Skip to main content

Full text of "Interpretaion Of Ore Textures"

See other formats





Call No, 5% U \ ^^T? Accession No. 

Author &a~sCc<w / ^ , S. 

Title \%^v\pvc>cJLEc^ c^. ove. tU4^ Y 4L&. 

This book should be returned on or before the date lasi marked below 

The Geological Society of America 
Memoir 45 




Ithaca, New York 

Reprinted 1953 

Made in United States of America 



Address All Communications to The Geological Society of Americ 
419 West 117 Street, New York 27, N. Y. 

The Memoir Series 

The Geological Society of America 

is made possible 

through the bequest of 

Richard Alexander Fullerton Penrose, Jr. 


Since the happy days of apprenticeship in Economic Geology on the U. S. Geologi- 
cal Survey when under the inspiring guidance of Waldemar Lindgren and Frederick 
L. Ransome he first acquired an interest in the application of geology to mining, the 
writer has been especially fascinated by the interpretation of ore textures under 
the microscope. Through twenty-five years of subsequent teaching of Economic 
Geology in the laboratory at The University of Chicago this interest has increased, 
as he has realized more and more fully how much the critical interpretation of the 
textures of ores as seen under the microscope contributes to a sound diagnosis of the 
origin of the ore deposits. 

In the study and interpretation of microscopic ore textures, as in most relatively 
new fields, a certain period of seasoning must be passed through before sufficient 
agreement develops among the research workers to warrant the preparation of a 
text. In the present case the writer believes that season has been sufficient and that 
the needs of students warrant the presentation in a brief book of the more salient 
information on ore textures now mainly scattered through the technical journals and 
governmental reports. Widespread agreement does not yet exist with respect to the 
significance of all ore textures. The book is intended to be not a manual but a labora- 
tory aid to graduate and advanced undergraduate students in their microscopic 
studies of ores, and the writer hopes it may also be of aid to research workers. 

In this as in most textbooks the function of the author is to an important degree 
that of a critical editor in dealing with the contributions of a host of fellow workers. 
In a very real sense therefore this book has many authors; as is natural, however, 
the writer has drawn heavily upon his own experience. In 1931 a committee on 
Mineral Paragenesis of the National Research Council prepared a report (Bastin 
et al. t 1931) that has been of particular aid in the preparation of this volume. To the 
other members of this committee, L. C. Graton, the late Waldemar Lindgren, W. H. 
Newhouse, G. M. Schwartz, and M. N. Short, the writer acknowledges his especial 

The writer has depended as heavily Upon illustrations as upon words having in mind 
a recent remark of President Stoddard of the University of Illinois: "A Principle is 
not really understood in the absence of examples." 

Pioneer efforts in a field such as this are sure to contain errors and omissions. 
The writer hopes his colleagues in the field of Ore Deposits will freely offer suggestions 
that may make any later book on this subject more nearly complete. 

During the first stages of critical editing of this book there has appeared from the 
other side of the world a book entitled Textures of the ore minerals and their signifi- 
cance by Dr. A. B. Edwards of the University of Melbourne, Australia (1947). The 
plan of organization of Edwards' excellent book is sufficiently different from that of 
the present book so that the two volumes should supplement each other usefully 
without undue duplications or conflicts of viewpoint. 


Ithaca, New York 

February 1, 1950. 



INTRODUCTION . . . ... ... 1 


General considerations . .... .... 3 

Common textural terms . . .... .... ... ... . 4 

Order of crystallization . . .... ... .... 7 

Textures resulting from late magmatic changes . . ... 8 

Textures produced by changes in equilibrium after the magmatic stage . . . . . 10 

Selected references . . . . . 13 


General . . . 14 

Gratification . . . . 14 
Cockade or ring structure . .17 

Fillings of minor fractures . . . ... 17 

Pore fillings .... . . .... 19 

Plant cell fillings ... . . . .... 19 


General . . . . . 20 

Buoyancy . . . 21 

Brownian movement ... . . 21 

Adsorption .... . . . 21 

Electrical charge . . 21 

Chemical reactivity . . . 22 

Formation of colloidal suspensions . . . 22 

Suspensoids and emulsoids . . . . ... 23 

Stabilizing agents . . . . ... 23 


General . ....... 25 

Rotund or spheroidal forms . 27 

Syneresis or shrinkage cracks ... .29 

Pellet texture 30 

Framboidal texture . . . . . ... 30 

Interference surfaces . . . ... 31 

Selected references . . 32 


General . . . . 33 

Pseudomorphic replacement textures 34 

Transecting or cross-cutting textures 37 

Guided penetration textures . 38 

General description 38 

Parallel orientation patterns . . . . 41 

Diffuse penetration and automorphic replacement textures 41 

General description 41 

Skeletal crystals or dendrites . . 44 

Selected references . 46 

Rythmic or diffusion banding 47 

Selective and nonselective replacement . . . 51 

Graphic replacement textures 51 

Large disparity in size between guest and host . 52 

Form of the guest-host contact . .... 52 

Hypogene versus supergene replacement . . . 52 

Synthetic replacements 53 





Solid inclusions . . . 56 

Fluid inclusions. ... ... . . 57 


Indirect geological or geochemical evidence . . . 60 

Automorphic outlines . .... . .... 60 

Partial automorphism in two intergrown minerals . . . 61 

Gravitative control in ore deposition . .... ... 61 

Simultaneous deposition ... ... . . . . ... 62 

Paragenetic diagrams . .... 63 


General . . ... 65 

Oolitic textures in ores. ....... . . . .65 

General ... 65 

Oolitic iron ores of France . . . . . 65 

Cleveland iron ores of England . . ... 69 

OSlitic iron ores of Newfoundland 69 

OGlitic iron ores of the United States . . . ... ... 70 

Evidences of algal and bacterial remains .... . . . 72 







Plate Facing page 

1. Textures of magmatic ores; pyrite in sandstone 80 

2. Textures of magmatic ores. . . 81 

3. Textures due to deposition in open spaces; exsolution texture 82 

4. Textures due to fracture filling 83 

5. Fillings of ruptures and of plant cells; colloidal textures . . . 84 

6. Colloidal textures 85 

7. Colloidal textures ... . 86 

8. Pseudomorphic replacement textures 87 

9. Replacement textures 88 

10. Replacements guided by mineral contacts 89 

11. Diffusion textures, mainly in silver ores 90 

12. Diffusion textures in silver ores 91 

13. Replacement of silver dendrites by calcite 92 

14. Replacement of silver dendrites; diffusion banding 93 

15. Diffusion banding; graphic replacement textures; synthetic replacements 94 

16. Synthetic replacements; inclusions formed by replacement 95 

17. Textures of oOlitic iron ores; pyrite; iron bacteria .... 96 

18. OQlitic iron ores; dynamo-metamorphic textures 97 


Figure Page 

1. Micrographic intergrowth of olivine and titanomagnetite 5 

2. Orbicular chromite 6 



3. Nodular chromite ore . . . . . . . 7 

4. Chain texture in chromite ore ... . . . . 8 

5. Corrosion of orthoclase phenocryst by chalcopyrite . 9 

6. Late pyrrhotite in magmatic ore. . . . .9 

7. Reaction rims in olivine gabbro . . . . . 11 

8. Crustification in vein from Newfoundland . . . .15 

9. Crustification in ore from Aachen, Germany . 16 

10. Fracture filling in ore deposition . . . . 18 

11. Colloform pitchblende. .. . . .... 28 

12. Replacement of galena along cleavages . 36 

13. Replacement of an igneous rock by stephanite . . 36 

14. Guided replacement of galena. . . 39 

15. Successive replacement of galena by two minerals . . 39 

16. Replacement of sphalerite by argentite and covellite . 40 

17. Complex replacement of galena . . 40 

18. Native silver replacing pearcite . . .42 

19. Diagram illustrating relation of diffusion banding to fractures . . 49 

20. Inclusions of sphalerite in quartz and freibergite . . . . 57 

21. Inclusions of chalcopyrite, petroleum, and marcasite in fluorite . . 58 

22. Diagram illustrating significance of protruding crystal boundaries . 61 

23. Diagram illustrating significance of depressed crystal boundaries ... 61 

24. Paragenetic diagram illustrating preferred usage of terms simultaneity, overlap, and suc- 

cessive deposition . . 62 

25. Paragenetic diagram, 1 Bote mine, Zacatecas, Mexico . . . . 63 

26. Composite paragenetic diagram for the Zacatecas district, Mexico 63 

27. Automorphic replacement of iron silicate oolites by siderite 71 

28. Contorted banding in ore from Rammelsberg, Germany 74 


The term texture as used here refers to the smaller features of ores which are de- 
pendent on the size, forms, and arrangement of the component minerals and in some 
cases upon such features as mineral cleavage, mineral parting, mineral contacts, 
and minor fracturing. These features can be observed in hand specimens either with 
the unaided eye or with the microscope. For those larger features of ore deposits 
that are dependent upon folding, faulting, jointing, bedding, etc., the term structure 
is applied. Since the distinction between texture and structure is in important part 
one of size, the writer reserves some latitude in the descriptions. This usage of the 
terms texture and structure conforms to that commonly adopted by petrographers, 
and this is desirable since ores are merely unusual rock types. 

Some textural terms well established among petrographers, are also appropriately 
applicable to ore textures and have been widely used in ore descriptions. Petrographic 
terms used in describing the shapes of crystals are: 

Idiomorphic or euhedral if the crystal contours are more or less perfectly developed. 

Xenomorphic or anhedral if crystal contours are not developed. 

Hypidiomorpkic or subhedral if crystal contours are partially developed. 

In all interpretations of ore textures it is important that the texture be visualized 
in three dimensions, although a single thin section or polished section displays only 
two dimensions. In magmatic ores of granular texture it is usually safe to assume 
that the third dimension is similar to the other two, but in magmatic ores showing 
either graphic or flow textures several sections cut in diverse directions may be 
heeded for complete textural interpretation. In ores of pneumatolytic and hydro- 
thermal origins directional differences in texture may be considerable and significant. 
For example, in one section mineral B may appear to be a simple inclusion in A , but 
when the third dimension is studied the "inclusion" may be found to represent a 
prong from a large area of mineral B projecting into mineral A . 

The techniques of the preparation of polished specimens of ores and of the identifi- 
cation of metallic minerals in polished specimens by etching and other tests are basic 
to the interpretation of ore textures. In this country these techniques have been 
described in an excellent text (Short, 1940) which also includes a section on micro- 
chemical methods. In German there is the comprehensive and profusely illustrated 
text of Schneiderhohn and Ramdohr (1931), which deals also in part with the inter- 
pretation of ore textures. 

In Chapter I the Magmatic Ores are considered separately because they are a 
variant of the great class of Igneous Rocks. The textural terms established by pe- 
trographers for the commoner igneous rocks and applicable also to magmatic ores 
form an appropriate background for the consideragion of textures formed by solutions 
that were not magmas though many of them were hot. Adherence to the widely 
accepted distinction between magmas and hydrothermal solutions does not negate 
the possibility of transitional relations between them. It does emphasize certain 
significant contrasts between most magmas and most hydrothermal solutions. These 
contrasts are particularly great between hydrothermal solutions and the basic or 



ultrabasic magmas with which most magma tic ores are associated. The two most 
significant features of hydro thermal alteration are (1) the intimacy with which the 
altering solutions have penetrated the rocks, often for long distances, and (2) the 
hydrous nature of many of the new minerals formed by their action. From this it is 
logical to infer that the solutions were highly mobile or tenuous Wd were highly 
aqueous. Magmas do not penetrate the rocks they intrude with an intimacy at all 
comparable, and there is ample evidence that the basic and ultrabasic magmas 
with which most magmatic ores are associated are relatively poor in water. 

Ores precipitated from hydro thermal solutions may be deposited (1) by precipita- 
tion in pre-existing open spaces, (2) by replacement, or (3) by a combination of 
these two processes. 



Magmatic deposits are unique among ore deposits in being direct crystallizations 
from magmas, usually deep-seated (plu tonic), rarely volcanic. It has long been recog- 
nized that magmas are complex mutual solutions and obey the laws of solutions 
(Grout, 1932, p. 209-210. Some geologists (Spurr, 192J) have emphasized the sup- 
posed igneous relations of the tenuous and highly aqueous solutions from which 
many ores were deposited by terming them "Ore Magmas." It seems more useful, 
however, to stress the contrasts between a magma, such as a lava flow, and a hot 
tenuous mineralizing solution that may penetrate a body of solid rock for tens or 
hundreds of feet replacing some or all of its minerals by new ones. Most magmatic 
ores are differentiation products from basic magmas, and, as these are character- 
istically poor in volatile components, we should expect that pneumatolytic and 
hydrothermal processes would be lacking or rare in association with such deposits. 
This appears to be the case. 

Ores can be proved to be magmatic only if it can be shown that their valuable 
metallic minerals were direct crystallizations from magmas and not later perhaps 
much later replacements of solid igneous rocks accomplished by solutions from some 
external source. In such diagnosis larger fields relations may be more valuable than 
small textural features. Most magmatic ores occur within bodies of undoubted 
igneous rock and show transitions one to the other. 1 The microscope is usually of 
great aid in confirming such transitions. In any one district magmatic ores are 
usually restricted to a single type of igneous rock or at most to closely similar igneous 
types. They may occur in the peripheral or in the central portions or may be scattered 
through the mass. The mineral composition of the ore usually aids in diagnosis, for 
worldwide observation has shown that most magmatic ores are valuable mainly as 
sources of iron, nickel, copper, chromium, titanium, and the platinum group. Nega- 
tive criteria, such as the absence or rarity of hydrothermal alteration and other 
replacement phenomena, may be of confirmatory value but must be used with caution 
since certain high-temperature replacements may take place without destroying the 
fresh "unaltered" megascopic appearance of the rock. 

Ore textures seen under the microscope are seldom conclusively diagnostic of 
magmatic origin but may have confirmatory value. In a few ores microscopic study 
shows that the metallic minerals are essentially contemporaneous with certain non- 
metallic minerals such as olivine that are known to be magmatic. In the chromite 
ore for example, shown in Figure 1 of Plate 1, some chromite grains (black) show 
automorphic crystal faces against olivine (or its alteration rim of serpentine), but 
some of the oliyine also shows automorphic outlines against chromite. This reversible 
relation indicates that chromite and olivine are roughly contemporaneous. As the 
olivine is magmatic the chromite is also. 

1 Less commonly dikelike masses of ore that are believed to be magmatic traverse, with sharp contacts, rocks that 
may or may not be genetically related to the ere, as in the case of some magnetite dikes. 



While the metallic minerals are, in some magmatic ores, contemporaneous with 
the silicates, much more commonly they are] either earlier or later. Figure 2 of Plate 1 
shows an ore in which euhedral chromite is enclosed in olivine (partly altered to 
serpentine). Such a texture might be due to automorphic hydro thermal replacement 
of an olivine rock by chromite analogous to the replacement of schist by pyrite cubes 
shown in Figure 4 of Plate 8. If such replacement were of the diffuse type it would 
be difficult to rule out this possibility on textural grounds alone. Commonly, however, 
automorphic replacements betray their replacement origin by being controlled in 
their distribution by textural features of the host such as fractures, mineral cleavages, 
and mineral contacts. In the absence of specific evidences of replacement the alterna- 
tive is that the chromite is magmatic crystallizing first and then being enclosed by 
the characteristically magmatic mineral olivine. This is here the preferred explanation 
and is harmonious with the fact that only in a few rare occurrences has chromite 
been shown to be hydro thermal (Sampson, 1931; Ross, 1929). 

In other ores such as the sulphide-bearing diabase (PL 2, fig. 1) the metallic 
components are the youngest and form the matrix between the silicates. The silicates 
are unaltered, and the metallic minerals pyrrhotite, chalcopyrite, and magnetite 
fill the interspaces around the plagioclase laths and clearly formed last. These rela- 
tions might be interpreted as the result of replacement of the silicates by the metallics 
through hydrothermal solutions from an outside source. However, the euhedral 
outlines of the silicates against the metallics indicate that such replacements did 
not take place. The alternative explanation, entertained by Schwartz (1925), is 
that the metallics like the silicates crystallized from the magma. 


As magmatic ores are igneous rocks, in general the accepted textural terms de- 
veloped for the description of the common igneous rocks may be applied also to 
magmatic ores. Only in a few instances does the great abundance of metallic minerals 
lead to textures not found in igneous rocks poor in metallics. Familiar textures of 
ordinary igneous rocks (Johannsen, 1931) that have also been recognized in magmatic 
ores are: 

(1) Granular 

(a) Xenomorphic 

(b) Hypautomorphic 

Poikilitic and Ophitic 

(2) Porphyritic 

(3) Micrographic 

(4) Banded textures 

Flow banding (fluidal or fluxion) 
Banding due to gravitative settling 

(5) Orbicular and nodular textures 

(6) Synneusis 

Xenomorphic (or Allotriomorphic) granular. Crystal grains are without facets 
and tend to be fairly uniform in size. This texture is seldom well developed in mag- 


matic ores but is approximated in the chromite ore shown in Figure 1 of Plate 1 
in which olivine and chromite indent each other and in which crystal facets are rare 
and rounded grain outlines prevail. 

Hypautomorphic (or Hypidiomorphic) granular. Some of the crystals show facets. 
The chromite ore pictured in Figure 2 of Plate 1 well illustrates this texture, as does 
also the sulphide diabase shown in Figure 1 of Plate 2. 

FIGURE 1. Micrographic inter growth of olivine (white} and titanomagnetite (black} 
From Foskaa, Norway. X 20. After Vogt. 

Poikililic granular. Optically continuous crystals of one mineral form the matrix 
for diversely oriented crystals of another mineral. 

Ophitic granular. A variety of poikilitic texture in which the diversely oriented 
crystals are plagioclase laths. 

Porphyritic. Large automorphic crystals (phenocrysts) of one or more minerals 
in a finer groundmass; in ores the groundmass appears to be always crystalline. 

Graphic and Micrographic. Two minerals mutually so intergrown as to resemble 
ancient cuneiform writing. Some micrographic intergrowths tend to radiate from a 
common center. In igneous rocks the minerals of such intergrowths are usually re- 
garded as contemporaneous. Micrographic intergrowths of olivine and titanomag- 
netite are shown in Figure 1, and a micrographic intergrowth of platinum and 
osmium is shown in Figure 3 of Plate 1. Both are unquestionably magmatic. 

Banded textures. Banding due to dynamometamorphism subsequent to ore dep- 
osition will be considered in a later section but banding is fairly common in ores 
unaltered by metamorphism, particularly in ores of iron and chromium. In many 
cases such banding is clearly the result of flowage in the magma during crystallization. 
An excellent example is given by Takeo Kato (1921) from the Akaishi mine in Japan 
in Figure 5 of Plate 1. In this ore layers of chromite alternate with olivine. The 
microscope reveals that many of the crystals in the chromite layers have been crushed 
along their borders but that the crushed-off grains have also been corroded. Kato 


concludes that "the banded ore was formed by a flowing motion of the crystallizing 
dunite magma, in which settling of the chromite crystals was going on." Banded 
textures due to flowage are also well developed in the magnetite ores of Kiruna, 

FIGURE 2. Orbicular chromite 

Tracing from a polished slab. Octopus mine, Siskiyou County, Calif. Black is chromite. White is olivine partly al- 
tered to serpentine. Scale in centimeters. After W. D. Johnston, Jr. 

In some chromite ores individual bands are so straight and persistent for long 
distances that the rocks at first sight resemble sediments. Johnston and de Souza 
(1943) have described such ores from Brazil (PL 1, fig. 4). At one mine 93 layers of 
chromite occur in a thickness of 20 meters. The chromite bands range from a few 
millimeters to a meter or more in thickness, and there is little or no regularity in the 
thickness or spacing of the bands such as is found in diffusion banding or in varves. 
At some of the Brazilian mines the bands are horizontal; elsewhere, as a result of 
faulting, they are steeply inclined or even vertical. Chromite bands never cut across 
olivine bands or vice versa. Johnston and de Souza attribute the banding to variable 
gravitative settling of both olivine and chromite crystals during cooling and crystal- 
lization of the sill-like bodies of magma. Under this concept all of the bands were 
originally nearly horizontal. Many of the chromite deposits of the Bushveld complex 
in the Transvaal are also flat-lying and bedlike though clearly magmatic. Banded 
chromite ores have also been described from Siskiyou County, California (Johnston, 
1936, p. 417-427). 

Orbicular and Nodular Textures. Textures resembling the orbicular forms oc- 
casionally found in the commoner types of igneous rocks have been described (John- 
ston, 1936, p. 418-426) in chromite ores from Siskiyou County, California (Fig. 2). 
Johnston reports (p. 418HU9) 

"The orbicular ore is made up of many spheroidal units, each of which consists of a nucleus of 
chromite surrounded by a shell with drawn-out pointed ends, composed, in the main, of partially 
serpentinized olivine. Each nucleus is composed of a number of chromite crystals. . . . Each sheath 
of olivine is surrounded by a thin outer sheath of chromite and commonly contains several concentric 
layers of chromite. Larger masses of chromite occupy the interstices between abutting units. In a 
few units, the chromite nucleus is absent. . . . The drawn-out spheroids are fitted together in parallel 
orientation. The elongation of the orbicular ore is linear and not platy, for cross-sections normal to 
the long axis of the units show only circular patterns.'* 


The parallel elongation of the spheroids is attributed to flowage before crystallization 
was complete. 

In the orbicules chromite nuclei appear to have acted as growth centers for olivine, 
and the interstitial chromite followed the olivine crystallization. 

Nodular structures are much more common than orbicular structures and have 
been described in chromite ores from Oregon (Fig. 3), California, Cuba, and Quebec. 

FIGURE 3. Nodular chromite ore 
Briggs Creek, Josephine County, Oregon. White matrix is serpentine. Scale in centimeters. After Diller and Johnston . 

These nodules have been interpreted as nuclei to which chromite was attracted 
during crystallization. 

Synneusis and Chain Structure. Synneusis (Vogt, 1921) is peculiar to igneous rocks 
and to the magmatic stage hi their development. Vogt (p. 321) states: 

"The individuals of a mineral, segregated from a magma at an early stage, frequently swam to- 
gether to assemblings or aggregates, the result of which is a structure, for which I propose the term 
together-swimming structure or synneusis structure." 

Vogt observed the texture not only in ores but in slags. In ores synneusis is a useful 
criterion of magmatic origin and moreover the mineral involved is early magmatic. 
Vogt's illustration of synneusis texture is reproduced as Figure 3 of Plate 2. On 
textural evidence alone this relationship might be explained as an automorphic 
replacement of olivine by chromite. However, fractures in the olivine and contacts 
between olivine grains exerted no control on the chromite distribution. It is more 
likely that olivine was still liquid when the chromite crystallized as Vogt concluded. 

The chain texture described by Sampson (1932, p. 130) from chromite ores in 
South Africa (Fig. 4) may be a variety of synneusis texture. Sampson does not reach 
a final conclusion as to its origin. The prevailingly automorphic outlines of the 
chromite plus the fact that in some specimens the chromite grains and chains have 
clearly been corroded seem to show that the chromite is older than the enclosing 


It should thus be clear that in ores that are products of normal crystallization 
from magmatic liquid solutions it is commonly possible to infer the order of crystal- 


lization of the several minerals from their textural relations. The principal criteria 
are (1) degree of automorphism of various minerals, (2) inclusions, (3) graphic 
intergrowths. (4) Synneusis texture, though rare, may also be of value. If the rock 
texture is xenomorphic granular (Fig. 1) the minerals are interpreted as essentially 
contemporaneous. In magmatic ores showing graphic or micrographic textures 

FIGURE 4. Chain texture in chromite ore 
Enclosed by labradorite. Forest Hill platinum mine, Bushvelt, S. Africa. X 17. After Sampson. 

(PL 1, fig. 3), the components of the intergrowth are generally interpreted as con- 
temporaneous. In ores that are not magmatic, graphic textures may be formed by 
quite different processes, and the two component^ may or may not be contempo- 

If the texture is porphyritic or hypautomorphic granular (including the poikilitic 
and ophitic), minerals completely or largely enclosed by others are interpreted a.s 
older than the enclosing minerals. Usually the enclosed mineral is euhedral or sub- 
hedral. Thus, for example, the phenocrysts of a porphyry are older than the minerals 
of the matrix, and the subhedral crystals of chromite in the chromite ore shown in 
Figure 2 of Plate 1 are older than the olivine that conforms to their crystal outlines. 


A crystallizing magma is a system characterized by changing conditions of equi- 
librium in the still liquid portions. Such changes in equilibirum may^lead to changes in 
the composition of the minerals deposited and/or to re-solution of minerals already 
formed. The most familiar example of depositional changes in composition is zoning 
in plagioclase feldspars; the outer zones usually are progressively richer in soda. 
Comparable phenomena in the metallic minerals of ores when of clearly magmatic 
origin seem not to be common. Zoning has, however, been observed in platinum 
grains hi the Urals, in which sharply demarked outer and inner zones differ notably 
in the amount of iron alloyed with the platinum. 

One of the most familiar examples of re-solution is the corrosion of quartz pheno- 
crysts observed in porphyries. Analogous phenomena are occasionally observed in 


FIGURE 5. Corrosion of orthodase (0) phenocryst by chalcopyrite (black) 
In chalcopyrite syenite from Park County, Montana. X 65. After Lovering. 


FIGURE 6. Late pyrrhotite (black) in magmatic ore 

Penetrating earlier labradorite (Labr), hypersthene (Hyp), and quartz (Qu) along fractures. Quartz norite from Rom- 
saas, Norway. X 16. After Vogt. There is no evidence of corrosion of the silicates by pyrrhotite. 

ores as for example the corrosion of feldspar phenocrysts by chalcopyrite (Fig. 5). 
Figure 2 of Plate 2 shows ore from Feragen, Norway, in which rounded and embayed 
chromite grains are surrounded by serpentine (after olivine). Donath (Abb. 20) 
interprets this as corrosion of chromite by still liquid olivine. 

Movements within the crystallizing magma may produce minor fracturing of 
crystals, and into these fractures the still molten components enter and crystallize 
forming veinlets of late magmatic minerals in the earlier magmatic minerals. Vogt 
(p. 643) describes this in a norite from Romsaas, Norway (Fig. 6). There is no tex- 


tural evidence of either corrosion or replacement of the silicates, and Vogt in- 
terprets the relations as a late magmatic injection of pyrrhotite. 

Although corrosion or resorption phenomena have been referred to as "replacement 
by resorption" the writer believes the magmatic conditions under which corrosion 
takes place differ sufficiently from those characteristic of replacement to justify 
separate terms. These differences will be more apparent after we have discussed 
replacement phenomena. 


After complete crystallization of a magma there may still be changes in equi- 
librium due mainly to the continued cooling of the rock. In magmatic ores, nearly 
all of which are plutonic rocks, cooling is usually slow, and minor textural modifi- 
cations may take place in the solid state with the formation of new minerals. 

Reaction Rims. One of the best examples of such changes is furnished by reaction 
rims, usually recognizable only under the microscope (Fig. 7). They are restricted, so 
far as known, to basic plutonic rocks and are fairly common in such rocks and the 
associated ores. They are particularly well developed between calcic plagioclase 
and olivine and in such situations are double, one rim having developed at the 
expense of plagioclase and the other at the expense of olivine. The boundary between 
the two rims is smooth and in places straight, and it clearly marks the original 
boundary of olivine against plagioclase. Between plagioclase and metallic minerals 
such as magnetite or pyrrhotite there is only a single reaction rim which has developed 
mainly at the expense of the plagioclase. Reaction rims are lacking between chemically 
similar minerals such as olivine and pyroxene. In contrast to replacements they never 
occur between grains of the same mineral species or along fractures or cleavages in a 
single mineral. 

Such rims are interpreted as phenomena of the solid stage immediately following 
complete crystallization from the magma. They record an interchange of substance 
between certain minerals of notably contrasting composition in response to declining 
temperatures; the new mineral or minerals being more stable at the lower temper- 
atures than were the original minerals. The new minerals are intermediate in com- 
position between the minerals from which they were formed. There is no evidence 
of addition of material from outside as in the process of replacement with which 
such reactions must not be confused. 

Unmixing or ex-solution textures. Solid solutions are common and are well exempli- 
fied among the common rock-forming minerals by the plagioclase feldspars which 
form a continuous isomorphous series from albite to anorthite. The student of ore 
deposits is of course concerned mainly with solid solutions of metallic minerals. 
However, the concepts of solid solutions and their unmixing seem first to have been 
developed by the metallurgist and later extended to the study of rock minerals and 
the metallic minerals of ores. It may be well, therefore, to consider first what the 
metallurgist means by a solid solution in alloys. If a melted alloy of copper with a 
relatively small amount of zinc is allowed to cool and solidify, microscopic examination 



shows that it is apparently homogeneous and that its properties differ only slightly 
from those of pure copper. If a series of other melted copper alloys with increasing 
proportions of zinc (but not exceeding 35 per cent) are prepared each alloy upon 
solidifying is homogeneous, and their properties depart from those of pure copper in 

FIGURE 7. Reaction rims in olivine gabbro 

From Lango, Norway. Between magnetite (black) and labradorite (ruled) there is a single reaction rim of hornblende. 
Between olivine (Ol) and labradorite the reaction rim is double; the part bordering olivine is hypersthene, and that bor- 
dering labradorite hornblende. Sp is serpentine. X 28. After J. H. L. Vogt. 

progressively greater degree as the zinc content increases. Such homogeneous solid 
alloys whose properties vary progressively with varying composition are termed 
solid solution. 

The limits of solid solubility in alloys vary with the nature of the components, 
with temperature, and probably also with pressure. In general the solubility in- 
creases with increase in temperature although a few exceptions are known. If, there- 
fore, a solid solution is near saturation with respect to one component at temperature 
T and the temperature declines to /, a separation of that component may take place. 
This is known as unmixing or exsolution. In alloys one or both components resulting 
from unmixing may be (a) metals, (b) compounds, or (c) solid solutions. The textures 
developed are varied but characteristic. 

The interpretation in recent years of certain textures observed in ores as due to 
unmixing was based on the close resemblance of such ore textures to those that 
result from unmixing in alloys. The argument from analogy is particularly close in 
the case of those magmatic ores which solidified from relatively dry melts. Origin 


by unmixing of certain of these ore textures has been verified experimentally. Thus 
certain ore intergrowths of magnetite and ilmenite if heated above 800C. merge 
to form a homogeneous solid. 

In ore deposits unmixing phenomena have been observed in certain members of 
four mineralogic classes the native metals, the oxides of the metals, sulpho-salts of 
the metals, and sulphides of the metals. Some solid solutions occurring in ores may 
show no unmixing textures if the natural changes in temperature and pressure have 
not transgressed the limits of solid solubility. In such cases only analysis will reveal 
the mixed character of the minerals. This is notably true, for example, with natural 
solid solutions of gold and silver. 

In magmatic ores, unmixing of solid solutions is a phenomenon of the post-mag- 
matic stage when initially high temperatures have declined through a considerable 

Presumably well-authenticated instances of unmixing in magmatic ore deposits 
involve the following pairs of minerals. 

(1) Magnetite Ilmenite 

FeO Fe 2 O 3 mFeTiOs nFe 2 O 8 

(2) Hematite Ilmenite 

Fe 2 O 8 mFeTiOa nFe 2 O 8 

(3) Chalcopyrite Cubanite 

Cu 2 S Fe 2 S 8 CuS Fe 2 S 8 

(4) Pyrrhotite Pentlandite 

FeS FeNiS 

The commonest type of unmixing texture is the grating type (PL 2, fig. 4) in 
which the crystal structure of one constituent controls the distribution of the other. 
The "inclusions" may be needles, blades, plates, or blebs. 

The following criteria seem valuable in the recognition of unmixing textures in 
ores (Schwartz, 1931). 

(1) The first criterion is mineralogic rather than textual. The products of un- 
mixing are usually minerals possessing certain mineralogic or chemical similarity. In 
the plagioclase feldspars the components are chemically analogous and are iso- 
morphous. Gold and silver are both isometric, crystallize in a face-centered lattice, 
and are similar in many of their properties. Arsenic and antimony are both rhombo- 
hedral and have similar crystal structures. Among the oxides and sulphides listed 
above as occurring as unmixing products in magmatic ores magnetite and ilmenitc 
though crystallizing in different systems are chemically similar except that ilmenite 
contains TiO 2 . Hematite and ilmenite are both rhombohedral and differ chemically 
only in that ilmenite contains titanium. Chalcopyrite and cubanite though belonging 
to different crystal systems are chemically very similar. Pyrrhotite and pentlandite 
though crystallizing in different systems differ chemically only in that pentlandite 
contains nickel. Solid solutions and ex-solution phenomena seem never to have been 


demonstrated between members of different mineral groups such as oxides and 
sulphides or native metals and sulphides. It has been suggested that in some instances 
gold, in particles too small to be seen under the microscope, is in solid solution with 
the pyrite, but such a relationship has not been demonstrated (Edwards, 1947, 
p. 92-93). 

(2) The inclusions may be blade-, plate-, rod- or bleb-like. There is no enlargement 
where the blades or plates join or cross; on the contrary many blades contract in 
width at intersections (PL 3, fig. 1), Replacement veinlets on the other hand charac- 
teristically enlarge at their intersections. (Compare Figure 6 of Plate 9.) 

(3) The boundaries of the included mineral are always sharp and smooth. Bounda- 
ries formed by replacement may be either smooth or ragged. 

(4) The included mineral is commonly absent outside the intergrowths. 

(5) The included mineral tends to be rather evenly distributed. 

(6) The included mineral usually occurs as disconnected units instead of forming 
a network as in replacement along mineral cleavages. 

(7) Orientation of the included mineral differs for each grain of the enclosing min- 
eral depending upon the crystal orientation of the latter. 

(8) The orientation of the included mineral is not related to the grain boundaries 
of the enclosing mineral (PL 2, fig. 6). 

(9) In general the absence of evidence of replacement indicates the possibility of 

Several of these relations must be recognizable in order to make a safe diagnosis. 
Exsolution phenomena in ores formed hydrothermally are considered later in 
this paper. 2 


References of particular importance as bearing on magmatic ore textures are 
listed in the end bibliography as follows: Per Geijer (1910), Schwartz (1931b), 
Singewald x (1913), Vogt (1921), Warren (1918). 

* For a more detailed discussion of solid solutions in ores the student is referred to Chaper IV In Edwards: Tex- 
tures of the ore minerals. This chapter includes lists of (1) oxides, (2) sulpho-salts. and (3) sulphides that have been 
shown experimentally to form solid solutions or in which unmixing has been inferred from textures. 



In Chapters II to IV textures indicative of deposition in rock openings are con- 
sidered with respect to deposition of (1) crystalline and (2) microcyrstalline and 
amorphous substances. Although one type of deposition usually dominates, both 
crystalline and amorphous substances differing in kind may be deposited simul- 
taneously or in quick sequence. 

Almost any type of opening in rocks may become the locus of ore deposition. 
Ores have been found in the following openings. 

(1) Pores (in sediments and pyroclastics mainly) 

(2) Vesicles 

(3) Plant cells 

(4) Miarolitic cavities 

(5) Solution cavities 

(6) Fissures 

(7) Faults 

(8) Joints 

(9) Bedding-plane fractures and "flats and pitches" 

(10) Tension fractures in folds 

(11) "Ladder vein" fractures 

(12) Irregular minor fractures 

Only certain of these require discussion from the standpoint of the textures developed 
in the ores. 

As a general rule minerals deposited hi open spaces are younger than those forming 
the walls of such spaces. There may, however, be occasional exceptions. For example 
the minerals of a fissure filling in granite are younger than the minerals of the granite, 
but if the granite has been altered hydrothermally they are not necessarily younger 
than these alteration products. 

If the fillings show layering or "crustification" they may be very significant in 
indicating the sequence of mineral deposition. 


Crustification is the deposition of mineral matter in successive layers or crusts 
upon the walls of openings in rocks. There may be many such layers, and usually 
the layer next the wall was deposited first. 8 In fissures the bands are symmetrical in 
kind on opposite walls but not necessarily symmetrical in width. The youngest 
bands may meet in the center of the fissure, or unfilled vugs, usually lenticular, may 
remain. Minerals that tend to crystallize in elongate forms, notably quartz, are 
often arranged with their long axes perpendicular to the base of the band, producing 

Reopened veins often have a new layer next one or both walls. 




what is kndwn as comb structure. There is no better illustration of (Trustification than 
the familiar one of von Weissenbach first published in 1836 and shown in Figure 3 
of Plate 3. This vein is unusual, however, in the large number of bands present. 

/ 13456 7 665 4 

FIGURE 8. Crustification in vein on Port au Port Peninsula, Newfoundland 

1. Massive limestone, 2. Limestone traversed by calcite seams, 3. Limestone breccia cemented by calcite, 4. Galena* 
5. Colloform sphalerite, 6. Mainly marcasite, 7. Calcite showing comb structure, 8. Vug. After K. deP. Watson. 

Such complete symmetry in kind between the bands of opposite walls appears to 
imply deposition from solutions that completely filled the fracture. 

The writer has observed typical crustification in veinlets as small as a centimeter 
in width though the bands were few in number. In veins of microscopic dimensions 
crustification is not observed, presumably because openings were quickly filled. 
Figure 8 illustrates this principle and also shows interesting variations from the 
regular crustification of Figure 3 of Plate 3. Some of the bands (notably 4 and 7) 
are crystalline, while others (notably 6) were deposited either in a colloidal or a 
microcrystalline state. Fillings of small fractures, while seldom showing crustification, 
occasionally show a segmented character. 


In some crustified ores many of the bands may be narrow and finely crystalline. 
This is notably true in the well-known "Schalenblende" zinc, ores of Aachen, Germany 
(PL 4, fig. 1). The texture of this specimen is a combination of crustification and 
cockade or ring structure (top portion). The banded layers are mainly sphalerite in 
varying shades, but in many specimens there are occasional bands of pyrite. The 

FIGURE 9. Cruslification in ore from Aachen, Germany 

Camera lucida drawings from parts of the specimen shown in Figure 1 of Plate 4, showing galena crystals deposited on 
curved surfaces of an older sphalerite band and enveloped by a younger sphalerite band whose surface curvature is con- 
trolled by the forms of the galena crystals. After Bastin. 

sphalerite bands show little or no regularity either in width or in spacing. Study by 
the writer of typical specimens in the collections of Cornell University shows that the 
sphalerite is crystalline showing, under low magnification, cleavage faces when 
freshly fractured. The cleavage faces show that the sphalerite crystals are oriented 
nearly perpendicular to the banding and range up to 0.1 mm in length. The oc- 
casional pyrite bands are much more coarsely crystalline. The bladelike crystals of 
pyrite are oriented perpendicular to the banding and range from 2 to 5 mm long; 
a few extend the full width of the band. Ehrenberg (1931), using hydrogen iodide as 
an etching reagent to distinguish sphalerite from wurtzite, has shown that much of 
the sphalerite is pseudomorphic after original crystalline wurtzite (hexagonal ZnS). 
Lengths of single wurtzite crystals range from 0.5 to 0.8 mm. 

Scattered crystals of galena up to 5 mm in diameter cap or are embedded in a few 
of the sphalerite bands as shown in Figure 1 of Plate 4 and more clearly in Figure 9 
which shows camera-lucida sketches from the same specimen. The bottoms of many 
of these galena crystals conform to the curved outline of the underlying sphalerite 
band. The galena crystals grew therefore on the curving free upper surface of a 
sphalerite band. The succeeding band of sphalerite enveloped the galena crystals, 
and the curvature of its upper surface was controlled in part by the shapes of the 
galena crystals. 

The textures exhibited by the Schalenblende of Aachen somewhat resemble those 
produced by diffusion into colloids (compare PL 5, fig. 4), but the regularity in spacing 
characteristic of diffusion banding is absent, and the curvature of the sphalerite 
bands depends not on surface-tension phenomena but upon the shape of the underly- 
ing surface. In the specimens seen by the writer or described in the literature there 
seems to be no evidence that even the finest of the mineralssphalerite was de- 


posited as a colloid. Shrinkage cracks so common in ores deposited as colloids appear 
to be absent. It is important to distinguish such finely crystalline crustification from 
the diffusion banding of colloids. 


This structure is a variant of crustification. Where a fissure becomes partly filled 
with a rubble of fragments torn or dropped from the walls these may become crusted 
with successive layers of mineral matter (PI. 3, figs. 5, 6). Typical occurrences in the 
Harz Mountains show the same succession of mineral bands on the walls and around 
many of the fragments. The absence of the oldest bands around certain fragments is 
explainable if these fell from the walls after mineralization had begun. In some 
cockade ore many fragments seem not to be in contact with other fragments or with 
the walls of the cavity. This absence (?) of support is probably apparent rather than 
real (Talmage, 1929). Also new minerals coating fragments may by their force of 
crystallization produce a moderate amount of dislodgement (Beck, 1909, p. 248, 
Emmons, 1940, p. 73). 


The filling of minor fractures is an important mode of mineralization; as replace- 
ment may also follow minor fractures it is important to distinguish the two processes. 

Openings formed by fracturing may be very irregular because of brecciation or 
somewhat regular as a result of fissuring. In either case the filling has sharply defined 
though often sinuous boundaries. Fissure fillings are commonly characterized by 
matching walls that is, by walls that would fit together at least moderately well if 
the filling were removed. Perfect matching is not always found, even in undoubted 
fracture fillings, since one wall may have moved in a direction inclined to the plane 
of the section available for study, and fragments torn from the wall in the fracturing 
may also have moved out of that plane. Typical fracture fillings are shown hi Figure 
10 and in Figure 2 of Plate 3. The prevalence of matched walls in fissure fillings is in 
marked contrast to the usual absence of matching in replacement veinlets such as 
those shown in Figures 3 and 4 of Plate 9. 

The fillings of minor fractures such as are observed under the microscope seldom 
if ever show crustification. The filling minerals are usually irregularly associated 
(Fig. 10), and it may not be possible to recognize age differences among them. 

In a few ores many of the small veinlets seen under the microscope are segmented 
veinlets; their component minerals alternating along the trend of the vein. Commonly 
only two vein minerals are involved, the vein at one place being composed entirely 
of mineral A while the succeeding segments are composed of mineral B. Occasionally 
a third mineral is present. The contacts between tie segmented vein minerals are 
sharp and usually are nearly at right angles to the vein walls. Segmented veinlets 
are common in the ores of the Cobalt and South Lorraine districts, Ontario, where 
the segments are commonly silver and calcite (PL 4, fig. 3), though in some cases 
silver and niccolite (PL 4, fig. 2). Sharwood (1911) has figured a segmented veinlet 
of gold and pyrrhotite filling a fracture hi arsenopyrite in the Homestake Mine. 


Many of the segmented veinlets of calcite and native silver in the ores of Cobalt 
and neighboring camps in Ontario, Canada, are unquestionably the result of late 
hypogene replacement of portions of the silver by calcite. The evidence of such an 
origin is conclusive in the ores of Gowganda and is identical in character with the 

FIGURE 10. Fracture filling in ore deposition 

Fractures in pyrite filled with quartz, sphalerite (SPH.) and a little chalcopyrite. The three minerals of the filling 
seem to be contemporaneous, and there is no evidence that any of them have replaced the pyrite. X 16. Specie Payment 
Mine, Clear Creek County, Colorado. After Bastin. 

evidence given hi this paper (PL 13) that calcite has replaced dendritic forms of 
native silver; indeed, some of the segmented veinlets are offshoots from partially 
replaced silver dendrites. 

Whether all segmented veinlets are the result of partial replacement as at Gow- 
ganda is uncertain. In the absence of evidences of replacement an alternative ex- 
planation would be the filling of fractures by two or more minerals whose grains 
crystallized from different centers until they finally interfered. 

Veinlets with AtUomorphic Crystalline Walls. An exceptional type of veinlet, well 
displayed in the cobalt-nickel-silver ores of Cobalt and South Lorraine, Ontario, is 
illustrated in Figure 3 of Plate 4 and Figures 1 and 2 of Plate 5. The walls of the 
veinlets tend to match roughly, implying that they are fracture fillings. Examination 
of the walls of the veinlet under high power reveals, however, that they are made up 
of automorphic crystals of the mineral that forms the walls of the veinlet. These re- 
lationships in ore from South Lorraine, Ontario, are shown at low magnification in 
Fig. 3 of Plate 4. The segmented veinlets of calcite (black) and silver (white) traverse 
safflorite. In the larger veinlet the tendency of opposite walls to match is conspicu- 
ous. When, however, the walls of such veinlets are more highly magnified (PL 5, 
fig. 1) it is seen that they are composed mainly of automorphic crystals of safflorite. 
The central part of Figure 1 Plate 5 is shown more highly magnified in Figure 2 of 
Plate 5. It is evident that no replacement of safflorite by the silver and calcite of the 
veinlets has taken place. 


These unusual veinlets are interpreted as ruptures produced during the primary 
mineralization and while the arsenides were still crystallizing. In these ruptures 
silver and calcite were deposited though not necessarily simultaneously. Very similar 
relations are shown in Figure 2 of Plate 4. Many of the veinlets just described are 
segmented veinlets. Still other features prove that these veinlets of the Cobalt ores 
are fracture fillings. In places the silver veinlets pass along their trends into breccias 
of angular fragments of arsenides enclosed by silver. One silver veinlet was observed 
to give way along its trend to a succession of silver-filled gash fractures oriented 
diagonally to the trend of the vein a typical rupture phenomenon. Finally, where 
a silver-calcite veinlet traverses ore showing the "tubercle" texture, the cap or outer 
layer of one of the tubercles lies on one side of the veinlet, while the main part of the 
tubercle lies on the opposite side. 

Dendrites (Swartzlow, 1934) may occupy minor fractures or may be developed by 
replacement. The delicate, branching, crystalline growths deposited along very small 
fractures are well illustrated by the manganese oxides on joint planes at Leadvilie, 
Colorado (PL 3, fig. 4). Replacement dendrites are shown in Figures 1 and 3 of 
Plate 12. 

Most fracture fillings in ores are crystalline, but some in crystalline material are 
colloidal and show collofonn textures. Figure 5 of Plate 7 shows small fissures in 
crystalline smaltite filled with colloidal native arsenic. 


Filling of pores as a mechanism of ore deposition becomes important mainly in 
rocks in which the percentage of pore space is high enough to give ample space for 
ore deposition and in which the pores are large enough to make the rock easily per- 
meable. Limonite and hematite often fill the pores of sandstones, but the volume of 
pore space is almost never sufficient to produce workable deposits. Perhaps the best 
examples of ores that are in part pore fillings are the copper, vanadium, and ura- 
nium deposits of the so-called "Red Beds" type (PL 1, fig. 6). In such sandstones one 
must be sure the material now filling the spaces is not a replacement of an older 
matrix. Remnants of any older matrix usually remain. 


The commonest examples of the filling of plant cells by metallic minerals are the 
pyrite nodules so often associated with coal beds. Examples are also found in asso- 
ciation with disseminated copper deposits in sandstones, the so-called "Red Beds" 
type of copper deposits (PL 5, fig. 3). The uncollapsed condition of the cells in many 
of these occurrences shows that the pyrite filling was deposited before the plant ma- 
terial was buried under any considerable load. Not uncommonly chaicosite has re- 
placed the pyrite. There is good evidence that in some such ores chalcocite has directly 
filled cell cavities as well as directly replaced wood (Papenfus, 1931, p. 318). 



Deposition of mineral matter in a very finely divided state may produce ore tex- 
tures very different from those that characterize coarse cyrstalline precipitates from 
"true" or electrolytic solutions such as have thus far been considered. Brief consid- 
eration of some general characteristics of fine dispersions is necessary before pro- 
ceeding to discuss their role in ore deposition. 

When a cube is divided into eight cubes of half the original diameter the surface 
exposed is thereby doubled. If the same cube is divided into cubes of one-fourth the 
original diameter the surface is quadrupled. Thus without change in mass the amount 
of exposed surface i^ notably increased, and all phenomena dependent upon the 
total amount of exposed surface of the solid are thereby greatly amplified. If the 
process of comminution is continued, ultimately a condition is reached in which the 
particles will be so small that they will not settle readily in any fluid medium in which 
they are suspended, and may remain suspended for months. Such systems are called 
colloidal dispersions or sols; the particles are the dispersed phase, and the fluid is the 
dispersing medium. This general principle applies regardless of whether the substance 
subdivided be solid or liquid and regardless of whether the surrounding medium be 
gas or liquid. Thus the importance of large surface areas of exposure may be mani- 
fest in such diverse phenomena as tiny drops of water suspended in air as fog, in 
small solid particles suspended in air as smoke, in fats suspended in water as cream, 
and in particles of solid gold suspended in water. 

Characteristics of colloidal dispersions are: (1) buoyancy, (2) brownian movement, 
(3) adsorption, (4) electrical charge, and (5) chemical reactivity. These will require 
comment in later paragraphs. 

The economic geologist is concerned almost exclusively with dispersions in liquid 
water which, however, is not pure water but carries electrolytes in solution and may 
also carry more than one substance in suspension. Further, he is concerned mainly 
with dispersions of solids in water but subordinatley with dispersions of liquids in- 
soluble in water. In further discussions these will be the type? of dispersions prin- 
cipally emphasized. Colloidal dispersions or sols have already been defined, and 
coagulates or flocculates from such dispersions will be referred to as colloidal floccu- 
lates, regardless of whether the flocculated particles be solid, liquid, or mixtures. 
The term gels is restricted to colloidal flocculates or coagulates from emulsoids which, 
like gelatine or colloidal silica, may show somewhat elastic, jelly like properties. Floc- 
culates from suspensoids are not jelly like. The textures characteristic of colloidal 
flocculates may be termed colloidal texture or, with certain limitations, colloform 
textures following the usage proposed by A. F. Rogers (1917, p. 518); Colloform 
refers only to the shape or form of the deposit regardless of whether the material is 
now collidal or exhibits only a relic colloidal or to use Wherry's (1914, p. 112) term, 
a metacolloidal texture. 

Most particles in colloidal dispersions are too small to be seen with the compound 
microscope, but with proper illumination, in the ultramicroscope, reflections of light 
from the particles can be seen, and it can be determined that they are in rapid motion. 




As the particles in most colloidal dispersions in water are much heavier than water 
we may inquire why they do not rapidly settle. If a cube of unit diameter is divided 
into cubes of a fourth that diameter each of these has one sixty fourth the mass of 
the original cube but has one sixteenth as much surface. Thus as materials become 
progressively finer-grained the ratio of surface to mass increases, and buoyancy tends 
more and more to delay settling. Thus it can be calculated (Burton, 1938, p. 4) that 
under quiescent conditions a sphere of silver 1 cm in radius in water will settle 1 cm 
in five-millionths of a second. A sphere 0.00001 cm in radius will require about half 
a day to settle the same distance, and a sphere 0.000001 cm in radius will require 
about 58 days. Thus if no factors other than buoyancy were involved colloidal parti- 
cles would tend to settle slowly. There are, however, other factors that add to the 
effects of buoyancy. 


The ultramicroscope shows that the motion of particles in colloidal suspensions 
lacks regularity, and that it changes abruptly in direction. This so-called Brownian 
movement is believed to be due to unequal bombardment of the small dispersed 
particles by the molecules of the medium. At one instant more fast-moving molecules 
may strike from one quarter, the next instant from a different quarter, causing the 
particles to move in successively different directions. In very fine particles the magni- 
tude of the random motion may be many times the motion due to gravitation and 
consequently may be much more important than buoyancy in keeping the particles 
in suspension. 

According to Burton (1938, p. 5) the feature common to all colloidal dispersions 
is the small size of the particles ; they are small enough not to settle out under gravity 
(about 10^ 5 cm in diameter) and large enough to have a very slow rate of diffusion 
(about 10~ 7 cm in diameter). 


In the interior of a liquid or a solid, molecules are attracted equally in various 
directions, and these attractions tend to neutralize each other. In a surface, however, 
molecules have their attracting forces neutralized on one side only, and surfaces 
therefore exert strong attractions on molecules as well as on atoms and ions in their 
close vicinity. This phenomenon, known as adsorption, is particularly well displayed 
in colloidal dispersions because of the large amounts of surface involved. Adsorption 
is often highly selective. 


Not only do fine particles dispersed through liquids adsorb molecules of the liquid, 
but they may adsorb ions as well, and as these ions carry electrical charges the solid 
particles also become electrically charged. Some substances adsorb positive ions more 
readily than negative ions, and vice versa. In water solutions the hydrogen and hy- 
droxide ions are principally involved. In dispersions of a single substance in a liquid 


the particles become similarly charged and tend to repel each other. This is an im- 
portant factor in keeping colloidal particles in suspension because flocculation takes 
place readily when the charges are neutralized. 

An important consequence of the electric charges on the particles of colloidal dis- 
persions is that when such dispersions are brought in contact with other colloidal 
dispersions whose particles are oppositely charged neutralization of both charges 
may occur and flocculation ensue. Similarly when a colloidal dispersion is brought in 
contact with a true or electrolytic solution carrying an abundance of oppositely 
charged ions neutralization with resultant flocculation may also occur. The peculiar 
effectiveness of aluminum sulphate in water clarification is due to the fact that it 
yields both negative and positive ions and is therefore effective hi flocculating both 
positively and negatively charged colloidal particles. 


The large amount of surface that characterizes colloidal dispersions of solids in 
either gases or liquids is in general favorable to chemical interaction between such 
solids and the enclosing fluids or with substances dissolved in such fluids. The prin- 
ciple is particularly well illustrated in oxidation phenomena as for example the ex- 
plosibility of magnesium dust in air as contrasted with the stability of a bar of mag- 

Thus colloidal dispersions have many unique characteristics, and these should be 
reflected in characteristic textures. 

In the naturally occurring colloidal dispersions in water the water contains dis- 
solved mineral matter. Deposition from such dispersions takes place by flocculation 
of the dispersed particles and by crystallization of the dissolved material. There seems 
to be no a priori reason why the two processes should not proceed concurrently, and 
indeed the ore textures indicate that such has often been the case. Furthermore, if 
chemical reactions in electrolytic solutions are the starting point in the formation of 
most natural colloidal dispersions, associations of this sort should be common. 


In the laboratory the two general methods of getting substances into the colloidal 
state are: (1) Larger particles are disintegrated until colloidal proportions are at- 
tained. This can be done in an electric ate or by fine grinding in the presence of some 
means of preventing flocculation of the particles. The formation of fine clays which 
have many of the properties of colloids in perhaps the most familiar geologic analogue 
of this method, but there seem to be few examples in the processes of ore deposition. 
(2) Substances in the molecular condition, usually in "true" or electrolytic solution, 
are built up to colloidal particles; precautions must be taken to prevent the particles 
from growing too large. This method involves a reduction in surface with conse- 
quent liberation of energy. It might, therefore, be expected to occur spontaneously 
and rather commonly in nature; and actually most natural colloids have formed in 
this way. Any type of chemical reaction whether of oxidation, reduction, double de- 
composition, or hydrolysis may initiate the building up of colloidal dispersions. 



Thus far features common to all colloidal dispersions have been stressed. Some 
significant contrasts, partly qualitative exist, however, between (1) dispersions of 
solids in water, and (2) dispersions of liquids in water or dispersions of certain sub- 
stances which though solid have many of the properties of liquids, as, for example, 
agar agar and gelatine. Recognition of these classes is of great importance to the 
chemist and the economic geologist. 

Dispersions of solids in a liquid are commonly termed suspensoids; in nature the 
liquid is usually water. Such dispersions are nonviscous, nongelatinizing, tend to be 
non-stable and easily flocculated, and once flocculated are not readily redispersed. 
To this class belong nearly all colloidal dispersions of the metals and their compounds. 

Dispersions of liquids in other liquids in which they are insoluble and dispersions 
of substances such as gelatine in liquids are usually termed emulsoids. They are often 
viscous and tend to gelatinize. They are likely to be more stable than suspensoids 
for they are not so readily flocculated by electrolytes. Gelatine and like substances 
tend to disperse readily and spontaneously in water. Redispersion after flocculation 
can be easily accomplished. Emulsoids of oil in salt water are of mucli importance to 
the petroleum geologist. Colloidal dispersions of SiO 2 in water can be prepared in 
the laboratory from soluble silicates, and under certain conditions such silica dis- 
persions coagulate to form stiff gels. In ore deposits silica now gelatinous is very rare, 
as we might anticipate, since silica gels are unstable and most ore deposits were 
formed long ago. Nevertheless some occurrences are known as in the upper portions 
of the Great Australia mine in Queensland, Australia (Levings, 1912, p. 478). That 
much of what is now quartz in certain types of ore deposits was deposited originally 
from colloidal dispersion may be inferred, however, from relic colloidal textures still 
visible in the quartz (PL 5, fig. 5). 


Despite the dispersive effects of Brownian movement and of like electrical charges 
on the particles, colloidal dispersions are unstable systems. The dispersed particles 
tend to flocculate easily, and the flocculated material tends to settle and separate 
from the fluid in which it was suspended. After flocculation redispersion is very diffi- 
cult and under natural conditions usually impossible. However, certain substances 
facilitate the formation of colloidal dispersions and tend to perpetuate or "stabilize" 
or "protect" them once they are formed. 

The presence of certain emulsoid colloids, such as gelatine, facilitates the formation 
of suspensoids, renders them more stable, and often make it possible to obtain more 
concentrated suspensoids and to redisperse a flocculate that otherwise could not be 
dispersed. Seemingly, the emulsoid is adsorbed on the surface of the solid particles 
of the suspensoid and this layer inhibits flocculation both by preventing the electrical 
discharge of the suspensoid particles and by offering a buffer to their coalescence. 
Thus the suspensoid acquires some of the properties of an emulsoid. 

In the laboratory certain common ore minerals are effective precipitants of gold 
and silver from solutions of gold chloride and silver sulphate (Palmer and Bastin, 


1913). Furthermore, with addition of a gelatine emulsoid to the solutions, the gold and 
silver can be obtained in beautiful colloidal suspensions instead of as crystalline pre- 
cipitates. Whether analogous processes operate in the natural transport and precipi- 
tation of gold and silver is not known. The emulsoid most likely to function in nature 
as did the gelatine in these experiments (Bastin, 1915) is a silica emulsoid. In further 
experiments silica was substituted for gelatine, and colloidal solutions of gold or 
silver were obtained with many of the ore minerals. 

Ordinarilly, flocculated suspensoids cannot be redispersed, but in the presence of 
"peptizers" some flocculates can be redispersed. Among the substances that might 
function as a peptizer in nature is hydrogen sulphide (Clark and Menaul, 1916) 
which is known in the laboratory to facilitate the formation of colloidal suspensions 
of certain sulphides and sulpho-salts in alkaline solutions. 



While it is now generally accepted that certain ores were deposited as colloids 
the proof of such an origin in many cases offers inherent difficulties. Examples of 
present-day deposition of such ores are not numerous and have not been adequately 
studied. Most of the ores thought to be of colloidal origin were formed long ago and 
have crystallized and developed new textures. In the siliceous ore from the Talisman 
Mine, New Zealand, the microscope reveals a granular texture in polarized light (PL 
5, fig. 6), but the same field in ordinary transmitted light (PL 5, fig. 5) shows an older 
and very different and much finer texture characterized by spheroidal forms outlined 
by bands of dark, opaque, accessory minerals. The obvious extreme fineness of this 
earlier texture and the presence of outlines that are almost perfect circles or arcs 
thereof suggest that the silica and probably the accessory minerals were originally 
deposited as very finely divided material of such low viscosity that surface tension 
could operate to produce spheroidal surfaces. Such deposits are commonly interpreted 
as colloidal. In general the finer the original grain, the more nearly circular the out- 
lines, and the smaller the scale on which the phenomena are displayed the stronger is 
the probability of colloidal origin. Typical deposits (PL 5, fig. 5) show occasional 
vugs bordered by spheroidal surfaces and other evidences of deposition in open spaces. 

To clarify the surface-tension explanation of such textures it is appropriate to re- 
call some general characteristics of surface-tension phenomena. The spheroidal out- 
lines assumed by one liqid when suspended in anbther of equal density, the spheroidal 
forms of small rain drops, and the incompletely spherical form of the bubbles in var- 
ious kinds of foams are all "surf ace- tension " phenomena. The molecules hi the center 
of a body of liquid are attracted equally in all directions by the surrounding molecules, 
but towards any free surface of the liquid the outward attraction in a direction nor- 
mal to that surface decreases becoming zero when the surface is reached. Thus mole- 
cules at any point on the surface are being pulled parallel to the surface and toward 
the interior but not outward. The effect is somewhat as if the mass of liquid were en- 
closed in a tightly stretched flexible membrane. The molecules of liquid tend to draw 
together as much as possible until the whole mass has the smallest possible surface 
compatible with its volume. Unless distorted by forces outside the liquid, this sur- 
face is spherical,^ 

Not only may wholly liquid substances assume rounded outlines because of sur- 
face tension, but liquids carrying much finely divided solid or semisolid matter in 
suspension (colloidal suspensions) may also assume spheroidal outlines under the 
influence of surface tension. Good examples of rotund surfaces are found among non- 
mineral substances in such familiar colloidal coagulates as glue and cottage cheese; 
the latter contains usually about 70 per cent of water. 

Not only have spheroidal deposits of silica like that shown in figure 5 of Plate 5 
been interpreted as colloidal, but many deposits of oxides, sulphides, and a few ele- 
ments that exhibit globular, botryoidal, reniform, or mammillary forms even on a 
relatively large scale have also been so interpreted. When the materials in their pres- 



ent state are fine-grained, when the spheroidal forms are developed on a small scale 
and approach sphericity, and when the material is notably hydrous, this interpreta- 
tion may be valid; for others, alternative explanations must be considered before 
final diagnosis. 

Several textural features of colloidal flocculates depend directly on their high water 
content which in turn is due to the large amount of surface available for water ad- 
sorption. The water content is a maximum in gels formed by the coagulation of 
emulsoids. A fairly stiff agar agar gel may be 99 per cent water, and gelatine gels 
that are 85 to 90 per cent water may be stiff enough so that their modulus of elastic- 
ity can be measured. In the laboratory silica gels can be prepared stiff enough to be 
cut yet contain over 90 per cent water. Natural silica gels occur at some hot springs, 
and a few have been recorded in mines, but no determinations of their water content 
seem to have been made. Natural opal is probably the solidified product of such gels, 
but during solidification the water content becomes greatly reduced. Ninety-four 
analyses of opal quoted by Hintze (1915, p. 1535-1538) show water contents mainly 
between 5 and 16 per cent. 

Little information is available concerning the water content of artificial flocculates 
from suspensoids the class that most concerns the mining geologist. Noncrystallized 
varieties of hydrous iron oxide listed by Hintze (1915, p. 2073) show for 77 analyses 
water contents ranging from 5 to 26 per cent; these were presumably more hydrous 
when first deposited. 

In the earlier studies of colloidal ore deposition emphasis was laid upon the rounded, 
more or less spherical forms and in 1917 Rogers proposed the term "colloform" for 
such forms when developed in open spaces (p. 518). 

"From the standpoint of physical chemistry amorphous solids are liquids. Now, the shape of a 
liquid unaffected by gravity or other external influence is spherical, and so we often find the hydrogel 
minerals in spherical, botryoidal, reniform, stalactitic, and mammillary forms. These forms inter- 
grade, so that one is often at a loss which term to use. I therefore propose the term colloform for the 
rounded, more or less spherical, forms assumed by colloidal and metacolloidal substances in open 
spaces. Some crystalline, not merely microcrystalline, minerals such as smithsonite, also occur in 
colloform crusts, and it should be emphasized that this term refers only to the shape or form, and 
not to the condition of the material " 

Lindgren (1925) pointed out that colloidal minerals can also be deposited by re- 
placement and cited marcasite "with colloform structure" that had replaced gangue 
minerals as well as pyrite and sphalerite (PL 5, fig. 4). He also cited as perhaps the 
most familiar example of colloidal replacement the development of rounded chert 
nodules in limestones and dolomites especially those nodules in which oolitic tex- 
tures of the limestone were preserved in the chert (PL 8, fig. 6). 

Other textural features besides "rounded, more or less spherical forms" suggest 
though they do not always demonstrate colloidal deposition. By definition Rogers' 
term "colloform" was restricted to spheroidal forms developed in open spaces, yet 
a comprehensive term is needed to include all textures that point to the original 
deposition of the material in the colloidal state. The writer would be willing to see 
Rogers' term "colloform" broadened to serve this purpose, but in the present book 
"colloidal textures" is used as the comprehensive term. 

The following textural features at least suggest colloidal deposition: (1) rotund or 


spheroidal forms (in part), (2) syneresis or shrinkage cracks (in part), (3) pellet tex- 
ture (in part), (4) framboidal texture, (5) interference surfaces (in part). 


By no means all of the rounded, more or less spheroidal forms found in ores indicate 
colloidal deposition. In the stalactites in certain zinc mines cross sections show 
rounded outer surfaces, and the successive bands of sphalerite and pyrite or marcasite 
of which they are made up also show rounded outlines. In many such stalactites the 
minerals are now crystalline, and there is no evidence that they were colloidal when 
deposited. In certain veins rounded outlines are developed by the successive deposi- 
tion of bands of different minerals around rock fragments as illustrated in the cockade 
or ring ore shown in Figures 5 and 6 of Plate 3. All the minerals of this ore seem to 
have been deposited in the crystalline rather than the amorphous state. In still other 
ores rounded outlines may be developed by the diffusion of the mineralizing solution 
outward from certain centers often combined with rhythmic precipitation (PL 14, 
fig. 4; PL 15, fig. I). In some such banded ores colloidal deposition is probable, but 
in others the ore minerals were crystalline as deposited. 

Rotund outlines combined with quite delicate banding (PL 4, fig. 1) are sometimes 
developed by layered deposition or crustification on the walls of openings or around 
angular fragments of wall rock or older ore minerals a delicate cockade texture. 
In Fig. 1 of Plate 4 galena forms crystals up to 5 mm across, and the associated sphal- 
erite though much finer is far above colloidal dimensions. 

There are, however, many other ores, mostly finer-grained, in which spheroidal 
outlines are developed in such delicacy, smoothness of outline, and close approach 
in cross section to a circle or an arc thereof as to strongly suggest that surface 
tension caused the spheroidal forms. Such textures are well displayed on a micro- 
scopic scale in the gold-bearing quartz from the Talisman mine, New Zealand. Figure 
5 of Plate 5 indicates that nearly complete reorganization of the silica to a micro- 
crystalline aggregate has not obliterated the original spherulitic forms. 

Original colloidal textures are most easily demonstrated in transparent ores, that 
can be studied under the microscope in transmitted light. In opaque ores that can 
be studied only in polished specimens by reflected light evidence of colloidal deposi- 
tion is likely to be presumptive rather than conclusive. In the covellite-chalcocite 
ore from Kennecott, Alaska (PL 6, fig. 1), both minerals are now crystalline. How- 
ever, spheroidal outlines suggest surface-tension phenomena, and the chalcocite 
seems to have been guided in its replacement of the covellite by fractures that were 
(1) radial, (2) concentric, or (3) netlike and hence strongly suggestive of the shrink- 
age cracks so characteristic of colloidal materials. Lasky believes the covellite was 
originally colloidal and became fractured with shrinkage. Subsequently it crystallized 
to its present radiating pattern and was partially replaced by chalcocite. 

Colloids deposited by replacement may exhibit rotund forms somewhat similar 
to those just described but apparently due to diffusion and rhythmic precipitation 
and not to surface tension. Figure 4 of Plate 5 shows partly oxidized ore from the 
Bigbug District, Arizona, in which fine-grained, curved concentric bands of marca- 



site are enclosed in a relatively coarse granular aggregate of sphalerite, pyrite, and 
gangue minerals. In places the banding of the marcasite transects the grain bound- 
aries of the matrix. According to Lindgren (1926, p. 140-141) this represents a re- 
placement of sphalerite and gangue by colloidal iron sulphide which later crystal- 
lized to marcasite. 

FIGURE 11. Colloform pitchblende 

Association of crystalline pyrite and chalcopyrite with pitchblende that was deposited as a colloid and shows spheroi- 
dal outlines. The three minerals are believed to be of nearly the same age. Wood mine, Central City, Colorado. X 17. 
After Bastin. 

Lindgren described excellent examples of replacement of limestone by highly 
siliceous ores from the Gemini mine in the Tintic District, Utah (PL 15, fig. 2). These 
are discussed at more length in the Chapter on Replacement. He believed (1915, p. 
231) limestone was first replaced by colloidal silica; metal-bearing solutions diffus- 
ing into this silica gel then deposited galena and sphalerite rhythmically to form the 
darker bands. In the lighter bands there are chalcedony spherules. The banding is 
due to diffusion and rhythmic precipitation. 

As might be expected, colloidal substances are weak materials and easily deformed 
by external stresses. Figure 3 of Plate 6 shows an ore from Silverton, Colorado, 
which is a breccia of pyrite fragments in a matrix of quartz with accessory sphalerite 
and galena. The pyrite fragments show the rotund outlines and the radial and con- 
centric shrinkage cracks characteristic of colloids. 

All or only some of the minerals of an ore may have been deposited as colloids. In 
the rather unusual ore shown in Figure 5 of Plate 7 older crystalline smaltite has 
been brecciated and in the fractures was deposited native arsenic having spheroidal 
outlines suggestive of colloidal rigin. Even among essentially contemporaneous ore 
minerals some may be deposL .^ as crystalloids and others as colloids. In the ore 
shown in Figure 11 the three-component minerals are nearly contemporaneous, yet 
the chalcopyrite and pyrite were crystalline whereas the pitchblende was deposited as 


a colloid as shown in the polished specimen both by spheroidal outlines and shrink- 
age cracks. The explanation of such associations is inherent in the nature of the min- 
erals. Just as some minerals such as barite and pyrite have a strong tendency to 
crystallize so the oxide of uranium is, in ore deposites, almost invariably deposited as 
a colloid. 


Fracturing, usually on a small scale and often microscopic, is common in ores 
whose other characters suggest colloidal deposition. Such fracturing is often either 
parallel to or approximately perpendicular to spheroidal outlines. When, colloidal and 
crystalline minerals are associated in the same ore, this type of fracturing is restricted 
to the colloidal minerals (PL 6, fig. 5). Such relations indicate that the fractures are 
not due to external stresses but are of internal origin. Cracks similar to those under 
discussion can be developed in the shrinkage of artificial gels and have been termed 
Syneresis cracks. In colloidal ores as in the artificial gels the cracks seem to have been 
formed by loss of water. 

In some ores the shrinkage cracks form a fine irregular pattern somewhat resembling 
that of crackled porcelain (PL 6, fig. 2). This ore is composed of very fine-grained 
chalcocite with a steely luster and a conchoidal fracture. Etching the polished sur- 
face with nitric acid brings out the fracture pattern. 

Shrinkage cracks developed in colloidal pitchblende from Joachimsthal, Bohemia, 
are shown in Figure 4 of Plate 6. Some of these fractures trend either parallel to or 
perpendicular to the curved boundaries of the pitchblende but others show diverse 
trends. The cracks in the pitchblende do not penetrate the surrounding mineral. 
The brecciated colloidal iron sulphide of Figure 3 of Plate 6 shows both radial shrink- 
age cracks and curving cracks that parallel spheroidal surfaces in fact most of the 
fragments are bounded by such fractures. 

Shrinkage cracks may traverse an association of two or three colloidal minerals. 
In the ore from the Cornwall mine, Missouri (PL 6, fig. 5), shrinkage cracks traverse 
globular intergrowths of chalcocite and covellite but do not extend into the surround- 
ing chalcopyrite. Rust (1935) interprets all these minerals as primary (hypogene). 

A fairly common type of shrinkage crack in colloidal ores is the bifurcating form 
shown in Figure 6 of Plate 6 also from the Cornwall mine, Missouri. The crack tra- 
verses an aggregate of tiny chalcopyrite spherules (often with pyrite cores) in a born- 
ite matrix. The sulphides have replaced a nodular area of dolomite in sandstone. 
Rust believes that the sulphides were deposited in colloidal form and that the cracks 
developed by shrinkage on drying. 

In cross sections of banded presumably colloidal ores, tapering cracks of the sort 
shown in Figure 6 of Plate 7 are sometimes present. This figure shows banded chalco- 
pyrite ore deposited on the wall of a fracture in limestone in the Cornwall mine, Mis- 
souri. These cracks are undoubtedly due to shrinkage, and those in the lowest (and 
oldest) band formed before the overlying band was deposited. The tapering forms 
show that shrinkage was greatest at the former surfaces of the successive bands. 

Microscopic shrinkage or syneresis cracks in very fine-grained materials suggest 


deposition of the material in a colloidal state. Such cracks are not, however, con- 
fined to colloids, for familiar mud cracks are of this class. 


In certain ores, most if not all of which have been developed by replacement, some 
of the metallic minerals when viewed under the microscope are tiny spheroids which 
may be homogeneous or may show concentric or radial structures. For such forms 
the term pellet texture is proposed. The pellets resemble oolites in form but are on a 
much smaller usually a microscopic scale and have been formed by quite differ- 
ent processes. 

Not all pellets are of colloidal origin. They are a common feature for example of 
the ores of Mount Isa, in Queensland (Grondijs and Schouten, 1937) which are re- 
placements of very line sediments. There the spheroidal pellets usually have as their 
core a crystal of pyrite, and there are gradations from spheres to automorphic crystals 
ofpyrite (PL 11, fig. 1). 

In other ores the pellets appear to have been deposited in a colloidal state. In ore 
from the Cornwall mine, Missouri (PL 6, fig. 5), a colloidal origin is favored by the 
shrinkage cracks so well developed within the pellets. Kidd and Haycock (1935) 
have described in the ores of the Eldorado mine, Great Bear Lake, very small pellets 
of pitchblende in quartz. These are clearly colloidal. They are often found between 
much larger spherules of pitchblende. 


The first description of this texture in ores from the United States appears to be 
of ore from the Cornwall mine, Missouri (Rust, 1935). They are shown in Figures 
1 and 3 of Plate 7. To this unusual texture Rust applied the term <% Framboidal" 
(after Framboise, the French for Raspberry) because of its resemblance to the ag- 
gregate fruits of the raspberry. The texture was recognized earlier by Schneiderhohn 
(1923) in the copper shales of Mansfield, Germany, but because of the extremely 
fine grain of these ores was not very clearly figured nor was a distinctive name given 
to it. A magnification of X800 was necessary to make these Mansfield structures 
visible. They consist of globular (rarely rodlike) masses 4 to 15 /* in diameter mainly 
of chalcopyrite, though bornite or chalcocite may substitute for chalcopyrite. Each 
globular mass is made up of many tiny chalcopyrite grains, and a few of the grains 
exhibit triangular outlines. Their maximum diameter is about 0.2 /*. 

Schneiderhohn believed these ores were deposited in colloidal form as a gel of mixed 
sulphides which later became very finely crystalline. He suggested ore deposition on 
the sea bottom and interpreted the small chalcopyrite spherules as representing 
colonies of sulphur-depositing bacteria and the tiny grains of which they were com- 
posed as "fossil bacteria." Not only the form of the spherules but the order of size 
he regarded as supporting the bacterial hypothesis. 

The "framboidal" spherules described by Rust from the Cornwall mine, range in 
diameter from 18 to 48 n or about 3 to 12 times as large as those in the Mansfield 
deposits. Most of them are pyrite rather than chalcopyrite. 


Schouten (1946) has published excellent illustrations of framboidal textures in the 
Copper shales of Mansfield and has figured similar textures from the ores of Meggen, 
Germany, from Redruth in Cornwall, England, and from Rio Tin to, Spain. A typical 
occurrence at Meggen is shown in Figure 2 of Plate 7. Schouten discusses Schneider- 
hohn's theory of a bacterial origin of this texture and rejects it in favor of an inorganic 

The recognition of framboidal textures in undoubted hypogene ore deposits such 
as those of the Cornwall mine, Missouri, Rio Tinto, Spain, and Redruth in Cornwall, 
England, rules out a bacterial origin. According to Rust (1935) the framboidal tex- 
tures as displayed in the Missouri occurrences 

"are probably metacolloidal in origin, having been formed by the crystallization of a globule of 
pyrite gel, crystallization starting simultaneously at many points. In an alternative hypothesis, the 
texture might be considered to have been formed by the bunching of tiny pyrite grains as they 
floated in a chalcopyrite gel. On the basis of uniformity of grain size, gradational types between the 
grain clusters and solid spheroids, and the fact that in non-circular, irregular pyrite masses, the 
'framboidal' texture may be produced artificially by etching with nitric acid, the writer prefers the 
former interpretation." (p. 408). 

Some of the pellets illustrated in Figure 5 of Plate 6 show incipient framboidal tex- 
ture, and the shrinkage cracks indicate a probable colloidal origin. 

The distribution of the small pyrite crystals shown in Figure 2 of Plate 7 is not 
suggestive of a floating together of the pyrite grains; it seems more likely that the 
framboidal masses are composite concretions, the tiny pyrite crystals having de- 
veloped at their present locations. Notable variations in the size of the pyrite crystals 
in closely adjacent framboidal clusters also favor this interpretation. (PL 7, fig. 3). 

Known framboidal textures are confined to fine-grained ores many if not all of 
which are of hypogene origin. At the Cornwall mine they are in ores that, on the 
basis of other evidences, were almost certainly deposited in colloidal form. Whether 
all framboidal occurrences were colloidal when deposited is uncertain, and such 
uncertainty obviously limits the value of this texture as a criterion of colloidal origin. 

In the Cornwall ores the framboidal clusters are of about the same order of size 
as the pellets already described. In some cases the cluster has a matrix of primary 
chalcopyrite; in others the matrix is secondary. Within most clusters there is fair 
uniformity in the size of the units, but in some clusters there is large size diversity. 
The average size of the units may differ considerably even in adjacent clusters. 
Pellets and framboidal aggregates are often associated in the Cornwall ores, and there 
are suggestions of transitions between types. 


In some minerals, judged on other grounds to have been deposited as colloids, as 
for example the pitchblende of Figure 4 of Plate 7, the contacts between different 
portions of the colloidal mineral are straight and in marked contrast to its curving 
outlines against other minerals. These boundaries resemble the nearly straight con- 
tacts of biscuits in a pan as contrasted with their rotund free surfaces. These phe- 
nomena in the pitchblende suggest interference in a yielding medium growing from 
many centers. The interference surfaces, if such they be, belong to the period of 
growth of the colloid and are slightly older than the shrinkage cracks which formed 


after growth had ceased. In Figure 4 of Plate 7 many shrinkage cracks are restricted 
to the individual pitchblende units and do not cross the interference surfaces that 
separate these units. Shrinkage cracks may follow along the interference surfaces as 
they may also follow spheroidal surfaces. Interference surfaces have usually not been 
distinguished from shrinkage phenomena but seem to be quite distinct, as suggested 
by a comparison of Figure 4 of Plate 7 with Figure 4 of Plate 6, both showing pitch- 

Interference surfaces are not confined to colloidal deposits but have been observed 
between the finely crystalline spherulites of certain lavas where neighboring spheru- 
lites interfere in their growth (Harker, 1909, p. 276-280). Such spherulites may show 
radial and/or concentric growth patterns. 


References of particular importance on colloids and colloidal textures are listed in 
the bibliography at the end of this paper as follows: Boydell (1924-1925), Burton 
(1938), Lindgren (1924), Rogers (1917), Rust (1935), Zsigmondy (1917). 



Replacement is the dissolving of one mineral or group of minerals and the im- 
mediate deposition of another mineral or group of minerals in the place thus vacated. 
There is no intervening formation of any open spaces of a magnitude visible with the 
compound microscope. Theoretically minute spaces must exist briefly. This phenom- 
enon is one of the most widespread and varied with which the geologist deals. Although 
its practical importance seems to be limited mainly to the geologic field, probably 
because of the slowness with which replacement proceeds, it holds much of theoretical 
interest to the physicist and chemist whose co-operation is needed for a full under- 
standing of the process. Replacement is particularly well displayed in the modifica- 
tion of limestones and dolomites through the silicification of their contained fossils, 
in the development in them of chert nodules, and in the formation within them of 
large bodies of ore composed mainly of sulphides of the metals. In the process of 
downward enrichment of ores, which is of major economic importance, replacement 
is the dominant mechanism. The substance replaced is referred to as the host, and the 
replacing substance as the guest. 41 

The clear relationship of most replacement phenomena to fractures, mineral cleav- 
age, contacts between mineral grains, etc., shows that solutions were the dissolving 
and depositing agency. These solutions were sufficiently mobile to be guided by open- 
ings of microscopic dimensions so water was probably their principal component. 
Many replacements have taken place at ordinary temperatures, particularly those 
accomplished by waters of surface origin. Other replacements have been accomplished 
by thermal solutions at temperatures ranging up to 500C. and possibly higher. The 
solutions responsible for the higher-temperature replacements were presumably gas- 

Essentially every mineral or rock species has under some circumstances under- 
gone replacement. Seemingly the only general rule is that rocks such as the carbonate 
rocks limestones and dolomites that are readily soluble under a great variety of 
natural conditions have been subject to replacement most commonly and on the 
largest scale. Minerals like quartz and rocks like quartzites, which under most natural 
conditions are difficulty soluble, may nevertheless yield to replacement under special 

In most cases the replacing (or guest) mineral is now crystalline and was probably 
crystalline when it was deposited. In other cases the guest mineral, though now 
crystalline, is believed to have been colloidal when deposited. Evidences for, this 
belief are mainly spheroidal outlines assumed by the guest mineral and the presence 
of concentric and radiating shrinkage cracks filled with later minerals. These features 
suggest crystallization of a gel and have already been described. 

In most situations where replacement has been active the replacing solutions have 
circulated along channels of submicroscopic size. These usually connect, however, 

* A synonymous term is metasome. 



with larger openings nearer the source of the solutions. In the submicroscopic open- 
ings the molecular attractions between the minerals and the solution must be im- 

In almost all cases there has been no detectable volume change in replacement. 
This of course would be expected if in a rigid mineral or rock there were no visible 
open spaces to start with and space for the guest minerals became available only by 
dissolving of the host. In many instances no portion of the replaced substance enters 
into the replacing substance, as in the replacements of limestone by quartz, galena, 
etc. In other instances chemical interchange is involved as when limestone is replaced 
by calcium silicates such as wollastonite or when, in the enrichment of a copper ore, 
chalcopyrite is replaced by chalcocite. 

In replacement we have a mechanism whereby large bodies of ore may be deposited 
without the previous existence of any considerable cavities. The material dissolved 
during replacement is of course eventually deposited but often far from its source. 
Other features of replacement can best be considered with the aid of specific illustra- 


Pseudomorphism is the presence, in a mineral or mineral aggregate, of forms, tex- 
tures, or structures not characteristic of itself but characteristic instead of other min- 
erals, or mineral aggregates, or even of organisms. The recognition of pseudomorphism 
presupposes a knowledge of the forms, textures, and structures of minerals, rocks, 
and organisms and of the conditions under which these normally develop. Pseudo- 
morphism results from the dissolving of one mineral or group of minerals and the de- 
positing of another mineral or group of minerals in its place. Usually the processes 
of solution and deposition go on concurrently through the medium of the same solu- 
tion and without the development of open spaces that is, by the process known as 
replacement or metasomatism. The most delicate and perfect pseudomorphs are de- 
veloped in this way. In other instances, however, minerals are dissolved by one solu- 
tion, and after an interval, during which open spaces exist, new minerals are deposited 
in these spaces by solutions of a different composition. This is not replacement. 
In all but exceptional instances it preserves only the outward form not the internal 
structure. The value of pseudomorphism as a criterion of replacement is greatly en- 
hanced by the finding of unreplaced remnants of the host. 

Features that are on record as preserved pseudomorphically in ores include: 

In Minerals 

Crystal outlines 
Twinning planes 

In Igneous rocks 
Flow lines 

Form, size and pattern of grains 
Porphyritic texture 


In Sedimentary rocks 
Stylolitic structure 
Form, size and pattern of grains 

Organic structures 

In Metamorphic rocks 

Interesting examples of the preservation of crystal outlines after replacement are 
found in the Black Hills of South Dakota where, in the Tertiary mineralization, dolo- 
mites have been extensively silicified. Under the microscope in polarized light the 
rock is seen to be an aggregate of interlocking quartz grains, but in ordinary light it 
has the appearance shown in Figure 1 of Plate 8; the crystal outlines of the original 
dolomite rhombs are clearly recognizable in spite of the complete silicification. 

An example of preservation of crystal faces of calcite after replacement by the 
silver haloid iodobromite is shown in Figure 2 of Plate 8. 

The preservation of mineral cleavage after replacement though not common is 
occasionally noted. Figure 12 is a camera-lucida drawing of ore from Neihart, Mon- 
tana. This shows a polished surface of galena with the familiar triangular pits formed 
during grinding and polishing by the tearing out of small inverted pyramids of galena 
between the cleavage planes. The sides of these pits define the three directions of 
galena cleavage. Parallel to these cleavage directions the galena has been replaced by 
polybasite, and the position of the galena cleavage plane from which the replacement 
started is still recognizable under the microscope as a faint line running from end to 
end of the polybasite lenses. Similar relations are shown in another specimen from 
the same locality (Fig. 14). 

Textures characteristic of igneous rocks are sometimes perserved in ores developed 
by replacement. Figure 13 shows a rich silver ore from Leadville, Colorado, in which 
the sulphosalt of silver, stephariite, has extensively replaced the matrix of a quartz 
porphyry while leaving the quartz phenocrysts intact and only partially replacing 
the sanidine. 

Textures characteristic of sedimentary rocks are rather common in replacement 
ores. An excellent illustration of preservation of bedding is shown in Figure 3 of 
Plate 8, taken in the Union mine in the Black Hills, South Dakota. The entire face 
shown is ore, mainly quartz with minor amounts of pyrite, fluorite, and gold. The 
bedding of the original shaly limestone is still beautifully preserved in the ore. 

Preservation of cross-bedding after replacement, though occasionally observed, 
is much less common than preservation of bedding, probably because the kinds of 
rocks in which cross-bedding is best developed namely sandstones are usually not 
easily replaced. 

Stylolites results from uneven solution under mild pressure in sedimentary rocks, 
and except in rare instances they are formed only in calcareous sediments. Stylolites 
found in ores, therefore suggest that the ores are replacements of limestones or dolo- 
mites. Stylolites are common in the "blanket" deposits of fluorite in Southern Illinois 



that are known to be replacements of limestone. The field shown in ordinary light in 
Figure 5 of Plate 8 is essentially all fluorite. 

As indicated in a later chapter oolitic textures are common in many iron ores; 
the oolites usually are hematite but sometimes chalmersite. These ores were formerly 
considered replacements of oolitic limestones, but it is now known that in most of 




FIGURE 12. Replacement of galena along cleavages 

From Neihart, Montana; replacement along cleavage planes by polybasite. The position of galena cleavage is still 
^cognizable in the polybasite as a faint line. After Bastin. 

FIGURE 13. Replacement of an igneous rock by stephanite (St} 
In silver ore from Leadville, Colorado, porphyritic texture still conspicuous. S Sanidine, Q Quartz. X 36. After 

these ores the oolites are original sea-bottom deposits. Preservation of oolitic textures 
by the replacement of limestones by silica is, however, common. In the Baraboo 
District, Wisconsin, chert nodules have been developed in an oolitic limestone; within 
the nodules are oolites identical with those in the bordering limestone except that 
they are silica rather than calcite (PL 8, fig. 6). Similar evidence would be necessary 
to establish the replacement origin of oolitic hematite or chalmersite ores. The forms 
and to some extent the concentric structure of oolites are preserved in quartz and 
pyrite in certain gold-silver ores of the Black Hills, South Dakota (PL 9, fig. 1). It 
is known on other grounds that these ores are replacements of dolomitic limestones. 
The preservation of oolitic textures is corroborative evidence. 

The forms and structures of organisms are not uncommon in ores that have re- 
placed sediments. Replacements of wood by sulphides are particularly well shown in 
sulphide nodules in disseminated copper deposits in sandstones. In such occurrences 
it is important to distinguish between the filling of the cell ulterior and replacement of 
the cell walls. Figure 3 of Plate 5 shows ore from the Sacramento District, New 
Mexico, in which pyrite (white) has filled the cells, and chalcocite (light gray) has re- 


placed the cell walls and filled small cracks. It has been inferred that the cells in the 
upper left of this figure were rotted and crushed at the time of mineralization, while 
the bulk of the specimen was not crushed. This suggests mineralization prior to deep 

One of the most striking instances of replacement of animal remains is afforded by 
the pyritic fossils of the Tully formation in New York. In the normal phases of this 
formation the fossils are calcareous, but over restricted areas the same species are 
pyritic. As no organisms are known to secrete pyritic shells pyrite must have replaced 
originally calcareous shells. The pyritic fossils average about one fifteenth the size 
of the same species in the normal Tully limestone. This dwarfing suggests that local 
high iron and sulphur content on the sea bottom in Tully times stunted the growth 
of the shells and also led shortly thereafter to their replacement by pyrite in the muds 
of the Devonian sea bottom. Obviously the interpretation of these pyritic fossils 
as due to replacement is based not on textural evidence but upon a knowledge of the 
composition of shells and the geologic conditions under which they are deposited. 

Spurr observed an extraordinary instance of replacement of a gastropod shell by 
silver at Aspen, Colorado. 

Edwards (1947, p. 28, Figs. 29, 30) has described and figured banded lead-zinc 
ores formed by replacement of schistose porphyry from Captain's Flat, New South 
Wales, and also by replacement of contorted slates from Mount Isa, Queensland. 
Microscopic study is often necessary to determine whether the mineralization pre- 
ceded or followed the metamorphism. If it followed the metamorphism the metallic 
and other minerals introduced by replacement should show no deformation. 


In the pseudomorphic replacements just described the forms, structures, or textures 
of the host are preserved and constitute evidence of replacement. In other instances 
no traces can be found in the guest mineral of the forms or textures of the host. If, 
however, the guest minerals transect features characteristic of the host, replacement 
can still be demonstrated. Figure 4 of Plate 8 shows a schist in which three pyrite 
cubes are developed. If they had antedated the development of the schistosity they 
should show fracturing and other evidence of deformation, and the schistosity of the 
rock should wrap around them. Instead, the pyrite crystals, unfractured, transect 
the schistosity. It is clear that schist once occupied the spaces now occupied by 
pyrite. The pyrite must therefore have developed by replacement after the develop- 
ment of the schistosity. Similar relations in gold ore from Timmins, Ontario, are 
shown in Figure 2 of Plate 9. 

Structures transected by replacing minerals include typical grain patterns of igne- 
ous rocks, bedding, flow structure, schistosity, and organic structures. 

In most cases of transecting textures the guest mineral develops its own character- 
istic crystal form as in the pyrite cubes of Figure 4 of Plate 8 and Figure 2 of Plate 
9. These replacements are therefore automorphic in contrast to the pseudomorphic 
. eplacements described hi the preceding section. Depending on the circumstances the 
same mineral species may replace the host pseudomorphically or automorphically. 


(Contrast Figure 4 of Plate 8 with pyritized fossils.) The reasons for this contrast are 
not fully known. Some minerals, such as quartz, barite, and pyrite, have a much 
stronger tendency than others to develop their own crystal outlines in replacement. 


Replacing minerals are often found along structural or textural features that would 
naturally be expected to guide penetrating solutions, such as fractures, bedding planes, 
schistosity, contacts between mineral grains of the same or different species, mineral 
cleavages, and mineral partings and twinning planes. 

Although much replacement is initiated from visible openings, one of the most 
striking features of the process is the ability of the solutions to penetrate along direc- 
tions of schistosity or contacts of mineral grains, where the highest powers of the 
microscope reveal no open channels. From such guiding features the solutions pene- 
trate and replace the bordering minerals. 

Figure 3 of Plate 9 shows anglesite (dark) associated with galena (light) from the 
Maury mine, Patagonia district, Arizona. Triangular pits in the galena indicate the 
three cleavage directions. The vein form of the anglesite and the fact that the vein- 
lets follow cleavage directions of the galena show that it either replaces the galena 
or fills fractures in it. The ragged outlines of the veinlets and the failure of their walls 
to match is evidence of replacement. Figure 14 shows replacement of galena by poly- 
basite, guided by galena cleavages and by the contacts of mineral grains. Commonly, 
as in Figure 14, only a single guest mineral is involved, but in other instances guided 
replacement involves a succession of replacing minerals as shown in Figure 15, in 
which galena was first replaced along quartz contacts and along cleavages by an un- 
identified mineral probably a sulphide of lead and silver and this in turn was re- 
placed by argentite. 

In other replacements one or more host minerals have been replaced by an aggre- 
gate of from two to many guest minerals. In the ore, shown in Figure 1 of Plate 10 
galena has been replaced by argentite along contacts of quartz with galena and along 
contacts of differently oriented galena grains. In two places, however, segments of 
electrum (stippled) take the place of argentite. The replacements are believed to 
be hypogene since electrum is not known as a supergene mineral. Alternative ex- 
planations are possible. Electrum and argentite may have simultaneously replaced 
galena through the agency of the same solution. If so, it seems strange that the two 
very different minerals should have replaced the galena at the same rate, as shown by 
the fact that the galena border remains smooth in passing from contact with argentite 
to contact with electrum. A perhaps more acceptable interpretation is that the galena 
was first replaced peripherally only by argentite and that through the agency of later 
solutions argentite was replaced selectively and only locally by electrum. 

In other ores (Fig. 16), comparable textures were developed by supergene enrich- 
ment. A zone of argentite and covellite has developed along the contacts between 
quartz and sphalerite. Although the replacement run varies in width its width does 
not change abruptly in passing from argentite to covellite. We seem here to have the 



alternative possibilities of (1) simultaneous replacement of sphalerite by covellite 
and argentite, or (2) replacement of sphalerite by covellite alone followed by later 
and localized selective replacement of covellite by argentite, or (3) peripheral re- 
placement of sphalerite by argentite beginning along quartz contacts and followed 



FIGURE 14. Guided replacement of galena 

Penetration of replacement along galena cleavages (at A), along contacts between galena grains, and along contacts 
between quartz and galena. Triangular pits developed in grinding and polishing define the cleavage directions in each 
galena crystal. At A replacement by polybasite has followed cleavages in the galena, and a faint line visible in the poly- 
basite marks the position of the galena cleavage from which replacement started. Camera lucida drawing from lead-silver 
ore from Neihart, Montana. After Bastin. 

FIGURE 15. Successive replacement of galena by two minerals 

Guided replacement of galena along quartz contacts and along cleavages by an unidentified mineral, probably a sul- 
phide of lead and silver (intermediate mineral). This in turn has been replaced by argentite. Camera lucida drawing of ore 
from Liberty Bell Mine, Telluride, Colorado. After Bastin. 

by selective, localized replacement of argentite by covellite. Nothing in the textures 
shown in Figure 16 justifies a choice among these three possibilities. However, in the 
laboratory sphalerite is not readily attacked by silver sulphate solution whereas covel- 
lite is. This favors the second as against the third explanation. In another ore (Fig. 
17) galena was first replaced by an undetermined mineral x\ this in turn has been re- 
placed by a somewhat irregular association of argentite and chalcopyrite either sim- 
taneously or in sequence. Most of the chalcopyrite follows the quartz-argentite con- 
ui tacts, favoring the interpretation that the chalcopyrite is the youngest mineral in 
the section and is a replacement mainly of argentite and subordinately of mineral x. 
While it is uncertain whether the relations illustrated in Figures 16 and 17 are the 
result of contemporaneous replacement by two guest minerals, simultaneous replace- 
ments by a group of several minerals are common, particularly in hypogene ore de- 
position. In the ore shown in Figure 2 of Plate 10 there are two groups of minerals, 
a relatively coarse group consisting of galena and quartz and a much finer-grained 



group consisting of carbonate, argentite and chalcopyrite. The minerals of this finer 
group replace the coarse quartz and galena. 

A very characteristic replacement textun is produced when a brittle mineral with- 
out cleavage, such as pyrite, is irregularly fractured and then partially replaced along 

O.I mm. 

O.3 mm. 



FIGURE 16. Replacement of sphalerite (spti) by covellite (c) and argentite (arg) 
q is quartz 1000 ft level of Tonopah Belmont Mine, Tonopah, Nevada. X 350 After Bastm 

FIGURE 17. Complex replacement of galena (gal) 

Along quartz (q) contacts Unidentified replacing mineral (x) is probably a lead-bearing silver sulphide This in turn 
has been replaced by an irregular association of argentite (arg) and chalcopyrite (black). Camera lucida drawing of pol- 
ished surface. Jim Butler Mine, Tonopah, Nevada. After Bastin. 

the fractures. On a small scale this texture is shown in Figure 5 of Plate 9 in which 
all stages from incipient to nearly complete replacement of Pyrite are shown. On a 
larger scale similar phenomena are shown in Figure 6 of Plate 9. None of the numerous 
pyrite areas are in contact with each other. Also, the walls of the veinlets between 
the pyrite areas do not "match", and groups of pyrite areas cannot be fitted together 
like the fragments of a jig-saw puzzle. All these features indicate replacement and 
are not harmonious with the alternative explanation of a simple filling of the inter- 
spaces of a breccia, although, as might be expected, in some instances both filling 
and replacement have taken place. In some instances faint lines along the middle of 
the veinlets mark the position of the crack along which replacement started. In still 
other instances replacement has proceeded further, and a second replacement mineral 
appears about midway between the areas of host minerals i.e., guest mineral No. 2 
begins to develop where guest mineral No. 1 began to develop. Replacement contacts 
are tight or sealed contacts and later replacements seldom if ever follow along them. 



The fractured and partially replaced pyrite shown in Figures 5 and 6 of Plate 9 
has no microscopic character that defines the crystallographic orientation of the 
several pyrite areas. If such a character were present it would probably reveal that 
entire groups of pyrite areas were parallelly oriented. In galena, however, triangular 
pits show the crystallographic orientation of each grain, and frequently neighboring 
areas of galena now separated by such minerals as anglesite are all parallelly oriented 
(PL 9, fig. 4). It is inferred that the galena was originally one grain and that replace- 
ment encroaching along cleavage directions has subdivided it. Such parallel orienta- 
tion patterns become a valuable criterion of replacement when supported by other 

On a much larger scale, parallel orientation of bedding in isolated rock fragments 
may Indicate replacement. 


Nature draws no sharp boundaries between the types of replacement just de- 
scribed that have been conspicuously guided by structural and textural features of 
the host and other replacements in which control by fractures, mineral cleavage, con- 
tacts of mineral grains, etc., has not been conspicuous. Nevertheless, in many re- 
placements the replacing solutions have evidently been able to penetrate the host with 
about equal facility in many or at least several directions, and it is sometimes dim- 
cult to determine why the guest mineral is where we find it with respect to the host. 
If fractures and grain contacts have guided replacement, these, as in the case of some 
fine-grained rocks, were so small or so numerous and diversely oriented that replace- 
ment proceeded about as readily in one direction as in another. Or the composition 
or porosity of the host or the composition of the replacing solutions were more im- 
portant than structural or textural controls. Such replacements show characteristic 
textures quite different from the guided replacements already described, and it may 
be helpful to group them under a separate caption of "Diffuse penetration textures." 
As in many instances the guest minerals show their own characteristic crystal out- 
lines, such automorphic replacements will also be considered. 

Unquestionably the physical character of the host is often responsible for diffuse 
penetration. It is common in ores that have replaced fine-grained sediments as in 
those of Mount Isa (PL 11, figs. 1,3). Occasionally the host is a fine-grained igneous 
rock such as felsite. The physical character of individual mineral grains of the host 
is sometimes important for Schouten (1946, p. 373, pi. 5) has shown that in the ores 
of Meggen, Germany, marcasite is more readily replaced than its associated pyrite. 

In other cases fair-sized crystal grains have been penetrated by the replacing solu- 
tions in diffuse fashion because they lacked both fractures and mineral cleavage; a 
good example is seen in Figure 18 where silver replaces pearcite, which lacks cleavage. 
The pearcite has characteristic curved outlines concave outward as if the silver had 
"bitten" into it. Conversely, the rounded silver outlines, convex outward, seem to 
indicate movement of the silver-bearing solutions outward from numerous centers. 



The chemical constitution of the host rather than its physical properties seems in 
many instances to have favored diffuse penetration, for some of the best examples of 
diffuse penetration are in ores in which the metallic minerals have replaced a car- 
bonate gangue (PL 11, fig. 4; PL 12, fig. 4). In such cases the host was apparently so 

FIGURE 18. Native silver replacing pearceile 

Aspen, Colorado. Camera Iuci4a drawing. X 360. After Bastin. The convex curved outlines of the silver against 
pearceite are noteworthy 

easily replaced in all directions that the guidance of cleavage and grain contacts be- 
came trifling advantages. It is not uncommon, however, to find even in the same ore 
specimen both diffuse penetration and guided penetrations in the replacement of 

Finally, in many instances the nature of the guest determines whether the replace- 
ment i^ of the diffuse-penetration or guided-penetration type. Most automorphic 
replacements are partly independent of the structural or textural features of the host, 
and certain minerals such as pyrite, quartz, and barite have a stronger tendency than 
others to assume their own crystal outlines in replacement. These crystal outlines 
transgress the textural features of the host (PL 9, fig. 2) . Some automorphic replace- 
ments are guided by major structural features of the host, such as bedding, but are 
independent of minor features. 

One of the simplest forms of diffuse-penetration texture is exemplified hi Figure 1 
of Plate 11. Both automorphic crystals and spheroidal pellets of pyrite have re- 
placed shale. Etching with acid reveals that some of the pellets are concentric. 

Slightly more complex is the so-called "Atoll Structure" well displayed in the ores 
of Mount Isa, Queensland (PL 11 fig. 3). Here spheroidal masses are made up of a 
shell of pyrite within which occurs another sulphide, either galena, sphalerite, or 
chalcopyrite. All these sulphides are replacements. There are often breaks in the 
pyrite shell, and through such breaks the galena or sphalerite within the pyrite shell 
are often continuous with areas of the same minerals outside the shell. The other sul- 


phides are considered younger than the pyrite. As mineralization proceeds they re- 
place the pyrite iwitil only remnants of the pyrite rims remain. They may also replace 
the shale. Rarely, at Mount Isa the "Atolls" occur in small groups somewhat re- 
sembling "tubercle" textures. 

"Pellets" may be clustered as is beautifully shown in a Swedish silver ore (PL 11, 
fig. 2) in which clusters of spheroids of native silver (white) are surrounded by ram- 
melsbergite, NiAsz. Both have replaced calcite (black). 

Next in complexity among the diffuse-penetration textures are those the writer 
has termed "tubercle" textures (PL 11, figs. 4-7). These textures are particularly well 
developed in the cobalt-nickel-silver ores of the Cobalt district and neighboring camps 
in Ontario, but they are also characteristic of the cobalt-nickel-silver ore type in the 
Erzgebirge of Saxony and of similar ore occurrences in several other widely separated 

To interpret properly the tubercle textures one must know the general order of 
deposition among the minerals as shown by other textural features. In the Cobalt 
district, Ontario, the order of deposition of the dominant ore minerals as shown by the 
relationships of veinlets, by the arrangement of minerals in small vugs or "nests " 
and by unquestionable replacement relationships is: older calcite, lollingite, niccolite 
and breithauptite, and native silver with younger calcite and minor polybasite. In 
the neighboring South Lorraine district the sequence as determined by the same 
criteria is: older calcite and a little quartz, safflorite with some arsenopyrite and 
cobaltite, nicolite with some tetrahedrite, and native silver and younger calcite. In 
both districts there is some overlap among the members of the sequence. 

The simplest tubercle texture is illustrated by Figure 4 of Plate 1 1 showing an ore 
from South Lorraine in which the tubercles are composed of a single metallic mineral 
arsenopyrite in a calcite matrix. The arsenopyrite (at right) shows its own char- 
acteristic crystal faces against calcite both on the inside and on the outside of the 
tubercles. At the extreme right a few small isolated arsenopyrite crystals occur in the 
calcite. The form of these "tubercles" is not what one would expect from deposition 
in open spaces. They are interpreted as automorphic replacements of the calcite by 
arsenopyrite through the agency of solutions penetrating the calcite in diffuse fash- 
ion from several feeding centers. 

Slightly later in the mineralization process in the same mine tetrahedrite was 
deposited by continued diffuse penetration as shown in Figure 6 of Plate 1 1 . This is 
more highly magnified than Figure 4 of Plate 11; it shows clearly the automorphic 
outlines of the arsenopyrite on both the inside and outside of the tubercles and shows 
the deposition of tetrahedrite through the replacement of calcite both inside and 
outside the arsenopyrite tubercles. Tetrahedrite within the arsenopyrite tubercles 
is in places continuous with that outside. It conforms to the automorphic outlines 
of the arsenopyrite and is therefore somewhat younger. In other ores of Cobalt and 
South Lorraine that show tubercle textures, lollingite or safflorite rather than ar- 
senopyrite formed the original tubercles, and niccolite and breithauptite rather 
than tetrahedrite formed the cores. In many instances native silver, niccolite, and 
breithauptite seem to have simultaneously replaced the original calcite cores of the 
tubercles. Much native silver is, however, somewhat younger than niccolite and 


breithauptite (PL 11, fig. 5). Why tubercle textures should be confined mainly to 
ores of the cobalt-nickel-silver type is not clear; the presence of an easily replaceable 
calcite gangue is probably one factor, and the strong tendency of certain arsenides 
such as arsenopyrite, smaitite, and safflorite to form automorphic crystals may be 

Keil (1931) has described from the cobalt-nickel-silver ores of Annaberg, Saxony, 
the textures shown in Figure 2 of Plate 12. Rather irregularly distributed crystals of 
native bismuth are enveloped in safflorite, and both lie in a quartz gangue. Beyond 
the main areas of safflorite in the upper part of this figure are several star-shaped 
twins of safflorite isolated in the quartz. It is difficult to see how isolated crystals of 
bismuth suspended in a tenuous mineralizing solution could have served as nuclei 
about which safflorite crystallized, and how isolated symmetrical star twins of 
safflorite could have remained suspended in a tenuous mineralizing solution and be- 
come enclosed in quartz. These textures resemble the tubercle textures found in the 
same ores and like them are interpreted as replacement textures. As a result of diffuse 
penetration of silica by the metal-bearing solutions scattered bismuth crystals wtre 
first deposited and safflorite was then deposited around them. 

Odman (1945) has described in a cobalt-nickel-silver ore from Sweden dendritic 
crystals of native silver surrounded by calcite. In Figure 4 of Plate 12 the boundaries 
of the calcite grains can be seen, and the silver crystals bear no relation to either the 
cleavage or the grain contacts of the calcite. Figure 5 of Plate 12, taken from ore 
from the same mine, clearly pictures replacement of calcite by diffuse penetration 
of silver-bearing solutions. The rudely tuberclelike outlines of some of the silver is 


Skeletal crystals, as delicate as many that can be formed artificially by crystalliza- 
tion from concentrated solutions, form naturally in ores by replacement. Such 
forms are particularly common in the ores of the cobalt-nickel-native silver type 
both in North America and in Europe. Most of these skeleton crystals are either 
native silver or native bismuth rimmed by cobalt or nickel arsenides; the matrix is 
usually a mixed carbonate such as dolomite. Figures 1 and 3 of Plate 12 show the 
typical forms of the skeleton crystals of native silver enclosed by lollingite from 
Cobalt, Canada. As native silver is isometric the skeleton crystals usually branch at 
right angles and show octahedral faces. Native bismuth is rhombohedral and shows 
skeleton crystals quite unlike those of native silver. 

Some misconceptions concerning these skeletal crystals can be corrected as the 
number of described occurrences has increased. In spite of the rather obvious en- 
closure of automorphic silver crystals by the arsenides some observers interpreted 
the silver as a later filling of the cores of arsenide dendrites. The arsenides however, 
have clearly replaced the silver of the dendrites in places. In Figure 3 of Plate 14 
some of the silver contacts are ragged replacement contacts eating into the elsewhere 
straight boundaries of the silver. Furthermore some of the arsenides enveloping 
silver such as rammelsbergite and safflorite are orthorhombic and thus could not 
have determined the isometric form of the enclosed silver. In the cobalt-nickel-silver 


ores of the Laver mine in northern Sweden some dendrites of native silver have 
arsenide envelopes (PL 14, fig. 3), but others (PL 12, fig. 4) are without any envelope 
of metallic minerals. 

Van der Veen (1925), who studied many occurrences of skeletal crystals of silver 
and of bismuth, nottJ that in some occurrences in the Erzgebirge and at Cobalt, 
Canada, the cores within the arsenide envelopes were calcite rather than silver. 
Even in the same specimen some cores were calcite and others were silver. However, 
the cores of calcite were identical in form with those of silver. Van der Veen believed 
that the original dendrites were calcite formed by the alteration of vein dolomite 
early in the mineralization. The native silver he regarded as a replacement of original 
calcite dendrites. How isometric dendrites could be produced by rhombohedral 
calcite is not explained. Furthermore the accessory minerals in some of the calcite 
dendrites are chalcopyrite, galena, argentite, sphalerite, and rarely proustite, all 
known to be late minerals in these ores. Their testimony points to the caclite of the 
dendrites as one of the latest rather than one of the earliest minerals of the ores. 
Also Keil (1931) has shown that at Marienberg, Germany, and at Cobalt silver 
dendrites are partially replaced by calcite. In Figure 1 of Plate 14 within the arsenide 
envelope the calcite is usually peripheral, enclosing replacement remnants of silver. 
Similar relations in ores from Cobalt are shown in Figure 2 of Plate 14. 

To check these age relations further Bastin re-examined a specimen of silver den- 
drites from Cobalt, Ontario, now in the University of Chicago collections and shown 
in Figure 3 of Plate 12. Megascopic examination of this hand specimen in reflected 
polarized light (through polaroid spectacles) shows that the matrix of the dendrites 
is an aggregate of diversely oriented carbonate grains and that the orientation of the 
dendrites is independent of either the cleavage or the grain boundaries of the matrix. 
The silver exhibits skeletal crystal forms of isometric pattern. The microscope shows 
that most of the borders of the silver are sharp, straight octahedral faces. Irregular 
borders are usually due to intergrowth with niccolite of a smaller order of grain 
size. Niccolite may occur even near the centers of the silver crystals but is most 
abundant just inside or just outside the silver border. The outlines of the very small 
niccolite inclusions are rotund, but the larger ones are automorphic against the silver. 
Clear evidences of replacement of niccolite by silver are very rare. 

The thin envelope of arsenides around the silver is mainly lollingite (PL 12, fig. 3). 
The silver usually shows sharp crystal boundaries against lollingite and any overlap 
in age between silver and lollingite is slight. The inner part of the lollingite is much 
intergrown with niccolite, but the outer part is free from niccolite. Peripherally the 
lollingite is automorphic against carbonate. 

In a few places the cores within the arsenide envelope are carbonate although 
they show the same forms elsewhere shown by the silver. There has thus either been 
a replacement of silver (and some niccolite) by carbonate or the reverse is true as 
Van der Veen concluded. In choosing an explanation it should be noted first that 
silver cores are the rule, and carbonate cores the local exception. Every gradation 
may be found from cores almost completely silver (with some niccolite) to cores 
completely carbonate. Within a single dendrite calcite may supplant silver for a 
short distance; then silver may come in again, all without change in the isometric 


outlines of the dendrite. In places, areas of silver with the "ragged" outlines so often 
characteristic of replacement are enclosed by calcite. The writer agrees with Keil 
(1931) that carbonate has replaced silver; usually the small niccolite areas originally 
enclosed by the silver have also been replaced by the carbonate, but in places a few 
niccoiite remnants remain. 

Perhaps the clearest evidence that the isometric dendrites of the ores of the cobalt- 
nickel-silver type were originally silver and were later partly replaced by calcite is 
found in the ores of the Miller Lake-O'Brien mine at Gowganda, Ontario. The writer 
collected and studied representative silver ores from this mine in 1948. The wallrock 
is exclusively Nipissing diabase much of which has been extensively replaced by 
calcite. In this rock native silver occurs as fracture fillings and as replacements; 
and dendrites or skeletal crystals are well developed among the replacements. In 
cross sections of the dendrites (PL 13) three minerals are shown native silver, 
loliingite, and calcite. Most of the silver dendrites have spider-like outlines, four 
or five irregular arms diverging from a center. In Figure 1 of Plate 13 the core of the 
dendrite is wholly silver. In Figure 2 of Plate 13 the core is mainly silver, but replace- 
ment (in part automorphic) by calcite has begun at the tips of two arms. In Figure 
5 of Plate 13 more than half of the original core has been replaced by calcite. In 
Figure 3 of Plate 13 there has been spotty replacement of the silver core at six points. 
In Figure 6 of Plate 13 calcite has completely replaced the silver core. 

Commonly the replacement of silver by calcite begins in the peripheral portions 
of the silver cores, often at or near the tips of the arms. Commonly small veinlets 
of calcite connect the calcite of the matrix with the calcite patches that are con- 
tiguous to the silver. These veinlets presumably represent the channels along which 
calcite-bearing solutions penetrated the lollingite to reach the easily replaceable 
silver. These veinlets are particularly well shown in Figure 3 of Plate 13 which also 
shows that the calcite usually replaces the silver automorphically. 


Table 1 lists the described occurrences of dendrites in ores of the cobalt-nickel- 
silver type and shows that (1) Bismuth dendrites are enveloped by some of the same 
minerals that envelop silver. (2) Among the 22 silver ores listed, 15 have cores domi- 
nantly silver which is usually free from accessories. Five are dominantly calcite, 
and two are dominantly proustite. Where calcite is the dominant mineral of the 
core there are usually present as accessories one or more of the minerals galena, 
sphalerite, argentite, chalcopyrite, and rarely proustite, all of which are charac- 
teristically minor and late minerals in this type of ore; this suggests that calcite is 
also a late mineral. (3) A great variety of minerals may form the envelope of the 
dendrites. Of these, lollingite and rammelsbergite are orthorhombic, niccolite is 
hexagonal, and pitchblende is amorphous; none of these could possibly be responsible 
for the isometric forms of the silver dendrites. These relations support the interpre- 
tation that the original minerals of the dendrites were silver and more rarely bis- 
muth and that the silver dendrites were not uncommonly replaced by calcite which in 
some instances was accompanied by galena, chalcopyrite, and proustite. 

Some of the more important references on dendritic and other textures in ores of 


the cobalt-nickel-silver type as listed in the bibliography are: Bastin (1922; 1925; 
1949), Keil (1931), Montgomery (1948), Odman (1945), Van der Veen (1925), 
Zuckert (1925). 


Diffuse penetration sometimes results in the deposition of minerals in concentric 
bands with a more or less regular spacing that seems to imply a rhythm in the pre- 

FIGURE 19. Diagram illustrating relation of diffusion banding to fractures 
Leadville, Colorado. Arrows show directions of diffusion. After Loughlin. 

cipitation process. Such banding has been termed rhythmic or diffusion banding. It 
is usually related to fractures in the host but is little influenced by lesser textural 
features. Figure 5 of Plate 14 shows diffusion banding developed during oxidation in 
an even-grained felsite from Tonopah, Nevada (Bastin and Laney, 1918 p. 45-46). 
The microscope reveals the cubical outlines of former pyrite grains that were dis- 
tributed rather evenly through the rock. The pyrite grains have been completely 
oxidized, but the product of their oxidation, hydrous iron oxide, has not remained 
where the pyrite grains were but has become distributed in bands. No banding was 
produced in the coarse fragments enclosed in the felsite. The general directions of 
flow of the solutions that produced the bands can easily be inferred. 

Diffusion banding is usually restricted to fairly fine-grained rocks. At Tonapah, 
oxidation banding occurs not only in the felsite shown in Figure 5 of Plate 14 but also 
in trachyte porphyry, a notably coarser rock. It is common elsewhere in sandstones 
and is found in some quartzites. It may also be developed in great delicacy in col- 
loidal silica as familiarly exemplified by agates. 

Most oxidation bands are yellowish brown; the coloring material is finely divided 
hydrous iron oxide. In the Precambrian quartzites of Baraboo, Wisconsin, the bands 
were presumably originally hydrous iron oxide, but this was dehydrated during the 
rnetamorphism of sandstone to quartzite to form fine red hematite. 

Although little influenced by the textural features of the rock in which it occurs, 
oxidation banding is usually related to joints and other fractures much as is hypo- 
gene diffusion banding (Fig. 19). 

Oxidation banding cannot be the product of a single episode of diffusion but repre- 
sents the composite effects of many episodes of wetting and drying, solution and 
deposition, during prolonged periods. 




Author and number 
of figure 

Dominant min- 
eral of dendrite 

Accessory minerals 
of dendrite* 

Rim or Rims 
of dendrite 

General matrix 


Van der Veen 31 




Same Mine 

Van der Veen 32 


Arg., Sph. 


Van der Veen 33 


Gal., Arg. 




Van der Veen 34 


Gal., Arg. 



Van der Veen 35 


Gal., Arg. 

Ni, Ra, Saf. 



Van der Veen 36 


Arg., Pr., Chpy. 

Intergrown Ni 
and Dol. 



Van der Veen 37 


Feathery Ar- 


Van der Veen 60 


Feathery Ar- 


Van der Veen 109, 


Few dendrites 
are calcite 

Saf., Sm., Ni, 



Odm? n 4 and 5 




Laver Mine, 

Bastin (1925) VII-C, 





Keil 49 


Safflorite, Co- 



Guild XIX-A, B 





Keil 20 


Ra, Saf. 



Keil 46 


Calcite replac- 
ing silver 

Ra, Saf. 


Neumann VII- 14 





Krieger 7, 8 


Nickel Skut- 

Qtz. Cal. 

Peak, New 



TABLE 1. Continued 

Author and number 
of figure 

Dominant min- 
eral of dendrite 

Accessory minerals 
of dendrite* 

Rim or Rims 
of dendrite 

General matrix 


Bastin (1922) 35 



Quartz, Ba- 


Zuckert, Table 4, 
Fig. 14 




Quartz, Dol. 


Montgomery, 3, 4 






Kidd and Haycock, 


Safflorite- Ram- 


Great Bear 

Kidd and Haycock 
743 and 4 


some Chpy. 


Great Bear 

Van der Veen 40 





Van der Veen 77 


Saf., Ra. 



Van der Veen 78 




Van der Veen 79, 80 


Saf., Ra. 



Abbreviations : 
















(Although diffusion banding is commonest as an oxidation phenomenon it is occa- 
sionally developed in hypogene mineralization^) The hypogene examples are most 
common in limestones and more rare in the carbonate gangue of ores. An example 
from the coablt-nickel-silver ores of Cobalt, Ontario (PL 14, fig. 4) shows concentric 
bands of smaltite developed by the replacement of calcite. Miller (1913, Fig. 11) 
illustrates a specimen showing 25 such bands in a calcite gangue. 

The rhythmic banding thus far described consists of bands of a single mineral 
alternating with remnants of the matrix which it has replaced. In other instances 
both bands consist of introduced minerals, and little if any of the original matrix 
remains. Such ores have been well described by Loughlin (Emrnons, Irving, and 
Loughlin, 1927, p. 202-205). In this ore (PL 15, fig. 1) limestone has been replaced 


by alternate bands of pyrite (light) and sphalerite (dark). Width and spacing of the 
bands are strikingly uniform. In the Leadville ores pyrite and sphalerite are the 
dominant minerals of the bands, but minor amounts of chalcopyrite or galena are iii 
places present. Since, in some ores in other regions all the bands are of one mineral, 
possibly in the Leadville ores the pyrite bands were first formed and the sphalerite 
deposited later by replacement of the limestone remnants between the pyrite bands. 
Because the successive bands of sphalerite are occasionally connected across the 
intervening pyrite band by tiny sphalerite veinlets, Loughlin (Emmons, Irving, and 
Loughlin, 1927, p. 202-205) believes that at Leadville the sphalerite bands were 
formed slightly later than the pyrite bands. Whether such age diversity character- 
izes ail double-banded ores is not yet certain. Double-banded ores almost identical 
with those of Figure 1 of Plate 15 form a minor feature of the fluorspar veins of 
Kentucky (Bastin, 1931). The bands are sphalerite and fluorite that have replaced 
vein calcite. 

While diffusion banding is largely independent of the minor textures of the rocks in 
which it is formed it may be markedly controlled by joints and other fractures and 
sometimes by bedding. Fractures, indeed, usually represent the channels through 
which the replacing solutions entered the rock as is diagrammatically shown by the 
arrows in Figure 19. 

Diffusion banding that partially simulates the natural occurrences can easily be 
produced in the laboratory. Liesegang (1913) allowed silver nitrate solution to diffuse 
into gelatine containing potassium bichromate and obtained a series of bands of silver 
bichromate spaced at progressively greater intervals away from the center of diffusion. 
Ostwald termed these bands "Liesegang rings" and believed that as the slilver nitrate 
solution diffuses outward a supersaturated solution of silver bichromate is formed; 
when the metastable limit of this solution is exceeded the silver bichromate is pre- 
cipitated, forming the first ring. Continued outward diffusion of the silver nitrate 
causes renewed formation of silver bichromate solution in the zone surrounding the 
first ring until precipitation again ensues, and so on. As the source of silver nitrate 
becomes depleted, the solutions must diffuse farther before supersaturation is reached 
and the successive rings are deposited at increasing intervals. 

These and other experimental productions of diffusion banding are instructive, but 
the conditions under which diffusion banding forms in nature may involve many 
other factors. Most artificial diffusion banding has been formed in gels that were inert 
media not reacting chemically with the diffusing solutions. The color banding in 
agates may not have involved chemical interaction with silica gel, but the develop- 
ment of diffusion bands in limestones or in carbonate gangues involved solution of the 
carbonates. The natural banding is not the product of a single episode of diffusion 
but the combined product of many episodes during which the diffusing solutions in 
many cases changed in concentration, composition, and temperature. 

Double banding due to diffusion (PL 15, fig. 1) would once have been interpreted as 
due to successive deposition in open spaces and hence as implying sudden and fre- 
quent alternations in composition of the mineralizing solutions. Under the diffusion 
explanation no such unusual changes need be postulated. 

Lindgren described very delicate diffusion banding in certain ores of Tintic, Utah. 


The intervals between these bands are 1 millimeter or less (PL 15, fig. 2). The bands 
are often curved in beautiful concentric arrangement. The light bands consist of 
spherules of chalcedony cemented by granular quartz with undulatory extinction. 
The dark bands contain irregular grains of galena, rounded grains of sphalerite, and 
a few small cubes of pyrite. According to Lindgren colloidal silica first replaced 
limestone; sulphide-bearing solutions diffusing through the gelatinous silica then 
deposited sphalerite and galena in successive bands. 


As already indicated certain rock types, notably carbonate rocks, are particularly 
susceptible to replacement, while others, notably shales and slates, are usually resist- 
ant. Obviously, too, replacement is as much dependent on the composition of the 
mineralizing solution as upon the composition of the rock; for example, relatively 
pure limestones are usually particularly susceptible to contact-metamorphic replace- 
ment, but at San Jose (Bastin, 1937), Tamaulipas, Mexico, the intrusive rock and 
chert nodules in the limestone rather than the pure limestone have been meta- 
morphosed. As the diorite is essentially free from quartz, solutions from the diorite 
magma were low in silica and probably rich in carbonates. They were thus incapable 
of reacting with the limestone. 

In monomineralic rocks such as limestones there is, except for scattered impurities, 
no opportunity for selective replacement; such rocks are usually replaced en masse 
up to a rather definite boundary. In rocks made up of many minerals on the other 
hand replacement is likely to be notably selective. For example, pyrite developed by 
replacement in granite is often confined to biotite grains, the iron of the biotite appar- 
ently entering into the pyrite. The potash mica, sericite, selectively replaces the 
potash feldspar orthoclase. Replacement that is at first highly selective may be less 
so and spread to other minerals as the process goes on. In the ore shown in Figure 2 
of Plate 10 replacement was at first confined to galena, but later, as at A, quartz 
began to be replaced. 

In automorphic replacements in which the replacing mineral assumes its own 
crystal outlines notably different relations exist. The crystals of pyrite shown in 
Figures 4 of Plate 8 and Figure 2 of Plate 9 may have begun their growth as selective 
replacements of some iron-bearing silicate, but as the crystal grows its faces come in 
contact with many rock-forming silicates, possibly a half dozen or more. As the 
pyrite crystal grows, always maintaining its cubical form, all these rock silicates are 
replaced simultaneously and at approximately the same rate. The full explanation of 
such replacements is not yet clear. 


Graphic intergrowths may be formed in several ways. They may be formed in 
igneous rocks as a result of simultaneous crystallization (Fig. 1), rarely they seem to 
have formed during contact metamorphism (Schwartz, 1931, p. 755, Fig. 18), and 
there is ample evidence that they are developed in ores by replacement. In an excel- 
lent example described by Lindgren from the Tintic district, Utah (PL 15, fig. 4), 
galena is traversed by bladelike crystals of barite and is in "graphic" association 


with tennantite and pearcite. The barite has not been replaced by the galena, but it 
has been replaced, in transecting fashion, by the tennantite and pearcite. Therefore 
these minerals must also have replaced the galena. 

Another excellent example of graphic texture developed by replacement (PL 15, 
fig. 3) pictures ore from the Engels Mine, Plumas County, California. Here the 
chalcocite in intergrowth with bornite can be traced into continuity with chalcocite 
that has replaced bornite along quartz contacts. 

In general, the evidences cited for a replacement origin of graphic associations of 
the type just described are (1) boundaries that are serrate under high magnification 
and from which tongues of the guest project into the host, (2) veinlets of the guest 
traversing the host (PI. 15, fig. 3), (3) occasional presence in the guest of a median 
line marking the position from which replacement started, (4) transitions from the 
graphic associations into unquestioned peripheral replacements. 

Schwartz (1930) has described a graphic association of cerussite and covellite that 
is a replacement of an original graphic association of galena and stromeyerite. 

Little or no system or regularity can be found in the graphic intergrowths formed 
by replacement. 


Under constant environment a group of crystallizing minerals tends toward moder- 
ate uniformity in grain size. Notable departure from such uniformity by certain 
minerals would, therefore, imply changed conditions, and hence age difference. The 
interpretation of phenocrysts of porphyries as differing in age from their matrix is a 
familiar application of this principle. There are notable exceptions to this rule, but it 
is valuable if used with caution. 

In replacement in ores the guest minerals may be notably smaller hi grain than 
those of the host (PL 10, fig. 2), or they may be notably larger (PL 9, fig. 2). 


In automorphic replacements the guest mineral develops its characteristic crystal 
faces, and its contacts with the host are smooth, straight crystal faces (PL 8, fig. 4). 
In pseudomorphic replacements the guest-host contact is likely to be curved or 
sinuous but may be either ragged or smooth, and the reason for the contrast is not 
always clear. In the ore shown in Figure 2 of Plate 10 very ragged contacts are de- 
veloped in the replacement of galena, whereas in the graphic replacements (PL 15, 
fig. 4) the galena contacts are smooth. Both are believed to be hypogene replacements. 

In some replacements the host shows scalloped outlines and appears "bitten-into" 
by the guest mineral. This texture is well shown in ore from Aspen, Colorado in which 
native silver contains abundant replacement remnants of pearcite (Fig. 18). 


It is usually important in the study of ore deposits to determine whether replace- 
ments were accomplished by deep-seated (hypogene) or surficial (supergene) solu- 
tions. The criteria thus far used are few and are not textural for replacement patterns 
are identical in hypogene and supergene replacements. 


The most readily available criterion is based upon knowledge of the geological and 
geochemical conditions under which various minerals form. This knowledge rests 
mainly on field observation and subordinately on laboratory experimentation. Cer- 
tain minerals form only under the near-surface conditions of the oxidized zone. Re- 
placements involving these minerals may safely be interpreted, therefore, as super- 
gene. A common example is the development of anglesite and cerussite along galena 
cleavages (PL 9, fig. 3). Replacements by limonite and malachite are also supergene. 

On the other hand, certain associations of lime silicates such as garnet, epidote, 
and wollastonite, which replace limestones, form only under conditions of hypogene 
mineralization. In certain rich ores of Tonopah, Nevada, electrum occurs as a replac- 
ing mineral. It is believed to be hypogene because in the zone of oxidation gold and 
silver tend to part company. Consequently, if they are redeposited from descending 
solutions they are deposited, not in alloy with each other, but as separate minerals 
such as argentite and relatively pure gold. 

Many minerals can form under either hypogene or supergene conditions, and these 
include native silver, silver sulphides, and probably the sulphosalts of silver, whose 
origin it is often of great practical importance to diagnose. 

In supergene replacement within the zones of oxidation and of downward enrich- 
ment two minerals seldom simultaneously replace the same host. They are usually in 
sequence such, for example, as galena first replaced by argentite and the latter by 
native silver. A few questionable instances of simultaneous supergene replacement by 
two minerals have, however, been described (Fig. 16). 

In hypogene mineralization apparently simultaneous replacement of one or more 
host minerals by an aggregate of many guest minerals is common. Certainly the 
several minerals of the replacing aggregate show no evidences of any considerable age 
diversity. The ore from Tonopah, Nevada (PL 10, fig. 2), is an excellent example. 
Many great replacement ore bodies in limestone like those of Leadville, Colorado, 
afford larger-scale examples. It must not be inferred that sequences may not develop 
in hypogene replacement, but they appear much less characteristic than in supergene 

Finally, depth relationships may be a criterion of hypogene vs supergene replace- 
ment. Replacements on an important scale at depths well below the effective circu- 
lation of meteoric waters suggest hypogene origin. On the other hand, the limiting 
in certain districts of some kinds of replacements to very moderate depths suggests 
supergene origin. At Aspen, Colorado, replacement of argentite and some other 
silver minerals by native silver is common near the surface but diminishes with 
depth and has not been noted below 1200 feet. The frequent limitation of chalco- 
cite to depths of a few hundred feet is another example. 


Replacements have been produced hi the laboratory by a number of experimenters. 
Some replacements were incidental to experiments to elucidate the chemistry of 
natural replacement processes (Young and Moore, 1916). The most extensive were 
those of Ray (1930) carried on primarily to aid in understanding the natural proc- 
esses of copper enrichment. Polished ore fragments composed of associations of 


pyrite, bornite, chalcopyrite, chalcocite and sparse enargite were placed in a steel 
bomb with distilled water and heated for 3 to 5 days at temperatures of 100 to 
175C. All the specimens showed complete dispersion of small bornite masses into 
the enclosing chalcocite presumably as a result of solid solution; there was also 
incipient replacement of bornite by chalcopyrite and of pyrite by chalcocite. Next, 
similar specimens were placed in pyrex bombs with distilled water and heated in a 
water bath at 90-100C. for from 7 to 20 days. Changes in the specimens were 
identical with those obtained at higher temperatures. Finally in similar experiments 
with pyrex bombs one or more of the substances copper sulphate, sulphur, sodium 
carbonate, and ferric sulphate were added to the water and heated at 90 to 100C. 
With these materials replacements were also readily obtained involving, in addition 
to those obtained with pure water, a replacement of bornite and chalcocite by covel- 
lite, of sphalerite by chalcocite, and of galena by covellite. In these experiments a 
number of natural replacement textures were imitated; delicate micrographic inter- 
growths of bornite and chalcocite and of sphalerite and chalcocite were produced. 

The most elaborate and instructive experiments were conducted by the Dutch 
geologist Schouten (1934). He made no attempt to imitate closely in the laboratory 
the complex natural conditions under which replacement proceeds either as regards 
composition of the replaced substances, or the composition, temperatures, pressures, 
etc., of the replacing solutions. The purpose was to obtain clearly displayed replace- 
ment phenomena by any rapid and effective means, to study the textures produced, 
and to compare them with natural textures. Some of the textures produced have not 
been observed in nature. 

Several of the principles of replacement deduced from the study of ores are con- 
firmed by these experimental results. For example, the form of many of the synthetic 
replacements shows they were guided by fractures (PI. 15, fig. 5) even where these 
fractures are too small to be detected by the microscope in the unaltered portions of 
the host mineral. Some small cracks present in the host were preserved throughout 
replacement, indicating that fractures though usually younger than their hosts may 
rarely be older. 

In some experiments large automorphic crystals proved more resistant to replace- 
ment than fine aggregates of the same mineral (PL 15, fig. 6). Similar relations are 
not uncommon in natural replacements. Such relations show that automorphic out- 
lines do not invariably indicate complete exemption from replacement. These obser- 
vations are of great aid in interpreting the relations of silver to niccolite in ores from 
Cobalt, Ontario (PL 11, fig. 5). In these ores silver has clearly replaced niccolite 
extensively in the central portions of many of the "tubercles", and many of the con- 
tacts between silver and niccolite are typically ragged replacement boundaries. Yet 
some of the larger niccolite areas (smooth gray) retain their automorphic outlines. 

While many synthetic replacements are guided by textural features of the host, 
others seem not to be so controlled. Figure 7 of Plate 15 shows the replacement of a 
phosgenite (PbCO 3 -PbCl 2 ) crystal by galena without visible relation to the cleavage 
of the host. Somewhat analagous relations in natural replacements are illustrated in 
Figure 4 of Plate 12. 

Pseudographic textures were developed by synthetic replacement of chalcopyrite 


by bornite, and textures closely resembling some exsolution textures were produced 
by the synthetic replacement of bornite by chalcopyrite. 

Many synthetic replacements were notably selective, and, according to Schouten, 
in general the attack on different minerals becomes more selective as the solutions 
are more dilute. Chalcopyrite was replaced by chalcocite synthetically much more 
rapidly than was pyrite a fact commonly observed also in natural occurrences. 

Volume relations in synthetic replacements are significant. Most natural replace- 
ments seem to have taken place with little if any change in volume. For the most part 
these replacements have occurred below the ground-water level under a fair amount 
of pressure due to overload. Schouten's experiments were carried out in covered 
dishes at atmospheric pressures and temperatures below 100C. In some of the ex- 
periments notable increases in volume produced disintegration, and notable decreases 
in volume produced porosity. Figures 1 and 2 of Plate 16 show development of porosity 
in the synthetic replacement of cuprite by native copper obtained by treatment by 
a reducing solution of Sn(OK) 2 in KOH. The physical conditions of Schouten's 
experiments are more closely analogous to those of the zone of oxidation than of 
the deeper zones, and close study of the natural replacements of the oxidized zone 
may disclose that volume changes there are more common than at greater depths. 
In nature moving solutions and open chemical systems contrast with the quiescent 
closed systems in most laboratory experiments. Anderson and Merritt (1937), in 
studying the shallow replacements involved in the alteration of anhydrite to gypsum, 
in dolomitization, in serpentinization, and in the formation of bauxite, found no 
evidences of volume changes; the replacements were largely pseudomorphic. 

Finally, synthetic replacement may offer a valuable new technique as an aid in the 
interpretation of microscopic textures, particularly those of supergene mineralization. 
In some cases hardly perceptible textures may be made clearly visible by synthetic 
replacement as an alternative to the customary etching methods. 



Solid inclusions may be either younger, older, or the same age as the mineral that 
encloses them but it may be difficult to determine the age relation. The inclusions 
may be irregularly rounded or bladelike, or may have well-developed crystal outlines. 
Their sizes may be remarkably uniform or highly variable, their distribution even or 
erratic, regular or irrregular. Regularity in distribution may be related to the boun- 
daries of the enclosing mineral or to its internal features such as cleavage, parting, 
and twinning planes. In the present state of our knowledge the interpretation of inclu- 
sions may best be presented through examples whose interpretation is fairly clear. 

Certain inclusions have been formed by replacement and are therefore younger than 
the mineral that encloses them. Figure 3 of Plate 16 pictures an ore from Mount Isa, 
Australia, hi which inclusions of galena occur in calcite. The white dotlike inclusions 
occur at the intersections of twinning planes and cleavages in the calcite. Proof of 
origin by replacement is afforded not only by this fact but by transitions from round 
inclusions to typical replacement veinlets that follow either cleavage or twinning 

Small inclusions of chalcopyrite in sphalerite are common hi ores. Van der Veen 
(1925) has figured an occurrence in which these inclusions are distributed along the 
cleavage directions and the twinning planes of the sphalerite (PL 16, fig. 4). Quite 
possibly this, like the inclusions of galena in calcite just described, is a replacement. 
However, in this instance we have not a sulphide and a carbonate but two sulphides 
capable of forming a solid solution (Buerger, 1934). In the unmixing of a solid solution 
the minor component often develops along crystallographic directions of the major 
mineral. Possibly, therefore, the texture shown in Figure 4 of Plate 16 is due to un- 
mixing, and therefore the chalcopyrite and sphalerite are contemporaneous. Simi- 
larly arranged inclusions of pyrrhotite in sphalerite have been described from the 
Malanas district in Sweden (Gavelin, 1939). The general criteria of unmixing have 
been discussed and apply to microscopic inclusions as well as to the textural patterns 
usually referred to as intergrowths. Unmixing is apparently limited to closely related 
minerals within the same chemical group such as oxides, sulphides, or elements. Solid 
solutions of oxides and sulphides are not known. Uniformity also characterizes most 
patterns due to unmixing. However, by reheating, Schwartz (1931) has produced 
unmixing textures in ores from Bisbee, Arizona, hi which chalcopyrite is very irregu- 
larly distributed through bornite. 

Inclusions may help interpret the age relations of ore minerals. Figure 20 is a 
camera-lucida drawing of an ore from the San Jesus mine, Zacatecas, Mexico. In the 
absence of any evidences of replacement the idiomorphic quartz is interpreted as 
older than the freibergite which conforms to its outlines. Unmixing appears to be 
ruled out because of the chemical dissimilarity of the three minerals. The localization 
of sphalerite inclusions in the peripheral portions of the quartz and in the Freibergite 
suggests that sphalerite began to crystallize after quartz crystallization was well 




advanced and continued to deposit with freibergite after quartz crystallization had 
ceased. The paragenesis is interpreted as follows: 




FIGURE 20. Inclusions of sphalerite in quartz and freibergite 
San Jesuo Mine, Zacatecas, Mexico. X 53. After Bastin. 

Loughlin (Loughlin and Koschmann, 1942) has described from the Magdalena Dis- 
trict, New Mexico, blebs of chalcopyrite in sphalerite that are confined to the outer 
portions of the sphalerite crystals and are roughly aligned parallel to the crystal 
borders. These inclusions probably had a different origin from the chalcopyrite inclu- 
sions of Figure 4 of Plate 16. Loughlin was inclined to ascribe them to replacement, 
but the evidence is inconclusive. They may well be of the same origin as the sphalerite 
inclusions in quartz shown in Figure 20. 


Fluid inclusions in ores have received less study than those in normal igneous rocks 
and hi pegmatites. Usually they are detectable only in transparent gangue minerals 
and a very few metallic minerals such as light-colored sphalerite that are transparent 
in thin sections. In a few instances they have been recognized in opaque metallic 
minerals such as dark sphalerite and galena. Fluid inclusions in ores; however, de- 
serve much more study because some of them undoubtedly represent minute samples 
of the solutions from which the ores were deposited. Even qualitative studies of such 
samples are of value. 

Newhouse (1932) has reviewed the data on the occurrence of fluid inclusions in 
ordinary rocks and in ores. Most such inclusions are irregular but crystal outlines 
have been observed in a few fluid inclusions in quartz and galena. Buerger (1932) 
has described tiny cavities in galena bounded by octahedral and cubical crystal faces 
and filled with a sodium chloride brine. He believes "negative" crystal faces may have 
developed as normal crystal faces lining a slight depression in a galena surface; the 
depression then became boxed in or covered by rapid growth of an adjacent galena 



crystal. A few of these cavities are up to 1 centimeter in maximum dimension. 
Newhouse (1932, p. 430) has observed similar cavities in galena from many other 
localities and has tested the fluids they contained. 

The principal substances identified in fluid inclusions are: 


SO 2 
H 2 S 

Radicles in Solution 
Basic Acid 


SO 3 

H 2 
C0 2 

Petroleum (rare) 


NaCl (in cubes) 

Newhouse studied 15 samples of galena and 4 of sphalerite from lead and zinc deposits 
in Europe and North America. In all of these the abundant basic radicles in the 


qX. IB.*I 

Pale B/u6 

Honey -yellow Fluor i 

FIGURE 21. Inclusions of chalcopyrite, petroleum and marcasitc in fluorite 

Negative crystal faces of fluorite (A) and inclusions of chalcopyrite (B), petroleum (C), and marcasite (D) along 
transition plane from yellow to blue fluorite. Cleveland Mine, Spar Mountain, Illinois. X 5. After B as tin. 

fluids of the inclusions were sodium and calcium, and the acid radicle was chlorine. 
Calcium is decidedly subordinate to sodium. The concentrations are estimated at 
12-25 grams of NaCl per 100 cc. or 4 to 9 times the percentage of NaCl in sea water. 
Even these roughly quantitative determinations of the composition of the fluid in- 
clusions indicate that, if these are valid samples of the solutions from which the ores 
were deposited, their composition and high concentration rule out both meteoric and 
connate waters as agencies for the deposition of these lead and zinc ores. The alterna- 
tive is a magmatic source. 

Study of fluid inclusions seems to furnish a new method for determining deposition 
temperatures of certain ores. Newhouse (1933) in studying cleavage flakes and pol- 
ished thin sections of light-colored sphalerite noted in the fluid inclusions an ap- 
proximate constancy between the size of the inclusion and the size of the enclosed 
bubble. This suggested that the combined composition of gas and liquid was the same 
in all inclusions from any one specimen. Probably when the inclusions formed only 
one phase was present the gas was completely dissolved in the liquid. On the stage 
of the microscope, Newhouse heated plates of sphalerite from 13 districts and found 
that, for any one district, the gas bubbles all disappeared at very nearly the same 


temperature for the Joplin, Missouri, district, for example, at from 125-135C., 
and for Southeastern Missouri at 105-135C. Newhouse believes these represent the 
temperatures of ore deposition. They support a hydrothermal origin for the Missis- 
sippi Valley lead and zinc ores. In applying this method polished surfaces are essential 
because of the high index of refraction of sphalerite. 

An apparently contemporaneous association of fluid inclusions of petroleum and 
solid inclusions of marcasite and chalcopyrite are shown in Figure 21. They occur in 
the outer portion of a fluorite crystal at the contact of a layer of yellow fluorite with 
an overlying blue layer. They are approximately contemporaneous with the enclosing 

Primary and secondary inclusions must be carefully distinguished for only the 
primary inclusions give information on the composition or temperature of the miner- 
alizing solutions. If planes of fluid inclusions cross grain boundaries without effect or 
change of direction they are almost certainly secondary. 

Ingerson (1947) has pointed out that if there is reason to believe that the fluid 
inclusions being studied were formed under considerable pressures (as for example 
the fluid inclusions in the quartz of certain pegmatites) it is necessary to estimate 
these pressures and to apply a pressure correction before any reasonable estimate of 
the temperature of mineral deposition is possible. The composition of the liquid of the 
inclusion must also be considered. 



The writer and others (1931, p. 576-577) have summarized the nature and value of 
this class of evidence as follows: 

"The methods of reasoning employed to prove successive deposition may be conveniently classi- 
fied under two types: 

I. Indirect geological or geochemical evidence. 
II. Direct textural or structural evidence. 

"This division is not without overlap, however, since geologic concepts usually underlie even the 
textural and structural evidences. The first type of evidence rests upon the a priori postulate that 
for the formation of certain minerals and rocks markedly different environmental conditions (e.g., 
temperature, pressure, solution composition, etc.) are necessary, than for the formation of others. 
This postulate is amply justified by a great body of observational data on mineral and rock occur- 
rence, checked and supplemented by a lesser amount of laboratory data on mineral and rock synthesis. 
From such data we know definitely the conditions under which certain rocks have formed and it has 
also become possible to formulate a classification of minerals in accordance with their conditions of 
formation as first emphasized by Lindgren [1907] and later by W. H. Emmons [1908]. Many minerals 
may form through a considerable range of environmental conditions but it is clear that others form 
only through a comparatively narrow range. 

"When minerals or rocks that require for their formation notably different environments occur 
associated, the inference is justified that a change in environmental conditions ensued after one and 
before the other formed, and hence that they are of diverse ages. 

"An example of the use of this type of evidence in the case of ore minerals is afforded by associa- 
tions of galena and anglesite. Although galena is deposited in nature under a wide range of condi- 
tions, it is not known to form under oxidizing conditions such as characterize the zone above the 
ground water level. Anglesite, on the other hand, is know to form only under oxidizing conditions. 
Furthermore, anglesite is commonly a product of the oxidation of galena. The association of these 
two minerals implies, therefore, a shift to highly oxidizing conditions after the deposition of galena 
and before the deposition of the anglesite and therefore implies age diversity. Other familiar 
examples are the presence of garnet crystals in limestone and of calcite in the vesicles of basalt. 

"Evidences of this type when used alone depend entirely for their validity upon the correctness of 
the geological or geochemical postulates. Fortunately in most cases textural or structural evidences 
of age diversity are also present." 


If the contacts of two crystalline minerals are in the main sharp and straight they 
probably represent crystal faces of one mineral or the other. Two alternatives then 
exist: (1) one mineral may be an automorphic replacement of the other, or (2) the 
minerals may have been deposited in succession, the younger mineral conforming to 
the crystal outlines of the older. The replacement alternative should be avoided unless 
evidence of replacement is definite, for even automorphic replacement usually be- 
trays itself by spatial relations to fractures, by selecting certain minerals as host, or 
in other ways. In the absence of evidence of replacement, successive deposition must 
be postulated, and the question then is which mineral is the older. If mineral A shows 
crystal outlines characteristic of itself but not of mineral B, the problem is very 
simple. In Figure 1 of Plate 17 older pyrite (white) shows its characteristic cubical 
outlines against both quartz (dark) and chalcopyrite (gray). The quartz, in turn, 
shows its characteristic hexagonal outlines against chalcopyrite. 

In other cases the problem is less simple. In Figure 22, mineral A may not show 
certainly identifiable outlines but its general outlines are convex or protruding out- 
ward, and its crystal faces join to enclose an isolated grain of A in B. Very probably 



the crystal faces are those of mineral A rather than B and therefore mineral A ante- 
dates B. In Figure 23, however, crystal boundaries of A are in general concave out- 
ward, and probably these are crystal faces of mineral B rather than A. Examples of 
"concave" areas of a younger mineral in an older are shown in Figure 5 of Plate 16. 
Such areas although described for brevity as "concave" are bounded wholly or in 


FIGURE 22. Diagram illustrating significance of protruding crystal boundaries 
FIGURE 23. Diagram illustrating significance of depressed crystal boundaries 

In Figure 22 (left) the outlines of mineral A, which are "convex" or protruding toward mineral B, indicate that A is 
probably older than B. In Figure 23 (right) the outlines of mineral A, which are "concave" toward mineral B, indicate 
that its boundaries are crystal faces of B and that A is the younger mineral. 

part by crystal faces and must not be confused with the caries developed by replace- 
ment (Fig. 18). 


Even in the same field of the microscope mineral A may in places show its own 
characteristic crystal outlines against mineral B and a short distance away the rela- 
tions may be reversed. In most such ores there is evidence of only one generation of 
A or of B. Such cases usually indicate simultaneous crystallization of A and B. 
Such an interpretation is often supported by chemcial evidence, as for example in the 
case of certain intercrystallizations of cuprite and native copper formed by the oxida- 
tion of chalcocite according to some such reaction as 3 Cu 2 S + 4 Oa 2 Cu + 
2 Cu 2 O + 3 SO 2 . 


Aside from its obvious control in the formation of sedimentary ore deposits gravity 
may be a factor in the deposition of other ore types. In some ore deposits the last 
minerals to be precipitated in vugs or other cavities were clearly deposited under the 
influence of gravity. Stalactites in the zone of oxidation are the most familiar examples 
of such deposits. In the fluorite deposits of southern Illinois layers of calcite or of 
barite crystals coat the upper surfaces of fluorite crystals in vugs but do not coat their 
flanks or downward facing surfaces (PL 16, fig. 6). The phenomenon is comparable to 



the heaping up of snow on the roof of a house and its absence from the walls and 
under the eaves. Such deposits are obviously younger than the minerals which they 
coat and which show no relationship to the direction of the force of gravity. 


The term simultaneous deposition has been used by geologists in somewhat diverse 
senses as indicated by such terms as "partial simultaneity", "essentially simultane- 


/***// JS 


Aftntrol / 


Mintrai / 

Successive Deposition 

FIGURE 24. Paragenetic diagram illustrating preferred usage of the terms Simultaneity ; 
Overlap, and Successive Deposition 

ous" and "complete simultaneity." In the interest of clarity the term is here used hi 
the description of ores in its narrow sense; substances are considered to be simultaneous 
only if their precipitation from solution began concurrently and ended concurrently. 
Under natural conditions simultaneous deposition, as so defined, is probably com- 
paratively rare. Overlap is partial simultaneity, and evidence of overlap consists, in 
the last analysis, of evidence of simultaneous deposition in one part of the specimen 
and of successive deposition in an adjacent part. Any overlap should preferably be 
described under the caption of "overlap" rather than "partial simultaneity." It is a 
common relationship in contrast to the rarity of simultaneous deposition. 

The age relations implied by the terms (a) simultaneity, (b) overlap, and (c) 
successive deposition, as used in this volume, are expressed diagrammatically (Fig. 24). 



In the simple form of paragenetic diagram (Fig. 25) the minerals of the ore are 
shown from top to bottom in the order in which their deposition appears to have 
begun-, minerals that are simultaneous are obviously interchangeable in position. 
The width of the diagram is adjusted to the width of the printed page, and therefore 
in some diagrams involving many minerals the lines will be short while if few min- 








FIGURE 25. Paragenetic diagram, EL Bate Mine, Zacatecas, Mexico 
Ore from 800-foot level. After B as tin. 

Ca and Mg 







FIGURE 26. Composite paragenetic diagram for the Zacatecas district, Mexico 
In terms of the radicals, basic and acid, that make up the minerals of the ores. After Bastin 

ends are involved they may be longer. Only the relative positions of the lines are 

In the interpretation of the microscopic textures of ores it is good practice to pre- 
pare a diagram of this sort for each polished specimen studied. From a series of such 
diagrams a composite paragenetic diagram for the ore body as a whole can be con- 
structed for seldom are all significant mineral relations displayed in a single polished 
specimen. Comparison of many diagrams may bring out similarities and significant 
variations in the order of deposition. 


Where mineralization has been fairly simple and uniform it may be possible to con- 
struct a composite paragenetic diagram for the whole district. Obviously many de- 
tails are lost in such a graphic generalization, which represents only the more com- 
mon relations for the district. 

As an aid to understanding the chemistry of the mineralizing processes it is often 
helpful to translate a paragenetic diagram from terms of minerals into terms of the 
radicles which make up the minerals (Fig 26). 

As there will be no overlap between (1) hypogene minerals, (2) minerals of the zone 
of downward enrichment, and (3) minerals of the oxidized zone, one can show in one 
composite diagram for an entire ore body or an entire district the paragenesis for all 
three depth zones, each zone being appropriately labelled. 

Some geologists introduce the period of maximum deposition into the paragenetic 
diagram by substituting for all or part of the straight line a bulbous enlargement 
whose widest part represents the period of maximum deposition. So much uncer- 
tainty enters into such attempts that they seem to be of doubtful value. 



Sedimentary ore deposits characteristically are associated with sediments of the 
more usual types commonly as beds or parts of beds; they usually show such charac- 
ters as bedding, cross-bedding, fossils, ripple marks, etc. These features are familiar 
to all students in geology and are not described here. In using such features as evi- 
dences of sedimentary origin one must remember that they are sometimes preserved 
in ores formed by hydrothermal replacement of sediments particularly limestones 
(PL 8, fig. 3). 

Genetically the sedimentary ores may be mechanical deposits, as placers, or 
chemical deposits as many iron and manganese ores; biochemical processes some- 
times play a role. To be properly classed as sedimentary they must have been syn- 
genetic deposits in at least their early stages, but their metal content may have under- 
gone subsequent redistribution by circulating solutions as well illustrated by the 
sedimentary "Red Beds" type of copper deposits in the Southwestern United States. 

The oolitic textures common in many iron ores are described because they throw 
much light on the origin of such ores more than has been generally recognized. 
Certain ore textures attributed to bacteria and algae are also briefly considered. 


Oolitic limestones afford the most familiar examples of oolitic textures and in 
early studies, oolitic iron ores were often interpreted as ferruginous replacements of 
odlitic limestones. As the oolitic iron ores have become better known it has become 
clear that the oolitic texture is an original character. Characteristic features of 
oolitic limestones have been described by Bucher (1918) and E. Cloos (1947). The 
present discussion is limited to the characteristics of oolitic iron ores. 

Oolitic textures are fairly common among the bedded iron ores, particularly those 
composed dominantly of limonite or hematite. In most instances the typical well- 
rounded forms were developed on the sea bottom (rarely lake bottom) by aggregation 
of fine ferruginous material around nuclei under gentle wave agitation. In a few in- 
stances the forms have been modified by dynanometamorphism, or replacement has 
altered the original forms and composition. 

The best illustrations and descriptions of the textures of the oolitic iron ores are 
those of Cayeux (1909; 1922) in his monumental volumes on the iron ores of France. 
The textural features of the French ores will therefore be described as a standard of 
comparison for those of other regions. From them much may be learned of the condi- 
tions of genesis of the ores. 


Oolites Defined. Cayeux defines oolites as small spherical or ovoid bodies of mineral 
matter of concentric structure and usually consisting of a core or nucleus enclosed by 



a cortical envelope made up of concentric layers (PL 17, fig. 3). Bodies of similar size 
and shape often associated with oolites but lacking concentric structure Cayeux terms 
false oolites. The latter are mineral fragments or fragments of the hard parts of organ- 
isms that have been rounded by wave action but have not been appreciably enlarged 
by accretion (PL 17, fig. 4). Oolites are usually simple, composed of a single nucleus 
and its envelope. Rarely they are composite, two, three or even six simple oolites 
being surrounded by a common envelope. In some deposits composite oolites are 
absent, in most they are rare; in a few nearly all are composite, 

Size of Oolites. Most oolites are 0.1-1.0 mm in maximum diameter; rarely a com- 
posite oolite or even a simple one may reach 2 mm. 

Packing of Oolites. Oolites may be so closely packed that there is almost no matrix 
(PL 17, fig. 3) or very few oolites may be in contact (PL 17, fig. 5). Usually matrix 
is subordinate to oolites in volume. 

Materials of the Oolitic Iron Ores. The same materials are found in oolites and 
matrix, but they are in different proportions, and their history has been different. 
They include siderite, chlorite, hematite, limonite, glauconite, magnetite and pyrite, 
and, as nonferruginous materials, silica, calcite, phosphate of lime, and a few others. 
Detrital minerals other than quartz and siderite are absent from most of the ores and 
are rare in others. The hard parts of organisms are often abundant. 

The nucleus of the oolite may be angular or rounded. It is sometimes ferruginous 
material which may or may not be similar to that of the envelope. In other cases it is a 
foreign body such as a quartz or calcite grain or a fragment of bone or shell or a frag- 
ment of an older oolite. Only exceptionally is it impossible to differentiate nucleus 
and envelope. The size of the nucleus is extremely variable with respect to the entire 
oolite (PL 17, figs. 3, 6). 

The envelope may be composted of one or more of the minerals siderite, chlorite, 
hematite, or limonite. Very rarely it may contain magnetite or pyrite. The concentric 
structure may be extremely fine and delicate or coarse and vague. It is usually most 
delicate in the chloritic oolites (PL 17, fig. 3). The iron oxides tend to mask the con- 
centric structure. 

The mineralogy of the oolitic iron ores of any one district is usually simpler than 
might be inferred from the comprehensive list of components given above for certain 
components are dominant and others rare or absent. For the French oolitic ores 
hematite is generally the dominant iron mineral in the Paleozoic ores and limonite 
in the later ores including the great deposits of Lorraine. Siderite, though subordinate 
to limonite, is of much importance in the ores of Lorraine (mainly as clastic grains) 
and not important in other French oolitic ores. In general carbonate-rich ores are 
also chloritic. Magnetite is of fortuitous occurrence in the Lorraine ores, and pyritic 
oolites are merely curiosities in the Oxfordian of Normandy. 

Solidity of Oolites. Cayeux's illustrations show that where the ferruginous oolites 
are hi contact each oolite maintains its smooth regularity of outline and is tangent to 
other ofllites without indenting them. Evidence is lacking of mutual accommodations 
of form such as would be expected in plastic materials as for example such accom- 
modations as may be observed between biscuits in a pan. Therefore the oolites were 


rigidly solid when deposited in contrast to the opinions of some writers (Bucher, 
1918) that the oolitic envelopes were plastic, gelatinous, and colloidal. Much of the 
material that fed the growth of the oolites was probably very fine-grained but it 
seems to have been aggregated into solid oolites. 

In a very few ores shallow cup-shaped depressions in the surface of one oolite may 
mark its places of contact with other oolites. This is clearly due to solution at points 
of contact subsequent to deposition and confirms the solid nature of the oolites since 
they have undergone no general deformation. 

False oolites also testify to the mode of origin of the true oolites. The false oolites 
as shown at A and B in Figure 4 of Plate 17 are fragments of fossils or of minerals that 
have unquestionably been rounded mechanically, presumably by wave and current 
action. They lack the envelope of concentric layers that characterize the true oolites. 
They are, however, frequently associated with the true oolites (PI. 17, fig. 5) and are 
similar to them in size and shape. Only in thin section are they readily distinguished. 
Indeed the distinction is purely one of definition for a false oolite may be regarded as 
an oolite composed wholly of a fossil or a mineral fragment. Oolites in a particular 
deposit tend apparently to build up by accretion to a certain order of size, while the 
larger shells and mineral fragments are worn down to the same order of size. Both are 
accomplished by wave action. There are transition forms oolites with very thin 
concentric envelopes (C in PL 17, fig. 5). 

False Oolites. Grains of various materials associated with the oolites and worn by 
wave action to the size and shape of oolites but lacking the concentric layers Cayeux 
terms "false oolites." In the formation of false oolites abrasion has been the dominant 
process. The distinction is no doubt justified as otherwise there would be no distinc- 
tion between an oolitic deposit and, for example, a sanstone of rounded grains. In 
quantity false oolites are usually very subordinate to the oolites, but some Devonian 
iron ores contain only false oolites. 

Most false oolites in iron ores show the organic structures of mollusks or of crinoids 
but may be partly or wholly ferruginous; there is every gradation from nonmineral- 
ized to completely mineralized false oolites. First there is usually a selective replace- 
ment of certain parts of the cellular material which tends to emphasize its structure; 
then there is a mass replacement penetrating inward from the periphery of the 
false oolite, that completely destroys the organic structure as shown at A and D in 
Figure 4 of Plate 17. If the false oolite were partly or wholly coarse calcite it maybe 
partly replaced by hematite with the development of very ragged boundaries as 
shown at F in Figure 4 of Plate 17. Finally the entire false oolite may be converted 
to hematite, the process usually being completed first in the smaller false oolites as 
at G. Usually in such ores the original surface outlines of a shell or other organic 
material has been destroyed or greatly modified by abrasion due to wave action. 
Figure 5 of Plate 17 pictures on a smaller scale oolitic ore from another French iron 
ore deposit. It shows some of the same features exhibited in Figure 4 of Plate 17. 
There are remnants of crinoids in the interior of some of the oolites and false oolites 
at (A and C). Unlike Figure 4 of Plate 17, however, are abundant crinoidal remains in 
the matrix that have been partly replaced by hematite. 


In Figure 6 of Plate 17 the oolites are nearly wholly hematite so dense that no con- 
centric structure is visible. Oolites with nuclei of irregular quartz grains are shown at 
A. The cement is siderite. 

Organisms. The hard parts of organisms found in the iron ores of France include 
fish remains, mollusks, crinoids, and Foraminifera. Others occur but only subordi- 
nately, Fish remains are present in all the ores but are particularly abundant in 
those of Lorraine. Mollusks and crinoids are in general the most abundant organisms. 
Foraminifera though widespread are in general rare and are mostly sea-bottom forms. 
All these fossils have been replaced in diverse degrees by various iron-bearing min- 
erals and by silica. Boring algae represented by tubes and casts are not known to have 
been important in the formation of the ores. 

Matrix or Cement. The matrix is usually subordinate in volume to the oolites. Its 
materials are in general the same as those composing the oolites but in different 
proportions. Calcite, often rhombohedral, is frequently abundant instead of rare as 
in the oolites. Clay also is a component. Cements composed largely of organic debris 
are extremely rare. 

Conditions of Deposition. The smoothly rounded outlines of the oolites and false 
oolites clearly indicate prolonged agitation by wave action. The nature of the fossils 
testifies to marine waters. Wave action was severe enough to comminute the remains 
of mollusks and crinoids and occasionally to break oolites, the fragments becoming 
the nuclei of new oolites. These features as well as evidences of contemporaneous 
erosion point to deposition in moderately shallow seas, and the great lateral extent 
of many of the deposits shows that these seas were extensive. 

Changes Subsequent to Deposition. Cayeux figures (Vol. II, Fig. 24) an ore from 
Lorraine in which wholly chloritic oolites have been deformed by unilateral pressure 
into irregularly spindle-like forms while the mechanically more competent grains of 
clastic siderite and phosphate of lime have not been distorted. Such pressure phe- 
nomena are, however, very rare among the French oolitic iron ores, whereas deforma- 
tion of oolites in limestones is fairly common as described by E. Cloos (1947). 

Replacement subsequent to the emplacement of the oolites has produced minor 
changes in some of the ores of Lorraine. In Figure 6 of Plate 17 secondary chlorite 
forms thin layers between the hematite of the oolites and the siderite of the matrix. 
Many of the hematite oolites are in contact, and at such contacts no chlorite is pres- 
ent. The chlorite therefore is not a part of the oolites but has been developed by re- 
placement of the siderite of the matrix subsequent to the deposition of hematite 
oolites in their present position. 

In a Silurian oolitic ore the originally smooth outlines of chloritic oolites have been 
rendered irregular through replacement by siderite beginning at the original periphery 
of the oolites and penetrating inward to varying depths (PL 18, fig. 1). The irregular 
inner border of the siderite hi places cuts across the banding of the chlorite layers, and 
therefore it formed after the banding was developed. Another picture of the same ore 
confirms the originally chloritic nature of the oolite envelope, for where oolites are in 
contact there is a smooth contact of chlorite with chlorite with no intervening siderite. 

The originally smooth outlines of oolites mav also become irregular through 


secondary enlargement. Hematite may develop in the matrix in continuity with that 
of the oolite and destroy its regularity of outline. 


The most important iron ores of England, the Jurassic ores of the Cleveland Hills 
in North Yorkshire, have been comprehensively described by Hallimond (1925) and 
are characteristically oolitic. Evidence is decisive that they were deposited essentially 
in their present condition. They differ from most other oolitic iron ores in that the 
iron is largely in the ferrous state as chamosite (a ferrous aluminous silicate) and 
siderite] the ferric oxides are rare or absent except as products of recent weathering. 

In the Cleveland ores the oolites average about one-sixtieth of an inch in diameter 
and are composed of chamosite and siderite. The chamosite is very fine-grained and 
is concentrically banded, but the siderite is granular and has usually replaced the 
chamosite beginning at the smooth periphery of the oolite. The inner border of the 
siderite zone is irregular. Some oolites are wholly chamosite, but oolites wholly sider- 
ite are rare. Some oolites have been replaced by calcite showing parallel orientation 
throughout a group of contiguous oolites. Many oolites have no recognizable nucleus, 
but others contain nuclei of shell fragments or broken oolites or crystals of chamosite. 
The matrix consists of siderite rhombs in a groundmass of fine chamosite. 

Concerning the British oolitic iron ores as a whole Hallimond says: (p. 10) 

"The odliths themselves are, in grading and distribution, practically equivalent to sand-grains. 
They are frequently polished and accompany polished sands. A striking feature is their uniform grad- 
ing, examples of small and incompletely developed oolites being very rare, and this fact seems most 
consistent with the view that growth in the later stages was extremely slow and tended to reach a 
limit when the increased weight of the particle resulted in a degree of mechanical erosion that bal- 
anced the chemical precipitation." 


Among the best examples of oolitic iron ores on this continent are the Ordovician 
ores of Wabana, Newfoundland (Hayes, 1915). The ores are principally hematite and 
chamosite an aluminous ferrous silicate. Siderite, abundant locally, is mainly con- 
fined to the matrix. The hematite is very finely divided and occurs principally in the 
oolites and subordinately in the matrix. In the oolites it is concentrically banded and 
is invariably associated with chamosite. When abundant it may mask the chamosite, 
but if the hematite is dissolved by acid the chamosite is always present. Usually the 
two minerals form alternate layers. In the leaner ores the envelopes of some oolites 
may be wholly concentric concentric layers of chamosite. In general chamosite is 
next to hematite in abundance. 

Most of the oolites have a nucleus which may be an angular grain of quartz (shown 
to be detrital by its fluid inclusions and its apatite prisms), shell fragments, or chamo- 
site or siderite. The matrix is predominantly siderite but may contain also hematite 
and chamosite as well as shell fragments. Siderite sometimes has replaced hematite 
and chamosite, but the reverse relation was never observed. The shell fragments are 
largely inarticulate brachiopods and hence are phosphatic. Fossil tubules of boring 
algae occur in oolites and in shell fragments. 

Ripple-marked surfaces in all the workable ore beds and the occasional presence of 


cross-bedding, indicate shallow water deposition, and graptolites and brachiopods 
indicate that the waters were marine. The common alternation of chamosite and 
hematite layers in the envelopes indicates that the two minerals are in the main con- 
temporaneous though in places hematite has clearly replaced chamosite. Such re- 
placements were diagenetic, taking place while the sediments were still unconsoli- 
dated. No important amounts of iron seem to have been added since sedimentation 

Oolitic Pyrite Deposits. Minor deposits of pyrite in the Wabana district, although of 
no economic importance are one of the few described occurrences of oolitic pyrite. 
Oolitic pyrite (PL 18, fig. 3) occurs in one to three beds lying 1 to 10 feet stratigraphi- 
cally above the highest hematite ores. There are no gradations from pyritic to hema- 
titic ores, and the two resemble each other only in texture. In some of the oolites layers 
of pyrite alternate with what appears to be calcium phosphate. Fragments of brachio- 
pods have been replaced to varying degrees by the pyrite. In contrast to the hematite 
ores graptolite remains, some replaced by pyrite, are common, indicating access to the 
open sea and its pelagic faunas. Unpyritized fossil fagments are often in contact with 
completely pyritized fragments indicating that these constituents were brought to- 
gether mechanically on the sea bottom after the formation of the pyrite. 

The pyritic beds alternate with fissile black graptolite-bearing shales. No pyrite 
masses transgress the bedding. There is no evidence that the pyrite beds are replace- 
ments of hematite-chamosite ores; Hayes considers that they are as truly original 
sediments as are the oolitic hematites, but deposited in deeper water where the de- 
velopment of hydrogen sulphide by organic reactions on the sea bottom led to the 
formation of iron sulphide and produced acid conditions inhibiting development of 


Newland and Hartnagel (1908) have described the oolitic iron ores of the Clinton 
formation of New York, which are dominantly earthy hematite and are in part 
oolitic. The oolitic grams are spherical or somewhat flattened and seldom over 1 mm 
in diameter. There is usually a nucleus commonly a quartz grain and an envelope 
of concentric hematitic layers. If the hematite of the oolites is dissolved by HC1 a 
gelatinous mass of transparent silica remains having the form of the original oolite 
The matrix is usually granular calcite. The oolites are usually closely packed. The 
rounded quartz nuclei, rarely over half the diameter of the oolite, contain fluid inclu- 
sions and small crystals of rutile and hematite. Their source is probably the Pre- 
cambrian crystalline rocks. In some of the Clinton ores the hematite is mainly a re- 
placement of the hard parts of bryozoans, crinoids and brachiopods and false oolites 
may be more abundant than oolites (PL 18, fig. 2). 

Some of the Clinton ores are directly overlain by limestone which shows no evi- 
dence of replacement by hematite and is not oolitic. Newland and Hartnagel interpret 
the ores as original sea-bottom deposits. 

The most detailed studies of the microscopic characters of the Clinton iron ores of 
New York are those of Ailing (1947). The Clinton ores are lenticular and not confined 
to a single horizon within the formation. The textures are characteristically oolitic; 



in the western part of the ore zone the nuclei are almost exclusively fossil fragments; 
in the eastern part nuclei of rounded quartz grains become predominant. The micro- 
scope shows that the fossil fragments were rounded, apparently by wave action, and 
their small openings were rilled by calcite, clay, and phosphatic material. Later the 

FIGUEZ 27. Automorphic replacement of iron silicate oVlites by siderite 

From new Priston, Kentucky. After Bucher. These relations are interpreted by Basttn as indicating automorphic 
replacement of solid iron silicate oolites by siderite as indicated by the transecting of the iron silicate laminae by the 
siderite crystals. 

fossiliferous material was replaced hi part by hematite and coated by concentric 
layers of chamosite-hematite often several hundred in number. All these processes 
are believed to have taken place on the sea bottom before the consolidation of the 
deposit and therefore were diagenetic changes. The source of the iron still remains 

In the Birmingham district of Alabama as in New York both oolitic ores and ores 
that are largely replacements of fossil fragments occur (Burchard, 1910). The oolitic 
ores are aggregates of somewhat flattened grains, A to rV inch in diameter, lying 
with their flatter dimensions parallel to the bedding. The oolites usually consist of a 
nucleus of rounded quartz and an envelope of successive layers of hematite associated 
with some siliceous and aluminous material. The matrix consists of hematite with 
some calcite. The so-called "fossil ores" of Birmingham consist of the fossil remains 
of bryozoans, crinoids, corals, and brachiopods many of which are broken and water 
worn. Usually either fossil or oolitic ore predominates in a bed, but in places they are 
mixed in nearly equal proportions. 

In a few of the American oolitic ores replacements subsequent to deposition appear 
to have modified the ores. In an oolitic iron ore from Kentucky (Bucher, 1918, p. 
598-601), in which oolites of iron silicate lie in a siderite matrix (Fig. 27), automorphic 
crystals of siderite project into the oolites. Bucher interprets this as evidence that the 
oolites were gelatinous and yielding, but the laminae of the oolites are not deformed 


but are sharply transected by the siderite crystals probably indicating automorphic 
replacement of solid iron silicate. 


The causes of the deposition of the great sedimentary deposits of iron ores that con- 
stitute the World's greatest reserves have been a persistent and vexing problem that 
is made more difficult because comparable ores seemingly are not being deposited 
today. Because some living algae and Bacteria can deposit hydrous iron oxides from 
ferruginous solutions (Harder, 1919) there has been much interest in the occasional 
occurrence of seemingly fossil algae and bacteria in or closely associated with some 
sedimentary iron ores. Conclusive evidence is still lacking that they played a major 
role in ore deposition. 

Cayeux (1909, Figs. 33-35) figures excellent examples of boring algae in both hema- 
tite oolites and matrix in a Silurian iron ore. Hayes (1915, PL XXI-B) describes and 
figures the tubules of boring algae from the oolitic hematites of Wabana, Newfound- 

Gruner (1924) has described and figured what appear to be bacterial and alga- 
forms in cherts from beds stratigraphically below the Huronian Biwabik iron-bearing 
formation of the Mesabi range in Minnesota (PL 17, fig. 2) 


Regional metamorphism is unfavorable for the formation of ore deposits but has in 
certain instances deformed pre-existing deposits and developed pronounced foliation 
in the ore. The changes produced are largely textural and mineralogic ; there is little 
change in gross chemical composition, and transfer of material has taken place only 
through very limited distances. Ores deformed by dynamometamorphism may look 
much like ores that have been formed by the replacement of metamorphic rocks, but 
in the latter only the wall rocks and not the ore minerals show evidence of deforma- 

In Europe the ores of Rammelsberg are an excellent example of regionally meta- 
morphosed ore deposits. These ore bodies are enclosed in slates of Devonian age that 
have undergone considerable distortion. The ores consist, in approximate order of 
abundance, of sphalerite, chalcopyrite, galena, pyrite, and arsenopyrite with a barite 
gangue, and in places are much contorted. (Fig. 28). Lindgren and Irving (1911) 
have shown that the ores have been regionally deformed with the development of a 
gneissoid structure. Pyrite and barite the components most resistant to deforma- 
tion form lenticles or nodules around which wrap bands of the weaker minerals 
galena, sphalerite, and chalcopyrite. Figure 4 of Plate 18 shows nodules of a relatively 
resistant coarse intergrowth of barite and sphalerite around which the weaker 
metallic minerals are moulded as thin laminae. 

Among the common ore minerals galena is particularly susceptible to deformation. 
In the initial stages such deformation may manifest itself in hand specimens by curved 
cleavage faces, and under the microscope in polished specimens by curvature in the 
lines of the characteristic triangular pits. More intense deformation reduces the 
galena to an aggregate of minute grains and develops planes of flowage forming a 
schistose mass commonly termed "steel" galena (PI. 18, fig. 5). Where more resistant 
minerals are associated with the galena the latter wraps around them in the fashion 
similar to that shown in Figure 4 of Plate 18. 

In some cases the harder minerals of the ore are broken by the differential pressure, 
and the fragments are dragged apart in the flowage of the steel galena. 

Movements within the ore minerals under differential pressures are not random 
adjustments but have been shown (Buerger, 1928) to be in part due to slipping or 
translation along definite crystallographic planes and in part to reorientation tending 
to bring certain crystallographic directions into parallelism with the direction of 
maximum pressure. Galena was found particularly susceptible to deformation, and 
pyrite particularly resistant. Sphalerite, chalcopyrite, and pyrrhotite were inter- 

In experiments on the deformation of ores and ore minerals under pressure (Buer- 
ger, 1928; Newhouse and Flaherty, 1930; Edwards, 1947, Fig. 33) certain metallic 
minerals that seldom show original or primary twinning may develop conspicuous 
secondary twinning under very mild external pressures that are insufficient to shear, 
brecciate, or granulate them. Sphalerite and chalcopyrite not uncommonly exhibit 
such twinning. This twinning must be used with caution as evidence of stresses of 




external origin since it may also develop as a result of mild internal stresses such as 
those incident to the unmixing of solid solutions (Buerger and Buerger, 1934). Etch- 
ing with appropriate reagents and/or examination in reflected polarized light is 
usually necessary to recognize such secondary twinning. 

FIGURE 28. Contorted banding in ore 
From Rammelsberg, Germany, a pyrite, b - chalcopyrite, c - galena. X 1. After Wiechelt. 

Finally, certain textures exhibited by some of the native metals some original 
textures and some the result of mild degrees of thermal or of dynamic metamorph- 
ism have been investigated by British metallurgists (Carpenter andTamura, 1928; 
Carpenter and Fisher, 1930; 1932) in a valuable study applying to the naturally oc- 
curring metals the methods used in the study of manufactured metals and alloys. 
Such studies may be illustrated by those on silver which are of especial geologic 

Native silver recrystallizes so readily under even mild pressures that in the grinding 
of specimens preparatory to polishing for microscopic study the surface layers un- 
dergo a distortion which masks the true texture of the metal beneath. By repeated 
etching this surface layer can be removed, and further gentle polishing combined 
with etching reveals the original texture. Twenty-one specimens of native silver from 
14 localities widely scattered over the world but mostly from Canada and the United 


States were prepared and studied under the reflecting microscope and by heat treat- 
ment (annealing). Some of the specimens showed a homogeneous granukr texture 
(PL 18, fig. 7) identical with that developed artificially in silver by annealing above 
200C. Silver from other localities had been partially recrystallized, and still other 
silver (PL 18, fig. 6) showed no recrystallization but was characterized by zonal 
structures. The silver that on geological evidence appeared to be supergene showed 
evidence of only partial recrystallization or none at all. In contrast, the silver from 
Cobalt, Ontario, and neighboring camps of the same type, now generally interpreted as 
hypogene, was all of the completely recrystallized type. Carpenter and Fisher (1932) 
conclude that this silver recrystallized in the presence of hot hypogene solutions above 
the normal atmospheric recrystallization temperature of silver (about 200C.). 

While temperature may well have been the dominant control for these contrasting 
textures, both time and pressure may have been collaborating factors, inasmuch as 
supergene native silver is invariably of shallow and recent origin, whereas the silver 
occurrences of Cobalt and neighboring camps are Precambrian and of more deep- 
seated origin. 


Ailing, H. L. (1947) Diagenesis of the Clinton hematite ores of New York, Geol. Soc. Am., Bull, vol. 

58, p. 991-1018. 
Anderson, G. E., and Merritt, C. A. (1937) The volume relations in open space replacements, Am. 

Assoc. Petrol. Geol., Bull., vol. 21, p. 1486-1493. 
Bastin, Edson S. (1915) Experiments with colloidal gold and silver, Washington Acad. Sci., Journ., 

vol. 5, p. 64-71. 
(1922) Primary native silver ores near Wickenburg, Arizona, and their bearing on the genesis 

of the silver ores of Cobalt, Ontario, U. S. Geol. Survey, Bull. 735, p. 131-155. 

(1925) Primary native silver ores of South Lorraine and Cobalt, Ontario, Econ. Geol., vol. 

20, p. 1-24. 

(1931) The fluorspar deposits of Hardin and Pope counties, Illinois, 111. Geol. Survey, Bull. 

58, 116 pages. 

(1937) In The geology and biology of the San Carlos Mountains, Tamaulipas, Mexico, 

Univ. Mich. Stud., Sci. Ser., vol. 12, p. 175-176. 

and Laney, F. B. (1918) The genesis of ores at Tonapah, Nevada, U. S. Geol. Survey, Prof. 

Paper 104, 50 pages. 
et al. (1931) Criteria of age relations of minerals with especial reference to polished sections 

of ores, Econ. Geol., vol. 26, p. 561-610. 

Beck, Richard (1909) Lehre von den Erzlagerstatten, Bd. 1, p. 248. GebrUder Borntraeger, Berlin. 
Boy dell, H. C. (1924-1925) The role of colloidal solutions in the formation of mineral deposits, Inst. 

Min. Metall., Tr., vol. 34, p. 145-337. 

Bucher, W. H. (1918) On oolites and spherulites, Journ. Geol., vol. 26, p. 593-609. 
Buerger, M. J. (1928) The plastic deformation of ore minerals, Am. Mineral., vol. 13, p. 1-17, 35-51. 
(1932) The negative crystal cavities of certain galena and their brine content, Am. Mineral., 

vol. 17, p. 228-233. 
Buerger, N. W. (1934) The unmixing of chalcopyrite from sphalerite, Am. Mineral, vol. 19, p. 525- 

and Buerger, M. J. (1934) Crystallo graphic relations between cubanite segregation plates, 

chalcopyrite matrix and secondary chalcopyrite twins, Am. Mineral., vol. 19, p. 289-303. 
Burchard, E. F., Butts, Charles, and Eckel, E. C. (1910) Iron ores, fuels and fluxes of the Bir- 
mingham district, Alabama, U. S. Geol. Survey, Bull. 400, 204 pages. 

Burton, E. F. (1938) The physical properties of colloidal solutions, 3d ed. Longmans, Green and Co. 
Carpenter, H. C., and Fisher, M. S. (1930) A study of the crystal structures of native copper, Inst. 

Min. Metall., Tr., vol. 39, p. 238-284. 
(1932) A metattographic investigation of native silver, Inst. Min. Metall., Tr., 

vol. 41, p. 382-433. 
and Tamura, S. (1928) The inner crystal structure of some native metals, Inst. Min. 

Metall., Tr., vol. 37, p. 365-386. 
Cayeux, L. (1909 and 1922) Les memoirs de fer odlithiquc de France, vols. I and II, Paris. Impri- 

merie nationale. 
Clark, J. D., and Menaul, P. L. (1916) The role of colloidal migration in ore deposits, Econ Geol., 

vol. 11, p. 37-41. 
Cloos, E. (1947) Odlite deformation in the South Mountain fold, Maryland, Geol. Soc. Am., Bull., 

vol. 58, p. 843-917. 
Donath, M. (1930) Geologish-mineralogische Studien an serbischen Chromiuagerstdtten, Ernst Mauc- 

kisch, Freiberg i Sa, p. 1-61. 
Edwards, A. B. (1947) Textures of the ore minerals and their significance, Australian Inst. Min. 

Metall., Melbourne. 185 pages. 
Ehrenberg, H. (1931) Der Aufbau der Schalenblenden der Aachener Belizinkerglagerst&aten und der 

Einflussihres Eisengehaltes aufdie Miner albildung, Neues Jahr. Mineral., Bd. 64, p. 397-422. 
Emmons, S. F., Irving, J. D., and Loughlin, G. F. (1927) Geology and ore deposits of the Leadville 

Mining District, Colorado, U. S. Geol. Survey, Prof. Paper 148, 368 pages. 



Bmmons, W. H. (1908) A genetic classification of minerals, Econ. Geol., vol. 3, p. 611-627. 
(1940) Principles of economic geology, McGraw-Hill Book Co., New York, 2d cd., 520 

Gavelin, Sven (1939) Geology and ores of the Malanas district, Sweden, Sveriges Geol. Unders. s. C. 

no. 424, 224 pages. 
Geijer, Per (1910) Igneous rocks and iron ores of Kiirunavaara, Luossavaara and Tuollavaara, P. A. 

Norstedt and Saner, Stockholm, Sweden. 
Grondijs, H. F., and Schouten, C. (1937) A study of the Mt. Isa ores, Econ. Geol., 

vol. 32, p. 415-421. 

Grout, F. F. (1932) Petrography and petrology, McGraw-Hill Book Co., New York, 522 pages. 
Gruner, J. W. (1924) Contributions to the geology of the Mesabi range, Minn. Geol. Survey, Bull. 19, 

71 pages. 
Guild, F. N. (1917) A microscopic study of the silver ores and their associated minerals, Econ. Geol., 

vol. XII p. 297-353. 

Hallimond, A. F. (1925) Iron ores: Bedded ores of England and Wales, Geol. Soc. Britain, Mem. 29. 
Harder, E. C. (1919) Iron depositing bacteria and their geologic relations, U. S. Geol. Survey, Prof. 

Paper 113, 89 pages. 

Harker, Alfred (1909) The natural history of igneous rocks, The Macmillan Co., New York. 
Hayes, A. O. (1915) The Wabana iron ores of Newfoundland, Canada Geol. Survey, Mem. 78, 163 


Hifitze, Carl (1915) Handbuch der Mineralogie, Veit and Co., Leipzig, Bd. la, p. 1535-1538. 
Ingerson, Earl (1947) Liquid inclusions in geologic thermometry, Am. Mineral., vol. 32, p. 375-388. 
Johannsen, Albert (1931) A descriptive petrography of igneous rocks, Univ. Chicago Press, vol. 1, 

267 pages. 
Johnston, W. D., Jr. (1936) Nodular, orbicular and banded chromite in northern California, Econ. 

Geol., vol. 31, p. 417-^27. 
and de Souza, H. C. A. (1943) Chromite at Campo Formoso, Baia, Brazil, Econ Geol., 

vol. 38, p. 287-297. 
Kato, Takeo (1921) Notes on the banded chromite ore from the Akaiski Mine in the Province of lyo, 

Japan, Geol. Soc. Tokyo, Jour., vol. 28, no. 331, p. 16-17. 
Keil, Karl (1931) Beitrage zur Kenntniss der Kobalt-Nickel-Wismut-Sttber Erzgange, Jahrb. fur Berg 

Huttenwesen in Sachsen. vol. 105, p. 95-132. 
Kidd, D. F., and Haycock, M. H. (1935) Mineragraphy of the ores of Great Bear Lake, Geol. Soc. 

Am., Bull., vol. 46, p. 879-960. 
Krieger, P. (1935) Primary native silver ores at Batopelas, Mexico and Bullard's Peak, New Mexico, 

Am. Mineral., vol. 20, p. 715-723. 

Levings, J. V. (1912) Inst. Min. MetalL, Tr., vol. 21, p. 478. 
Liesegang, R. E. (1913) Geologische Diffusioncn, Rev. by Adolph Knopf, Econ. Geol., vol. 8, p. 

Lindgren, Waldemar (1907) The relation of ore deposition to physical conditions, Econ. Geol., vol. 2, 

p. 105-127. 
(1915) Processes of mineralization and enrichment in the Tintic district, Econ. Geol., vol. 

10, p. 225-240. 
(1924) The colloidal chemistry of minerals and ore deposits, in The theory and application of 

colloidal behavior, R. H. Bogue, Ed., McGraw-Hill Book Co., vol. 2, Chap. 8, p. 445-465. 

(1925) Gel replacement, a new aspect of metasomatism, Nat. Acad. Sci., Pr., vol. 11, p. 5-11. 

(1926) Ore deposits of the Jerome and Bradshaw Mountains quadrangles, Arizona, U. S. Geol. 

Survey, Bull. 782, 192 pages. 
and Irving, J. D. (1911) The origin of the Rammelsberg ore deposit, Econ. Geol., vol. 6, 

p. 303-313. 
Loughlin, G. F., and Koschmann, A. H. (1942) Geology and ore deposits of the Magdalena mining 

district, New Mexico, U. S. Geol. Survey, Prof. Paper 200, 168 pages. 
Miller, W. G. (1913) Cobalt-nickel arsenides and silver deposits of Temiskaming, Onatario Bur. Mines, 

19 Ann. Rept., pt. 2, 279 pages. 


Montgomery, Arthur (1948) Mineralogy of the silver ores of Gowganda, Ontario, Univ. Toronto, 

Studies, Geol. Ser., no. 52, p. 23-38. 
Neumann, H. (1944) Silver deposits at Kongsberg, Norges Geologiske Undersokelse, no. 162, PL 

VII, Phot. 14. 
Newhouse, W. H. (1932) Tlte composition of vein solutions as shown by liquid inclusions in minerals, 

Econ. Geol., vol. 27, p. 419-436. 
(1933) The temperature of formation of the Mississippi Valley lead-zinc deposits, Econ. Geol., 

vol. 28, p. 744-750. 
and Flaherty, G. F. (1930) The texture and origin of some banded or schistose sulphide 

ores, Econ. Geol., vol. 25, p. 600-620. 
Newland, D. C., and Hartnagel, C. A. (1908) Iron ores of the Clinton formation in New York 

Slate, N. Y. State Mus., Bull. 123, p. 5-182. 
ftdman, O. H. (1945) A nickel-silver mineralization in the Laver Copper Mines, No. Sweden, Sveriges 

Geol. Undersok, no. 470, ser. C. 
Palmer, Chase, and Bastin, Edson S. (1913) Metallic minerals as precipitants of silver and gold, 

Econ. Geol., vol. 8, p. 140-170. 
Papenfus, E. B. (1931) "Red bed" copper deposits in Nova Scotia and New Brunswick, Econ. Geol., 

vol. 26, p. 314-330. 

Ray, J. C. (1930) Synthetic sulphide replacement of ore minerals, Econ. Geol., vol. 25, p. 433-451. 
Rogers, A. F. (1917) A review of the amorphous minerals, Jour. Geol., vol. 25, p. 515-541. 
Ross, Clarence S. (1929) Is chromite always a magmatic segregation? Econ. Geol., vol. 24, p. 641-645. 
Rust, George W. (1935) Colloidal primary copper ores at Cornwall Mines, southeastern Missouri, 

Jour. Geol., vol. 43, p. 398-426. 
Sampson, Edward (1931) The origin of chromite, Econ. Geol., vol. 26, p. 662-669. 

(1932) Magmatic chromite deposits in southern Africa, Econ. Geol., vol. 27, p. 130. 

Schneiderhohn, H. (1923) Chalkographische Untersuchung des Mansfelder Kupferschiefers, Neues 

Jahrb. Mineral., vol. 47, p. 14^38. 
and Ramdohr, Paul (1931) Lehrbuch der Erzmikroskopie, Berlin, Geb. Borntraeger, 714 

pages, 235 figs. 
Schouten, C. (1934) Structures and textures of synthetic replacements in ore spaces, Econ. Geol., vol. 

29, p. 611-658. 
(1946a) The role of sulphur bacteria in the formation of the so-called sedimentary copper ores 

and pyritic ore bodies, Econ. Geol., vol. 41, p. 517-538. 

(1946b) Some notes on micropseudomorphism, Econ Geol., vol. 41, p. 348-382. 

Schwartz, G. M. (1925) A sulphide diabase from Cook County, Minnesota, Econ. Geol., vol. 20, p. 


(1930) Pseudo-eutectic textures. Discussion, Econ. Geol., vol. 25, p. 658-663. 

(1931a) Intergrowths of bornile and chalcopyrite, Econ. Geol., vol. 26, p. 186-201. 

(1931b) Textures due to unmixing of solid solutions, Econ. Geol., vol. 26, p. 739-763. 

Sharwood, W. J. (1911) Analyses of some rocks and minerals from the Homestake Mine, Lead, South 

Dakota, Econ. Geol., vol. 6, p. 729-789. 
Short, M. N. (1940) Microscopic determination of the ore minerals, U. S. Geol. Survey, Bull. 914, 

2d ed., 311 pages. 
Singewald, J. T. (1913) The titaniferous iron ores of the United States, U. S. Bur. Mines, Bull. 64, 

145 pages. 

Spurr, J. B. (1923) The ore magmas, McGraw-Hill Book Co., 915 pages. 
Swartzlow, Carl R. (1934) Two dimensional dendrites and their origin, Am. Mineral., vol. 19, p. 

Talmage, Sterling B. (1929) The significance of unsupported inclusions, Econ. Geol., vol. 24, p 

Uglow, W. L. (1917) Gneissic galena ore from the Slocan district, British Columbia, Econ. Geol., vol. 

12, p. 643-662. 
Van der Veen, R. W. (1925) Miner agraphy of ore deposition, G. Naeff, The Hague. 


Vogt, J. H. L. (1921) The physical chemistry of the crystallization and magmatic differentiation of 

igneous rocks, Jour. Geol, vol. 29, p. 318-350, 426-443, 515-539, 627-649. 
Waldschmidt, W. A. (1925) Deformation in ores, Coeur D'Alene district, Idaho, Econ. Geol., vol. 

20, p. 573-586. 
Warren, Charles H. (1918) On the microstructure of certain titanic iron ores, Econ. Geol., vol. 13, 

p. 419-446. 
Wtiterry, E. T. (1914) Variations in the compositions of minerals, Washington Acad. Sci., Jour., vol. 

4, p. 111-114. 
Young, S. W., and Moore, N. P. (1916) Laboratory studies in sulphide ore enrichment, Econ. Geol., 

vol. 11, p. 574-581. 

Zsigmondy, Richard (1917) The chemistry of colloids, John Wiley and Sons. 
Zuckert, Richard (1925) Die Paragenesis von Gedeigen Silber und Wismut mil Kobalt-Nickel-Kiesen 

und der Uranpechblende zu Skt. Joahlmsthal in Bohmem, Mittel. der Abt. fur Gesteins, Erz, 

Kohle und Salz-Untersuchungen. Preuss. Geol. Landes., Berlin. 





1. Texture approximating xenomorphic (allotriomorphic) granular in chromite ore from Castle 
Crag mine, Shasta County, California. X 15. Chromite (black) and olivine indent each other 
and are essentially contemporaneous. Crystal facets are rare, and rotund grain outlines pre- 
vail. After Diller. 

2. Hypautomorphic granular texture in chromite ore, Dolbear mine, Siskiyou County, Califor- 
nia. X 15. Chromite largely euhedral and enclosed in olivine which is partly serpentinized. 
After Diller. 

3. Graphic intergrowth of platinum and osmium in a polished nugget from the Ural Mountains. 
Etched with Aqua Regia. Magnification not stated but considerable. After Duparc. 

4. Layered chromite (dark) in serpentine (after olivine), Campo Formosa, Baia, Brazil. Foun- 
tain pen is 12 cm. long. After W. D. Johnston and De Souza. 

5. Hand specimen of banded chromite ore from the Akaishi mine, Japan, showing fluxion 
structure. About half natural size. After Takeo Kato. 

6. Pyrite, partly replaced by network of bornite and chalcocite, all forming the matrix of sand- 
stone. Visingso, Sweden. X 111. After O. H. Odman. 


BA3TIN, PL. 1 



BA8TIN, PL. 2 





1. Hypautomorphic granular texture in sulphide diabase from Cook County, Minnesota. The 
rock consists of fresh olivine (O), augite (A), plagioclase (P), and sulphides (black). Euhedral 
crystals of feldspar are embedded in the sulphides. After Schwartz. 

2. Corrosion of chromite (black) by Olivine that is now completely serpentinized. Feragen, 
Norway. X 35. After Donath. 

3. Synneusis or "swimming-together" texture in chromite (black grains) embedded in olivine 
(gray). Dunite from Hestmando, Norway. X 19. After J. H. L. Vogt. 

4. Unmixing (exsolution) texture in cumberlandite from Cumberland, Rhode Island. Magnetite 
(dark) and ilmenite (light). Dark spot at upper left is olivine. X 190. After C. H. Warren. 

5. Exsolution or unmixing texture in cubanite (Cu) and chalcopyrite (Cp). Republic Mine, 
Fierro, New Mexico. X 90. After Schwartz. 

6. Unmixing or exsolution texture in granular ilmenite rock from St. Urbain, Quebec. Ilmenite 
dark, hematite light. X 62. After C. H. Warren. The lack of relationship between the hema- 
tite bands and the boundaries of the ilmenite grains shows the the banding developed after 
the ilmenite grains had formed. 




1. Unmixing (exsolution) texture in intergrowth of chalcoyprite (light) and bornite. X 560. 
Note narrowing of chalcopyrite bands near intersections. The texture was induced in ore 
from Globe, Arizona by artificial heating and quenching. After Schwartz. 

2. Fractures in pyrite (light gray) filled with chalcopyrite (dark gray). Ruisenor mine, Cuba. 
X 64. After Graton. The matched walls of the veinlets indicate fracture filling. 

3. Gratification in vein in Churprinz mine near Freiberg, Saxony. From walls toward the cen- 
ter the bands are (a) brown sphalerite, (b) white quartz, (c) green fluorite, (d) brown sphal- 
erite, scattered, (e) barite, (f) marcasite, narrow, (g) barite, (h) fluorite, (i) marcasite, (j) 
white calcite, (k) wine yellow calcite. The repetitions of barite, fluorite, and marcasite are 
noteworthy. The vein traverses gneiss. After von Weissenbach. 

4. Dendritic deposit of manganese oxide in joint plane in white porphyry. Leadville, Colorado. 
After Loughlin. 

5. Cockade or ring structure in ore from Alacran Mine, Zacual, Mexico. X $. After Spurr. 

6. Diagrammatic sketch identifying the composition of the rings shown Figure 5. 



I Tuff 4 

t Silicate Quartz. 

I fttyrttie 









1. Delicate trustification in finely crystalline zinc ore or "Schalenblende" from Aachen, Ger- 
many. The specimen is mostly sphalerite with scattered galena crystals in certain bands. 
Courtesy of U. S. National Museum. 

2. Segmented veinlet of niccolite (nc) and silver (s) in smaltite (sm) in ore from Cobalt, On- 
tario. Note that in many places automorphic crystals of smaltite form the walls of the vein- 
let. X 87. After F. N. Guild. 

3. Segmented veinlets of native silver (white) and calcite (black) in safflorite (gray). Note the 
tendency toward matching of the walls in the larger veinlet. Frontier Mine, South Lorraine, 
Ontario, Canada. X 30. After Bastin. 




1. Irregular segmented veinlets of native silver (light gray) and calcite (black) in safflorite (dark 
gray). Note the automorphic outlines of safflorite against the native silver. Frontier mine, 
South Lorraine, Ontario. X 40. After Bastin. 

2. Central part of the field shown in Figure 1 magnified 150 diameters to show the automorphic 
crystals of safflorite (darker gray) projecting into silver (light gray). After Bastin. 

3. Filling of cells of wood by pyrite (white). Most of the filling took place before any collapse 
of the cells, but the cells in the upper left corner appear to have been partly rotted and col- 
lapsed when the pyrite was deposited. Chalcocite (light gray) occupies the spaces between 
the pyrite and has largely replaced the cell walls. X 35. Ady Cupb mine, Sacramento Dis- 
trict, New Mexico. In Permian beds. After R. P. Fischer. 

4. Marcasite in botryoidal masses with spheroidal structure that has replaced sphalerite (gray) 
and gangue (black). Iron Queen mine, Bigbug District, Arizona. X 80. After Lindgren. The 
spheroidal banding is probably due to diffusion and not to surface tension. 

5. Colloform texture on microscopic scale developed in siliceous gold ore from the Talisman 
mine, Karangahake, New Zealand. X 35. After S. F. Adams. 

6. Same field as in Figure 5 but in polarized light. 





BA3TIN, PL. 6 





1. Covellitc (cv) and chalcocite (cc) ore from Bonanza mine, Kennecutt, Alaska. The texture 
is interpreted as metacolloidal on the basis of the spheroidal forms and because the chalcocite 
in replacing covellite seems to have been guided by concentric and radial and network frac- 
tures presumably formed by shrinkage. X 32. After Bateman and McLaughlin. 

2. Irregular fracture pattern or "crackled" texture brought out by etching of chalcocite with 
nitric acid. X 65. Interpreted as due to shrinkage of colloidal chalcocite. Kennecott mine, 
Alaska. After Bateman and McLaughlin. 

3. Brecciated colloform pyrite, Red Cloud mine, Silverton, Colorado. The pyrite fragments lie 
in a matrix of quartz carrying a little sphalerite and galena. Hand specimen. After F. L. 

4. Colloform pitchblende in a gangue of barite. X 35. Radial shrinkage cracks are well de- 
veloped. Joahimsthal, Bohemia. After Schneiderhorn and Ramdohr. 

5. Forking shrinkage crack developed in and confined to globular masses of mixed rhalcocite 
and covellite in a matrix of chalcopyrite. X 155. Cornwall mines, Missouri. After G. W. 

6. Bifurcating shrinkage crack developed in globular area of sulphides that has replaced a dol- 
omite nodule in sandstone. The sulphides are aggregates of tiny chalcopyrite spherules, often 
with pyrite cores, in a bornite matrix. Cornwall mine, Missouri. X 23. After G. W. Rust 




1. A single framboidal group of pyrite crystals from the same specimen as Figure 3 but mag* 
nified 700 diameters to show the crystalline outlines of the pyrite grains. After Rust. 

2. Pyrite crystals disseminated and also in framboidal groups in a matrix of sphalerite. Meggen, 
Germany. X 480. After Schouten. 

3. Framboidal texture in masses of pyrite (p) in the Cornwall mine, Missouri. The pyrite clus- 
ters lie in a matrix of chalcopyrite (cp) and are interpreted by Rust as due to the crystalliza- 
tion of a gel globule about many centers. X 122. After Rust. 

4. Spheroidal outlines, interference surfaces, and shrinkage cracks in pitchblende (light) from 
Great Bear Lake, Canada. The dark-gray mineral surrounding the pitchblende is quartz. 
The shrinkage cracks are filled in part with covellite. X 60. After D. F. Kidd. 

5. Brecciated smaltite (white) whose fractures are filled with colloform native arsenic (gray) 
that has been partially replaced by siderite (dark gray). X 22. After Keil, Freiberg, Germany. 

6. Shrinkage cracks developed in successive bands of chalcopyrite deposited on the walls of a 
fracture in dolomite. Cracks in lower band were developed before upper band was deposited. 
X 10. Cornwall mine, Missouri. After G. W. Rust. 









1. Silicified dolomite, Black Hills, South Dakota. In polarized light the rock is seen to be an 
aggregate of interlocking quartz grains. In ordinary light the crystal outlines of dolomite 
grains are clearly recognizable in spite of the complete silicification. After J. D. Irving. 

2. Replacement of calcite by iodobromite (pebbly gray) with preservation of the crystal out- 
lines of calcite. X 300. Chanarcillo, Chile. After Whitehead. 

3. Preservation of bedding in ore that has replaced limestone, Union mine, near Terry, Black 
Hills, South Dakota. After J. D. Irving. 

4. Transection of schistosity by pyrite cubes. About natural size. After Bastin. The perfect 
pyrite cubes show no distortion and sharply transect the straight schistosity. 

5. Stylolites (jagged black bands) preserved in fluorite. Microphotograph, Spar Mountain, 
Hardin County, Illinois. After Bastin. These fluorite deposits are replacements of limestones. 

6. Oolitic texture in weathered chert nodule in Prairie du Chien limestone near Baraboo, Wis- 
consin. Except for their siliceous character the oolites in the chert are identical with those in 
the enclosing limestone and have formed by replacement. X 5. After Bastin. 




1. Oolite outlines and concentric structure preserved in auriferous quartz and pyrite that have 
replaced dolomitic limestone. Siliceous gold-silver ores of the Bald Mountain area, Black 
Hills, South Dakota. X 22. After Connolly. 

2. Transection of schistosity by a cube of pyrite. Hollinger mine, Porcupine district, Ontario, 
Canada. X 78. After Graton. 

3. Replacement veinlets of anglesite developed along cleavage directions of galena (light). 
Maury mine, Patagonia district, Arizona. X 43. After Bastin. 

4. Galena (with triangular pits) partly replaced by anglesite (dark). The parallelism of the tri- 
angular pits in the three galena areas indicates that they once formed a single galena area 
and have been isolated by replacement. Maury mine, Patagonia district, Arizona. X 38. 
After Bastin. 

5. Pyrite (white) partly replaced by chalcocite along irregular fractures. Visingso, Sweden. 
X 100. After Odman. 

6. Pyrite (light gray) partly replaced by bornite (dark gray). Butte, Montana. X 65. After 
Graton. Note that none of the pyrite areas are in contact. 





BA8TIN, PL. 10 




1. Guided replacement of galena (gal) along its boundaries with quartz (q) and along contacts 
between galena grains by argentite (arg) and electrum (stippled). The argentite and electrum 
may both have replaced galena contemporaneously or electrum may be a late local replace- 
ment of argentite. Extention mine, Tonopah, Nevada. After Bastin. 

2. Partial replacement of galena (gal) by a fine aggregate of a light-colored carbonate, argentite, 
and some chakopyrite. There has also been a slight local replacement of the quartz (q) as 
shown by the ragged boundary of the quartz at A as contrasted with the smooth boundary 
at B. Tonopah-Belmont mine, Tonopah, Nevada. X 100. After Bastin and Laney. 




1. Spherical pellets and crystal grains of pyrite developed by diffuse replacement of shale 
(black). Mount Isa, Queensland. X 460. After Grondijs and Schouten. 

2. Clusters of spheroids of native silver (white) surrounded by rammelsbergite (gray). Both 
have replaced calcite (black). X 140. Laver mine, Sweden. After Odman. The very unusual 
spheroidal form of the silver suggests that it may have been deposited as a colloid. 

3. Atoll shapes of pyrite (white, high relief) with galena cores. Outside the atolls are a carbonate 
(black), sphalerite (gray), and galena (smooth white). All have developed by the replace- 
ment of shale. Mount Isa, Queensland. X 54. After Grondijs and Schouten. 

4. Tuberclelike growths of arsenopyrite replacing calcite (black). Frontier mine, South Lor- 
raine, Ontario. X 27. After Bastin. 

5. Ore from the Townsite mine, Cobalt, Ontario, in which native silver (white) has partly re- 
placed the arsenides of tubercles and the calcite (black) outside the tubercles. X 29. After 

6. Tuberclelike growths of arsenopyrite (white) and tetrehedrite (light gray) in calcite (dark 
gray). X 68. After Bastin. Note that tetrehedrite occurs both inside and outside the arseno- 
pyrite tubercles. 

7. Tubercle texture in rich silver ore from the Townsite mine, Cobalt, Ontario, f natural size. 
After Bastin. 


B AS TIN, PL. 11 








1. Skeleton crystals of native silver enveloped by lollingite in a matrix of calcite. X 1J. Casey 
mine, Cobalt, Ontario. After Bastin. 

2. Native bismuth as crystallization centers for safflorite rosettes. These lie in a quartz matrix. 
Isolated in the quartz can be seen four small star-shaped triplets of safflorite. Annaberg, 
Saxony. X 58. After K. Keil. 

3. Skeleton crystals of silver (white) enveloped in lollingite (light gray) in a calcite matrix (dark 
gray). X 25. Same specimen as Figure 1. Casey mine, Cobalt, Ontario. After Bastin. 

4. Dentritic crystals of silver in calcite in cobalt-nickel-silver ore from Laver mine, Sweden. 
X 23. After Odman. The grain boundaries of the calcite can be dimly seen, and the silver 
bears no relation either to these or to calcite cleavage. 

5. Diffuse replacement of calcite by native silver (white). Laver mine, Sweden. X 77. After 



The six figures form a sequence of cross sections of silver dendrites enveloped by lollingite and lying 
in a matrix of calcite. The dendrites range from 1 to 3 mm. across. Miller Lake-O'Brien mine, Gow- 
ganda, Ontario, Canada. After B as tin. 

1. Silver core (white) of the dendrite forms a five-armed cross enveloped in lollingite (gray) 
Two small veinlets of calcite (black) traverse the lollingite and enlarge slightly where they 
encounter the silver. 

2. Dendrite showing a four-armed silver core in which the tips of two of the arms have been 
replaced by calcite (black). 

3. Dendrite in which several calcite veinlets traverse the lollingite and enlarge abruptly where 
they reach and replace the silver. 

4. Another dendrite showing that it is usually the outer portions of the silver core that are the 
first to be replaced by calcite. 

5. Another dendrite in which portions of the five-armed silver core have been replaced by cal- 
cite. These are the portions adjacent to two calcite-filled fractures which probably served as 
feeders for the replacing solutions. 

6. In this section no silver remains within the dendrite. The core is now wholly calcite but the 
cross-shaped form of the original silver core which it replaced is perfectly preserved. 










1. Silver dendrites (white) enveloped by rammelsbergite-safflorite all in a calcite matrix (black). 
The silver of the original dendrites has been partly replaced by calcite, usually peripherally. 
X 50. Marienberg, Erzgebirge, Germany. After Keil. 

2. Skeleton crystals originally of silver (white) now largely replaced by calcite (black) are en- 
veloped in safflorite and cobaltite. In parts of the larger dendrite a narrow rim of silver 
remains; in the smaller one an irregular remnant of silver remains. X 100. Cobalt, Ontario. 
After Keil. 

3. Skeleton crystals of silver (white) enveloped by rammeisbergite (crusts) in a matrix of cal- 
cite. X 37. Laver mine, Sweden. After Odman. The outlines of the silver are mostly auto- 
morphic but in places are very ragged due to replacement by rammeisbergite. 

4. Rhythmic deposition of smaltite in calcite. Cobalt, Ontario. X 8. After Guild. 

5. Diffusion banding due to oxidation in fine, even-grained felsite. Tonopah, Nevada. Nearly 
natural size. After Bastin and Laney. 




1. Diffusion banding developed in limestone in South Mover mine, Leadville, Colorado. Light 
bands are pyrite, and dark bands are sphalerite. X 2. After Loughlin. 

2. Diffusion banding in jasperoid from Gemini mine, Tintic, Utah. Light bands are chalcedony 
and quartz. Dark bands show grains of galena, sphalerite, and some pyrite. Ordinary light. 
X 15. After Lindgren. 

3. Graphic association of chalcocite (cc) and bornite (b). The chalcocite of the intergrowth is 
continuous with chalcocite that has replaced bornite along quartz contacts. Engel's mine, 
Plumas County, California. After A. F. Rogers. 

4. Graphic association of galena (white) with tennantite and pearceite (both gray). Barite blades 
(black) have not been replaced by galena but by tennantite and pearceite. It follows that the 
latter must also have replaced the galena. Tintic, Utah. X 480. After Lindgren. 

5. Synthetic replacement of pyromorphite (gray) by galena (white) along a multitude of small- 
fractures which are unrecognizable even under high powers of the microscope in the unal- 
tered pyromorphite. X 130. After Schouten. 

6. Synthetic replacement of pyromorphite by galena (white) in which relatively large auto- 
morphic crystals of pyromorphite have been resistant to replacement whereas interstitial 
fine-grained pyromorphite has been completely replaced by galena. X 80. After Schouten. 

7. Synthetic replacement of phosgenite (dark) by galena (light). The replacement seems to 
have been of the diffuse type uninfluenced by the well-defined cleavage of the phosgenite. 
X 33. After Schouten. 










BA8TIN, PL. 16 





1. Complete synthetic replacement of cuprite by metallic copper (white) with preservation of 
cuprite structure and with loss of volume as shown by the numerous vugs (black) and poros- 
ity. X 39. After Schouten. 

2. Same specimen as Figure 1 but in the initial stages of replacement. White is copper, and 
gray is cuprite. X 39. After Schouten. 

3. Dotlike inclusions of galena (white) at intersections of cleavage and twinning planes in cal- 
cite (dark). Mount Isa, Australia. X 58. After Grondijs and Schouten. Note transitions from 
dotlike inclusions to replacement veinlets. 

4. Inclusions of chalcopyrite in sphalerite distributed along twinning directions and cleavage. 
X 72. After Van der Veen. 

5. Quartz (dark) conforming to the outlines of older sphalerite (light) which in part are straight 
crystal faces and in part rotund as is so common in sphalerite crystals. Carmen mine, Zaca- 
tecas, Mexico. X 25. After Bastin. 

6. Calcite crystals (white) deposited on fluorite in Hillside vein, Rosiclare, Illinois. The calcite 
has been deposited like snow on a roof mainly on upward-facing surfaces of the fluorite indi- 
cating a gravitational control in its deposition. X f . After Bastin. 




1. Cubical pyrite (white) showing automorphic outlines against quartz (dark) and chalcopyrite 
(gray). The quartz in turn shows automorphic hexagonal outlines against the chalcopyrite. 
In the absence of any evidences of replacement, the order of deposition is interpreted as py- 
rite, quartz, and chalcopyrite. San Jesus mine, Zacatecas, Mexico. X 28. After B as tin. 

2. Supposed remains of Iron Bacteria from chert in Pokegama formation conformably underly- 
ing the Biwabik iron-bearing formation of the Mesabi district, Minnesota. X 1540. After 

3. Oolitic iron ore showing closely packed odlites of chlorite and hematite with concentric struc- 
ture (A). A highly hematitic and nearly opaque oolite is shown at C. Nuclei of siderite are 
shown at B. The sparce matrix is mainly chlorite with some siderite. X 23. La Ferriere aux 
Etangs, Orne, France. After Cayeux. 

4. Iron ore from Thoste,France, composed of false odlites in a matrix (E) of calcite. A to D are 
crinoid plates partly replaced by hematite. F = false oSlites in which calcite has been exten- 
sively replaced by hematite. G = hematite oolites. X 32. After Cayeux. 

5. Iron ore from Mazenay, France, showing both false oolites composed of crinoid remains (A) 
and oolites (Bto E) in a matrix largely of granular calcite but with some organic remains. 
X 32. After Cayeux. 

6. Iron ore from Mexy, France, with oolites of hematite (black) in a matrix mainly siderite. At 
A are quartz nuclei. Around most of the oolites a narrow band of chlorite (B) has developed 
by the replacement of the siderite matrix. In places, as at C, only ragged replacement rem- 
nants of siderite remain. Where hematite oolites are in contact no chlorite band is present. 
X 29. After Cayeux. 










1. Silurian oolitic iron ore showing peripheral replacement of chloride oolites (A) to various 
depths by white rims of siderite. The smooth outer border of the siderite marks the original 
border of the chlorite oolite. The inner border of the siderite is irregular and in many places 
transects the concentric banding of the chlorite. Nuclei of quartz are shown at B and C, and 
nuclei of siderite at D and E. The matrix consists of siderite rhombs embedded in chlorite 
and hematite. X 23. La Ferriere aux Etangs, Orne, France. After Cayeux. 

2. Clinton oolitic iron ore from Wolcott, New York, composed of rounded and semiangular fossil 
remains partly replaced by hematite. These lie in a matrix of calcite (E). B == oolite composed 
of a nucleus of a rounded bryozoan fragment largely hematized and enveloped in a cortex of 
hematite. C bryozoan remains almost completely filled and replaced by hematite. D - 
crinoid remains partly filled and replaced by hematite. X 14. After Cayeux. 

3. Oolites of pyrite (black) in a matrix of crystalline quartz (c), Wabana, Newfoundland, a * 
pyrite odlite, b = pyritized shell fragment partly replaced by quartz (e), d quartz in oo"lite. 
X 56. After Hayes. 

4. Dynamometamorphosed ore from Rammelsberg, Germany, showing flowage of finely granu- 
lated sulphides around nodules of relatively coarse barite and sphalerite. Natural size. After 
Lindgren and Irving. 

5. Hand specimen of "steel" galena from the Slocan district, British Columbia showing charac- 
teristic banded structure produced by crushing under differential pressure. X 1$. After Uglow. 

6. Zonal texture brought out by etching in native silver from the Elkhorn mine, Montana. 
X 100. After Carpenter and Fisher. 

7. Homogeneous recrystallized texture brought out by etching in native silver (of Precambrian 
age) from Cobalt, Ontario, Canada. X 100. After Carpenter and Fisher. 


Aachen, Germany, Zinc ores of 16 
Adams, S. F. 84 
Adsorption in colloids 31 
Akaishi mine 5 
Allotriomorphic granular 4 
Atoll texture 42, 90 
Automorphic outlines 61 

replacements 37, 41, 42, 51, 71 

Banded textures 5, 74, 80 

Banding due to gravity 6, 80 

Bateman, A. M. 85 

Beck, Richard 17 

Bedding preserved in replacement 35, 87 

Birmingham, Ala., Iron ores 71 

Brecciation in colloidal ores 28 

Brownian movement in colloids 21 

Bucher, W. 67, 71 

Buerger, M. J. 57, 73 

Buerger, N. W. 56, 74 

Buoyancy of colloidal particles 21 

Burchard, E. F. 71 

Burton, E. F. 21 

Bushveld, Chromite in 6 

Carpenter, H. C. 74, 75, 97 

Cayeux, L. 65, 96, 97 

Cell structures preserved in replacement 36, 84 

Chain structure 7, 8 

Chemical reactivity of colloidal particles 22 

Chromite ores 6 

Clark, J. D. 24 

Cleavage preserved in replacement 35, 36 

Clinton iron ores 70, 97 

Cockade structure 17, 82 

Colloform texture 26, 28, 84, 85 

Colloidal deposition, Principles of 20 

ore textures 25, 85, 86 

replacements, 26, 33, 51 

silica 23, 51 

suspensions 22 
Colloform textures denned 26 
Comb structure 14 
Connolly, J. P. 88 

Cornwall mine, Missouri 29, 30, 31 
Corrosion in magmatic ores 8, 9, 81 
Crustification 14, 15, 16, 82, 83 
Crystal outlines preserved in replacement 35 
Crystallization, Order of 7 

Dendrites 19, 44, 48, 49, 82, 91, 92, 93 

de Souza, H. Capper A. 6, 80 
Diffuse penetration textures 41, 90, 91 
Diffusion banding in colloids 28 
or rythmic banding 47, 93, 94 
Diller, J. S. 7, 80 
Donath, M. 81 
Duparc, Louis 80 

Edwards, A. B. 13, 37, 73 

Ehrenberg, H. 16 

Electrical charge on colloidal particles 21 

Emmons, W. H. 17 

Emulsoids denned 23 

England, Iron ores of 69 

Exsolution textures 10 

False oolites 66, 67, 96 

Fischer, R. P. 75, 84, 97 

Flaherty, G. F. 73 

Flow banding 5, 80 

Fluid inclusions 57 

Fossil forms preserved in replacement 36, 37 

Fracture fillings 17, 18, 82 

Framboidal texture 30, 86 

France, Iron ores of 65 

Gavelin, S. 56 
Gels 23, 26 

Water content of 26 
Gowganda, silver ores 46 
Grain size in replacement 52 
Graphic textures 5, 80, 94 
Graton, L. C. 88 

Gravitative control in ore deposition 61 
Grondijs, H. F. 30, 90, 95 
Grout, F. F. 3 
Gruner, J. W. 72, 97 
Guest mineral, denned 33 
Guided penetration textures 38, 89 
Guild, F. N. 48, 83, 93 

Hallimond, A. F. 69 

Harder, E. C. 72 

Hartnagel, C. A. 70 

Haycock, M. H. 49 

Hayes, A. O. 69, 72, 97 

Hintze, Carl 26 

Host mineral, denned 33 

Hypautomorphic granular 4, 8, 80, 81 

Hypidiomorphic denned 1, 5 

Hypogene replacement 52 




fdiomorphic defined 1 

Fluid 57 

Solid 56, 57, 58 
due to replacement 56, 95 

unmixing 56 

Ingerson, E. 59 

Interference surfaces in colloids 31 

frving, J. D. 73, 87, 97 

Johannsen, Albert 4 
Johnston, W. D., Jr. 6, 80 

Kato, Takeo 5, 80 

Keil, Karl 44, 45, 48, 86, 93 

Kidd, D. F. 49, 86 

Kiruna 6 

Koschmann, A. H. 37 

Krieger, P. 48 

Levings, J. V. 23 

Liesegang, R. E. 50 

Lindgren, W. 26, 28, 50, 51, 73, 84, 94,|97 

Loughlin, G. F. 49, 57, 82, 94 

Levering, T. S. 9 

Magmatic changes, late 8 

ores 3 

McLaughlin, D. H. 37, 85 

Menaul, P. L. 24 

Metacolloidal texture 85 

Metasomatism defined 34 

Micrographic intergrowths, Synthetic 54 

textures 5 

Miller, Willet G. 49 

Montgomery, A. 49 

Moore, Neil P. 53 

Mount Isa, Queensland 30, 37, 41, 43, 56 

Neumann, H. 48 
Newfoundland, Iron ores of 69 
Newhouse, W. H. 57, 58, 73 
Newland, D. C. 70 
Nodular textures 6, 7 

Odman, O. H. 44, 48, 80, 88, 90, 91 
Oolitic textures in ores 36, 65, 88, 96, 97 

preserved in replacement 36, 87 

Open spaces, Deposition in 14 

Ophitic texture 5 

Orbicular textures 4, 6 

Order of crystallization 7 

Organic forms preserved by replacement 36 

Overlap in ore deposition 62 

Palmer, Chase 24 

Paragenesis 60 

Paragenetic diagrams 63 

Parallel orientation patterns 41 

Pellet texture 30, 42, 90 

Peptizers 24 

Plant cell fillings 19 

Poikilitic texture 4, 5 

Pore fillings 19 

Porphyritic texture 4, 8 

textures preserved in replacement 36 

Pseudomorphic replacement textures 34, 87 

Pyritic oolites 70, 97 

Ramdohr, Paul 1, 85 
Ransome, F. L. 85 
Ray, J. C. 53 
Reaction rims 10, 11 

Definition 33 

Depth relations 53 

Grain size of guest and host 52 

Graphic textures in 51 

Hypogene versus supergene 52 

Inclusions due to 56 

Influence of composition of host in 33 

Parallel orientation patterns in 41 

Simultaneous by several minerals 38 

Temperature range of 33 

Tight contacts in 40 

Veinlets formed by 38, 88 
textures 33 

Colloidal 26, 33, 51 

Synthetic 53 
Ring structure 16 
Rogers, A. F. 26, 94 
Ross, C. S. 4 
Rupture veinlets 19 
Rust, G. W. 31,85, 86 
Rythmic or diffusion banding 28, 47 

Sampson, Edward 4, 7 
Schalenblende 16 
Schneiderhohn, Hans 1, 30, 85 
Schouten, C. 31, 41, 54, 86, 94, 95 
Schwartz, G. M. 4, 12, 52, 56, 81, 82 
Sedimentary ore deposits 65 
Segmented veinlets 17, 83 
Selective replacement 33, 55 
Sharwood, W. J. 17 
Short, M. N. 1 

Shrinkage or syneresis cracks 29, 85, 86 
Silica, Colloidal, in ores 23 



Simultaneous deposition 62 

Siskiyou County, Chromite in 6 

Skeletal crystals or dendrites 19, 44, 48, 49, 92, 


Solid solutions 11 
Spheroidal forms in colloids 27 
Spurr, J. E. 3, 82 
Stabilizing agents 23 

Stylolites preserved in replacement 35, 87 
Surface tension in colloidal ore deposition 25 
Suspensoids defined 23 
Syneresis or shrinkage cracks 29 
Synneusis 7, 29, 81 
Synthetic replacements 53, 94, 95 

Talmage, Sterling B. 17 

Transecting replacement textures 37, 88 

Tubercle texture 43, 90 

Unmixing of solid solutions 10, 81, 82 

Van der Veen, R. W. 45, 48, 49, 56, 95 
Veinlets with automorphic walls 18 
Vogt, J. H. L. 7, 81 
Volume relations in replacement 34, 55 

Warren, Chas. H. 81 
Whitehead, W. L. 87 

Xenomorphic denned 1, 4 
granular 8, 80 

Young, S. W. 53 

Zoning in magmatic ores 8 
Zuckert, R. 49