958 0.909

R(0.56) 0.1 7 0.2 .5 G.123 0.242 U.257 0.295 0.228 CG.Z4

0.66 0.78 C.64 0.74 0.68 0.68 0.64 0.69 0.55 0.68 G.84 o.85 0.85 S.C, 1.00 i4'5 1003 10 L .05 1.07 -j93 1.91 .97 L.87 0.93 j0a94

1 TIESCHITZ - NO RUST H3 OCAAISK - MIXED NO RUST H4 3 CATALIA FINER POwDER, DARK, NO RUST H5 4 COLLESCIPOLI - V. RARE RUST H5 5 PANTAR - ASU H5 6 LANCO4 - VVR RUST H6 7 QUEEN'S MEROY - W ROX MINOR RUST H6 8 AVERAGE FILTER RESPONSES 342

A(III )-14

ZHONY~rhIT3 TYPE H

FILTER

0.C36j 342 C34 .4,- Z B434 029 0.468 0.500 z 53~ C)9

C2 599 03E 0.665 fC° ° CI 699

33 Ct 9735

, 947 04 033 D,353 028 1' 101 C23

R (0,56) 0,048

9 AVERA3E VARIATION 343

A(III )-15

CHO4NDRITES - TYPE L,,

FILTER

0. 360 ,595 3,599 C0636 0. 610 00383 ,666 0.678 C. 732 C, 689 0.718 L2745 0725 0.434 0 791 01790 C.S11i 0. 797 0.009 0.468 0.861 C0.857 .875 io Sb4 0.50:3 0.91i 0. 910 0.919 u.37 ul 972 0.954 Co965 01.03 ,.566 Jo j 0.599 2.025 1.019 ;.022 0.632 - 045 .083.043 ,=41 .345 5.002 i3 072 "1,08 1.372 f,699 3.510 ~°.95 , 955 1.313 CJ763 1-398 ...'53073 i.e 3o837 C.988 O.9050.855 04 932 0,93 " 878 O,855 ~3399 O=947 ).896 368 3,9o jC 0.027D.:35 lj,936 '-94: 0, 904 0.5019 Ji 353 0,965 0.948 0.971 " ,36 0.014 1i101 3,998 ".994 1.027 014l

R(3,S6) 3188 0,221 0,237 0.25 03012

U 0,63 0~64 3, 7 33,5 0,02 B 3.82 3.82 3.84 0.83 ". i V ;,3 a.S i.A3GJ R 1.07 1.39 1,37 1.08 0.01 I 0.96 .97 1.-i C.98 0.

BALD MOUNTAIN - rO RUST LA, CYNTHIANA FIFNE POWDER L4 SARATOV - ASU L4 AVERAGE FILTER RESPONSES AVERAGE VARIATIG\ 344

A( III )-b

CONDRITES - TYPE L5

FILTER 1 2 3 4 5 6

0.360 0.570 0.567 0.645 0.5,7 0.572 C0036 0.383 ~.674 0.634 .68. 0o.67 ,.649 o.029 D0402 0.725 0,682 0.7 6 346oS 0.695 0.020 0,434 0,803 02763 0a786 J0755 0.776 C0018 o,468 32858 0827 o.840 J.825 Q.837 L. 12 0.500 C.909 0,883 0.890 0.884 0,891 0.009 0.533 03965 0,951 3,954 0.95' 0,955 .CC05 0.566 i00 1G030 .0312 i.C3 i.000 0. 0.599 .02c 1.030 1.030 1.028 i.027 0,004 a 6-02 -147 ,360 lJ64 . ', 57 . .357 ,. 5 0.665 19079 13085 1.099 ,.035 1.087 0.006 C.699 ,~093 1.105 -,13:.120 i.04 ~3 CI8 0.729 1,399 1103 - i9128 ,1095 .104 0'.12 ;,753 -- 348 1073 J;;_v .8 7 1.93' 1,34 .v39 1949 .,983 41i C.855 0.79 3.9036 03953 0,313 .859 0 C61 0C9 6 0,737 08560 3.99 .75i J.814 ".70 C.947 .738 3,868 0.91C 0W751 04817 0.072 --Vi3 1,781 5,911 2439 ".,764 ,d854 0,07 1-33 .834 4957 0982 .536 a.92 -11-7 .101 3,894 1000 i.024 0.891 3.952 , 06

R('.561 .0,249 C.2180 Z..72 Ca24: .21.0 .0364

U 4.62 0.61 0.68 0.56 0.62 00.3 S3 283 0.79 3.-i 0.79 0.80 02

) _; A .3 i. R 605 1.09 c.0 0oC i,08 Ca01 I .86 0.96 .433 3.56 .,92 C.36

1 AUSSOl FINER POWDER (NO.-STAINED) L5 2 HOMESTEAD COARSER POWDER L5 3 KNYAHIA WHOLE ROCCK L5 4 SHELBURNE RARE RUST L5 5 AVERAGE FILTER RESPONSE.S 6 AVERAGE VARIATIONJ 345

A(IIll-17

CHONDRITES - TYPE L6

FILT-' 9+ 5 6 7 8 0.416 0.363 0.455 %.577 0.6o45 0.526 0.603 0.582 S383 .,584 3.539 ",685 0.744 . 648 :.727 0.683 0.402 0.503 0.744 03797 0.713 C.786 0.738 0.620 3.434 3,747 0,7003 0.821 3,864 C.793 0.852 0.812 0.724 0.815 0.,777 3.87Z 0.913 3.8483 0.898 0.866 0.786 ".500 0.877 0.846 0.920 30.947 0.905 30.939 0.915 0.857 0.533 .951 0.938 C,970 0.980 0.965 C0980 0.967 0O941 X 566 1.011 3,599 .022 1033 1,324 3035 1.013 1.002 0=632 1.056 lo324 1.026 10o38 ±.024 1.052 1.0379 -2665 21.3 , 375 1.334 ... 40 1.088 0 .699 1.079 034 8 9 1*039 .. 014 1.042 1.065 - "56 1.064 f,993 .997 1.317 0.982 0.954 C 934 5.988 0.985 0,856 C.907 0.881 0.818 5,816 0.896 0.827 ; 55 l.785 ^.773 .0685 1697 0.792 .665 .9036 3.638 '0.722 0.715 0. 619 Cab38 0.747 o.947 .637 0.721 :.715 0o764 0.618 0.634 0.748 0.564 3,588 0.3762 3.752 Co656 0.767 0.597 1.0.53 0.6b4 2.750 .6312 3.794 08 7 ,.;721 .o695 3.844 .. 32 3.864 3.8 4 ~.777 .0742 C.736

R(0.56) 0.273 0.314 0.332 0.328 30366 0C343 0.271 0,352

0,53 0.51 .63 0,69 0.59 .66 G0.48 ,,78 0.8C .82 0.87 0.76 0.74 1.00 .03 1.03 1.0 1.00 3C5 1.02 ;099 10C6 1.01 3,55 4*.J1 - u.74 T,74 C34 Co73 1.030.84 0.84

1 ALFIANELLO L6 2 ANDOVER - SOME RUST L6 3 AUMALE - NO RUST L6 4 BRUDERHEIM - ASU L6 5 BUSCHOF - NO RUST L6 6 CABEZO DE MAYO - FINER, 13 RUST L6 7 COLBY(wIS) - MIXED V RARE RUST L6 8 DRAKE CREEK -L6 346

A(III)-17

CHONDRITES - TYPE L6

FILTER 9 1 1 12 13 4 15 16

0.360 0.488 0585 j.564 0.47 L.549 2472 0.634 0.534 0= 383 0.597 0.571 0.671 0.588 %.665 0.573 0.702 0.642 0.402 0.664 C.7221 C.657 V. 728 .642 0.758 0.703 0.434 G,755 0.797 0.804 %.7470.808 0;733 C.831 0.786 0.468 %. 620 C0.859 0.856 .,861 Oa 9C3 0.8844 0.844 -u 503 ;,879 0,937 a,937 t872 (.910 S.871, 0.899 ,0533 0.955 (.963 0.962 v.967 0.948 0,976 0.961 0.566 -. 000 0.599 .. C21 La023 1.018 .032 1.025 1.018 0.632 1.C37 1.032 ..055 1022 2.028 ..034 0.665 ., 56 _.348 176 1.78 ~1.51 .. 038 0.699 1.C57 1.052 1.C57 1o061 1 .038 1.C83 1.046 1.054 0.729 1.043 1.043 .064 1.026 1.067 1.037 1.039 . 763 8.987 1.31 1.3L6 Q.972 1.005 0.9~8 .807o 0.861 0.927 C0.91 C0 871. .843 0,82665 0.903 0.870 a 855 ., 737 La 831 :768 2.716 <.718 -727 3.796 0,3747 C.67. 0.775 0.704 (.651 ,.655 0.664 C.746 0.686 C.947 C.666 0.774 0.74 0.6'7 655 0.662 0.749 Go0 684 0.797 0.7544 0.667 C.692 0.759 0.719 1.053 ,.753 0.830 0,777 0.747 L.739 0,757 o0.827 0.767 .. 8i1 ,0868 0.823 SoB.,9 C. 792 0 -.816 0.875 0.820

R(C,561 0.264 0.317 .25E 06.31 0.296 0.273 C.289 0.306

;.55 0.63 :.62 4.61 ".53 C. 65 0.59 0.79 C.83 0.83 C.78 0.8 3 0.77 0.85 0.8 1.00 I9C3 1.00 1.04 .02 1.04 1.06 1.03 1.04 :.79 087 0.82 0.79 ,.77 3.8 C. 84 0.80

9 GIRSENTI - VR RUST L6 LEEDEY - ASU L6 NERFT - RARE RUST L6 12 ST, MICHEL FINER POWDER L6 13 TOURINNIES LA GROSSE - RARE RUST L6 14 UTRECHT - MIXED MINOR RUST L6 15 ZAVID - VR RUST L6 16 AVERAGE FILTER RESPONSES 347

A(III -17

CHONDRITES - TYPE Lb

FILTER 17

.360.350 .056 C.383 .C56 0.402 %434 0.468 C.034 C0.500 0. 025 Ca 533 C4566 co599 C.02S.00. 2 C0632 0.012 0.665 C.699 0.729 w,763 0.807 0.28 C0855 . 042 Co 936 C.049 C, 947 L48 1.053 C, C43 10 "I J. 0 v A v.034

R(C.56.d 060

0.05 C.0 .C2 0.0I.

17 AV-RASE VARIATION 348

A(IIl)-18

CH0NuRIT=S - TYPE LL

FILTER 2 3 4 5 6 7

0.360 03 555 03 518 " ,,561 ,.5:6 ,.538 0.383 %06930.612 C.685 C,670 0.641 0.660 0.C27 0.672 3,753 0.7L0 0.723 0.026 .:434 19823 j,763 v.791 0.48 0,874 0.8632 %0874 0.870 0.842 0.858 ,001 7 0=500 C.927 U.892 0.922 Co979 0.897 0.912 0. 533 0.96. 0.973 C 9 30967 0.970 00 oCK 0.566 :,033 ,030 2.000 0.0 , 599 ,322 1.014 l0.5 6 i..055 1.034 .°C36 .* 015 1.032 C.66b5 i.060 1.060 i.054 o.026 1.0C52 0.015 0. 699 "3042 I G93 1.073 1.=063 ".021 1.058 0.022 %4729 ._090. C.066 2.055 0. 999 1, 046 0.029 : 974 0 t.942 _, 7 3d837 0.859 0.978 0.949 C.945 0.825 0.913 0.053Co C,5.2 U.555 0.878 0.850 0.816 0.057 J3 326 00658 0765 0.058 C,701 0.947 0.8i9 0.772 0,6.6 v.637 30749 0.064 0 842 .77J .-2755 0.072 -330 G,719 3.870 0.795 0.562 0.781 767 0,9'i3 0.841 08.93 0.708 3.825 0.070

R(C.56) C.27i 0.272 0.369 0.234 ,.227 0,275 0.C38

C.62 3.57 0.61 Co6 . 57 0,60 0.02 ,,85 . 79 3185 .8O3 6082 !083 ^°2 ,00 S1.00 .. 01 .CO 0.0 0108 .*35 2o 4 0.02 3.81 Oo9, DDLr"%aL 0392 038 r.76 -.85 0.36

V4VILOVKA - VR RUST LL SOKO-BANJA - MIXED NO RUST LL4 OLIVENZA LL6 JELICA - V RARE RUST LL6 MAN6HOOM - NO RUST (GODP) LL6 AVERAGE FILTER RESPONSES AVERAGE VARIATION 349

A(IIl)-9

CHONDRITES - BLACK

FILTER 1 2 3 4 5 6

C.360 ..816 0.779 0.859 .738 0.798 0.039 C 3932 ,871 0.83 C.892 C,774 0*842 0.,39 .-422 .897 3.856 .934 -z795 i.863 .,.38 C0434 0.927 0.892 0.929 0.848 0.899 0.029 .468 953 097 7 0920.942 3677 .922 25 C0500 . 970 C0937 0.953 0.895 0.939 0.022 [.533 v.993 0.978 039868 .965 CG981 ,C.C9

;.599 ,0G8 1.016 io316 .029 1.017 G.006 C.632 .01i5 -.027 :.028 I- .C28 3oC07 C.665 z%0291.043 .03'4oS2 ,.044 -0109 C, 699 L 5 1,C49 .349 .33 -&054 0.0314 ..729 134 x.364 .,.6-. -..a97 .. 64 ~0,16 Co763 .,021 1,362 1,053 -G061 0 2I0 .8 7 C.963 .. 328 1.340 -.075 ..*026 G.C32 W.855 %.886 0.9938 .a20 1C53 .9839 3.052 .,96 L1858 C98e 33 1.,4J3 4,975 0, 59 .;947 8677 0,993 7.31273 .989 Co356 00C5 wG912 0,36 4.a43 .093 .016 0052 1.035A.0530.94 ±.&54 1.1Z9 .035 0.047 :131o , 967 1.060 1.073 1 12o 1.056 0.04

R(0,56) .108 0O037 o,.oa08 . 85

U i.84 ,.=1 C088 0.77 3082 .04 B 0.94 0.9. 0.94 3,87 0.91 C=03 V =33 L3 . . .°- eV v R 1.02 J.05 i.O5 1.09 .05 C.2 cI9 a.*03 .044 .C9 .0C2 0.05

i FARMINGTON - NO RUST ('S) L5B 2 PARAGOULD - NO RUST ('1$) L5B 3 SEVRUKOVO - NO RUST (NS) L 8 4 RUOSE CITY - W ROX (NS) NO V RUST H B 5 AVERAGE FILTER RESPO:SES 6 AVERAGE VARIATION 350

A( II)-20

CHONDRITES - TYPE C.

FILTER 3

9.676 0.576 0.0 4.698 .o698 0.700 0.700 0.0 0.434 0.787 0.787 0.0 0.468 ,.837 0.837 0.3 0.500 ;,.887 0.887 0.0 0.0 )2533 3.98 ~.982 4-.2000 0.0 0.566 . 000 0.3 .599 C 0.0 0.632 i,.047 1,047 3.665 1.062 0.0 3.D 2 699 -* 37; ., 73 0.729 .o094 ,1094 0.0 0.763 . 098 .0398 0.O ,.807 . 3to7 3,.655 .0 57 i.338 S9306 " 37 0.0Do? 0.947 1,050 AC 03 1.375 .. 375 1.C53 0c85 o.085 0*0 r.101 ..0395 ;.,095 0.0

R(5.56) 0.040 0.040 3.j

0, 70 0.70 0.0 iio 82 1.00 0.0 -. 08 ".0 . 37 37 0.0

ORSUEIL (XS) ci AVER4AGE FILTER RESPONSES AVERASGE VARIATION 351

A(III)-21

CHONDRITES - TYPE C2

FILTER 3 4 5 6 7 8

C0360 0.640 0.634 0.424 0.490 0.513 0.419 0.479 0.514 G,383 V.712 0.703 0.535 b58 0.590 0.529 0.573 0.6C4 ;3757 0,753 0.669 .6 13 " .641 J.671 C2434 0.859 0.857 0.729 ,.789 0.786 0.745 0.767 0,790 G0.468 t.921 0.900 0.814 .Uo867 0.873 0.54 0.874 0. 871 0.5003 0.964 0.95.4 0.883 0.941 0.939 3.925 0.955 0.937 Z) 533 :.3i2 1.303 3,964 31:. 984 G.973 0.995 0.989 6, 566 C0.99 0.983 0.971 1.002 voqd6 0.989 b.9bZ 1.003 0.934 0.988 5.632 0.999 C.977 . 972 0.956 0.973 0.665 C.953 0.963 1. 318 J.963 1.39 0o936 0.972 0-699 .,936 03916 C.965 0.961 1o032 0.930 C.966 0.729 3.963 5097: 0.969 .2354 0.927 0,977 2.753 0,950 0.959 1.058 Lc964 *.979 . 54o 0.938 0.989 0.8C7 0.910 .046 C, 997 0.989 1.045 0. 959 G.912 1.034 0.93: 0.855 C'*908 1.027 i.013 0.994 0.969 2.953 0.9680 0.947 0.972 0.995 S947 2.949 ,.944 1274 0.986 1013 0003 0,95 0.981 ..035 10120 1.040 i.053 0.976 1.012 1.137 'oC77 .044 1.i39 1.015 1.057 1*C01 .c iO3 1,009 1.153 i,090 1.051 1.154 10025 6 95

R(3 56) 3.034 0.035 C.038 0.'43 ;.043 C.045 0. C6 0.041

.* 67 U.67 0.48 0.54 0.56 0.47 0.52 G.56 0.87 O.6 C.76 0.81 0.81 0.77 G079 .9OG 0.00 1.00 1.05 0.95 0.96 4e 'JO o.99 0o.99 CS 0.96 0.99 h1e.. . 0.94 0.95 .23 . 02 0.98 4..

COLD BOKKEVELT (32-100) (NS) C2 COLD BOKKEVELT (100-2007) (NS) C2 COLD BOKKEVELT (GT 20*M) (NS) C2 MEHEI - (GT 20OH) (NS) C2 MURCHISON - (NS) C2 MUkRAY - ASU (NS) Ca NOGDYA - INS) C2 AVERAGE FILTER RESPONSES 352

A( III I-21

CHODNDRITE5 - TYPE C2

FILTER 9

C. 360 C.070 0.383 0.059 Z.4C2 ,048 ,.434 ,C39 0i468 0z023 0.500 0.019 0.533 0.C13 C.566 0.0

0. 532 0. 017 0.665 ,.0 24 7a699 60,034 .o729 Z.C35 0.763 .a0i8 $0-807 5.044 Cz 855 C 043

&947 .'46 3OS Ga 047 1.053 C.052 .,101 .Cs54

R( 56) 004

U 0.06

V 0.0 R 5.03 I 0.05

9 AVERAGE VARIATION 353

A( II)-22

CHONDRITES - TYPE C30

FILTER I 2 5 6 7 8

0.360 0(i627 0.660 0,582 0.509 0.625 0.618 0.617 0.021 0,383 ,715 L.732 0.683 3,757 0.712 C.716 0.7'78 0 0! 4 3,402 ^.,767 30078i 00744 v. 841 0.781 C.778 0.768 03012 0.434 30849 0,849 0.842 3.865 0,856 0.850 0.007 0,468 0.891 v.894 0,902 S,896 Qa912 0.905 0.900 0.006 2a921 0924 C0.94 4..941 0 .940 C.943 0.935 0.533 3.970 0,969 0.976 3.977 0.980 0.978 0.975 0.004 0~.66 1.000 1.030 1.000 0.0 i 599 10030 1.017 1.034 0.632 1.030 IsD3c i.039 1.032 1.033 0.003 C,6655 1040 337 1.056 ". C56 1.043 1.050 1.047 0.007 J.699 4.144 1. 58 e69 1.334 1.-54 1. 55 .o049 1.067 i.033 1.054 1.058 0.013 0.763 1.052 1.068 " "86 .47 1.058 0,015 .C.33 1.0C 70 .005 1.013 i.C35 0.018 C.855 o1345 .0193 ",955 0- 976 1.004 0.025 9:, 6 .3998 i;1s33 9.946 '.955 0.99Z 3.3 2 8 V947 i. J1 ir 27 %.935 949 1994 0.988 0c.31 .014 .922 C.972 1C53 o.9'37 3.914 0.036 0.986 1%015 0.996 3.954 ).03 0.910 0.966 0.039 .. C.995 1.026 1.003 0.994 0. 913 0.921 0.975 0.039

R(3a56) 0.086 0.073 0.116 ,172 0.106 0.106 0.111 0.027

J.67 C0.70 0.63 0,67 0.66 0.66 S.86 0387 3.85 3.85 0 *87 ". 87 3.86 1.00 1.0 1.00 1.00 1.00 0.010.0 C..a 1.06 1,07 i,04 1.0'S 12.3 112CO 0 5 0.. a2 B.G5., ,4 1625 G.98 *099 -io112

FELIX (NS) C30 KAINSAZ - (NSI C30 OR.,4ANS - NO ALT (NS) C30 WARRENTN - (N ) ( S 20M) TVJ C30 WARRENTON - (NS) (CO3-2C0M) TVJ C30 WARENTON - (XS) (L-0) M) TV3J C30 AVERAGE FILTER RESPONS S AVc ASE VAkIATION. 354

A(IIll)-23

CHONDRITES - TYPE CSV

FILTER 1 2 3 4 5 6 7 8

0.360 w.595 0.499 0.667 0.694 .453 C0651 0.610 02333 .705 0.579 0.73 1 0.747 3.517 0.773 0.710 0.680 0,402 20764 00650 0.793 03797 0.584 0.815 0.753 0.737 0.434 0.844 0.765 0.881 3.707 0.873 0.827 0.68i008-4 0.463 0.881 0.842 0.935 0.930 0e.792 C0.914 0.875 C.500 C; 917 3.902 0.968 0C962 0.857 0.937 0.9i7 0.923 03533 C 974 0.967 0.996 0.989 0.943 0.974 0.975 0.974 0,566 1.00 10000 1.000 1.000 J.000 1.000 0.599 1o024 1.024 0.980 Cs997 l.039 10033 1.026 1.016 0.632 ;..037 .0336 C.968 0.990 10061 1.033 1.037 1.023 0.665 1.057 i056 0.966 0.985 .. 087 1.057 1.049 i. 037 0O699 13075 1.070 C * 961. 0*984 1.100 1.047 0.729 1.087 1. 076 0.954 C0977 A.105 1.093 1.067 1.051 0.763 *.090 1.084 0.949 G.973 1.69 0.807 ".073 i.371 0.940 0.962 3.387 1.092 1.0C49 ,. t -.°39 0.855 1.057 3.927 0.949 1.064 1.077 1.0c28 1.022 3.903 1.6 , 4. 044 C.915 3.934 1.067 1.091 1*016 0, 947 *.069 s349 00908 ,.064 1.103 1.120 12333 L> 73 )0899 -.913 1.059 i.053 1.080 1.0339 .594 0.903 4.055 1.111 1.026 0.897 1.119 61±0 -.094 1043 0.889 1.053 1.031 1.oi8 1 eCo " R(i. 56) 386 3.1,7 3).49 .. 515 0.089 0.069 0.071 0.075

0 65 0. 54 .9 '.49 C.74 0.68 0.65 .086 0.79 3.89 0 * 74 C0*69 0.64 0.84 1 03 l.00 1.00 1.0 i.00 1,08 _038 0.97 0.99 1.06 1.05 1 1.11 09 ,.93 4.95 ,Ll.O i.W5 1.05

ALLE"DE - NO ALT ('NS) cjv33V GROSNAJA - (,S) (LT 200M) TVJ POA GROSNAJA - (NS) (1300-2C M) TVJ C3V GRSNAJA - (NS) (32-0%P,) TVJ C3V LEOVILLE - AS ( ,S) CIV MOKOIA ASU (NS) C3V VIGARANO (NSJ C3V -AVERAGE FILTER- RESPOSES 355

A(III-23

CHONDRITES - TYPE C3V

FILTER 9

2.360 3.383 :0.75 0,402 0.058 0.434 0.468 0.039

0.503 Is0^8 0.533 5... 0 01 0,566 0.599 G0015 C.632 1.023 0.665 0.699 0.43 0,729 G0.763 :.053

0,906 U.947 0.059 . C23 :0 C63 1053 ,057 1.10; 3.071

R(0.561 3.017

3.07 i;,5 bo04 £. .

9 AVRIAGE VARIATION