s 14, GS: Cq^cA Sax0>O£jU CIR 3{ Goniol ithon, etc. 23 Halimeda 4 m.% MgCO (range and av. ) after elimination of organic matter and water from the analyses m% ca 3 (po 4 ) 2 (range and av. similarly calculated 1.79-11.08(9.61) 4.61-14.10(8.02) 0.09-1.11(0.50) *6. 18-15. 73(11. 12) 0.24-1.28(0.54) 0.00-9.72(3.35) Trace Trace-9.96(3.38) Trace Trace-8.57(2.34) Trace Trace-0.99(0.54) 7.86-13.74(10.86) Trace-1.12(0.21 0.80-20.23(3.43) 5.41-13.79(8.51) 7.79-14.35(10.94) 6.61-14.95(10.52) 13.84 0.17-11.08(5.41) 0.49-8.63(2.42) 0.79-6.68(2.97) 0.00-1.00(0.22) 0.20-0.45(0* 0.00-1.78(0. 0.16-6.02(2. 0.75-2.49(1. 3.65-15.99(8 32) 29) 60) 82) .28) 10. 93-25. 17( 0.02-1.09(0. 17.74) 65) Trace(one had 0.20) Trace-1.85(0.28) Trace-0.78(0.15) Trace-1.14(0.17) Trace Trace-8.47(2.08) Trace-0.57(0.19) 74.73-91.74(82.91) Trace(two had 0.07, 0.40) Trace Trace-0.85(0.10) Trace Trace (one had 0.77) 6.57-49.56(17.12) Trace-0.43(0.05; Trace * One value (Heliopora cerulea, Philippine Islands) of 0.35. limestone of the Oregon Quadrangle, Illinois, that were partly filled by residual dolomite sand, and Stout (1940) described local areas of the Peebles Dolomite of Niagaran (Silurian) age in Pike, Adams, and Highland Counties, Ohio, that have been rendered soft and porous through leaching. 20 ILLINOIS STATE GEOLOGICAL SURVEY CIRCULAR 298 Table 2. - MgCO in Calcareous Skeletal Materials (Compiled from Chave (1954a). The analyses include 43 from Clarke and Wheeler (1922), for which water temperatures were known.) Number of Percent Weight Percent Classification samples arago nite MgC0 3 Foraminif era 23 <4-15.9 Sponges 3 5.5-14.1 Madreporian corals 10 100 0.12-0.76 Alcyonarian corals 14 6.05-13.87 Echinoids 25 4.5-15.9 Echinoid spines 12 <4-10.2 Asteroids 9 8.6-16.17 Ophiuroids 6 9.23-16.5 Crinoids 22 7.28-15.9 Annelid worms 12 0- 99 6.4->16.5* Pelecypods 11 0- 100 0.09-2.80 Gastropods 7 5- 100 0.08-2.40 Cephalopods 3 0- 100 0.05-7.00 Decapod crustaceans 6 5.2-11.70 Ostracode crustaceans 6 <4-10.2 Barnacles 9 1.35-4.60 Calcareous algae 15 7.7-28.75 * The highest values cited apply to small amounts of calcite found in the more aragonitic materials formed at higher temperatures. MAGNESIUM DISTRIBUTION IN CARBONATE ROCKS AND SEDIMENTS Magnesium in Skeletal Carbonates Clarke and Wheeler (1922) presented an extensive series of original chem- ical analyses of invertebrate skeletal materials, summarized here as table 1, as well as an evaluation of earlier literature on the subject. Some of the materials contained more than 20 wt percent MgC03 . Even though the available data on water temperatures at the points where the various organisms were collected were incomplete, Clarke and Wheeler were able to show quite clearly that the percentage of MgC03 in the skeletal material of alcyonarian corals and echinoderms increased with water temperature. Fossil crinoids contained distinctly less MgCOo than their present-day counterparts, the MgC03 content of sea urchin spines differed from that of shells and teeth, and the magnesia and phosphorus contents differed for lobsters of different ages . Chave (1954a) showed by X-ray diffraction that the MgCO-, in invertebrate skeletal materials is present in solid solution, and that the amount of solid solu- tion in aragonitic materials (rarely more than 1 wt percent MgCOo) is in general much less than that in calcific materials (rarely less than 1 to as much as 20 or more wt percent MgC03) (see table 2) . In all groups of calcitic organisms for which sufficient data were available, a linear or nearly linear relation could be shown between Mg content of the skeleton and water temperature. An increase in phylogenetic level of the organism is in general accompanied by a decrease of MgC03 in the skeleton and by a decrease in the slope of the wt percent MgC03 vs. temperature plot. Chave, like Clarke and GEOCHEMISTRY OF SEDIMENTARY CARBONATES - II 21 Table 3. - Mineralogical Classification of Foraminifera by Families (after Blackmon and Todd, 1959) Imperforate Perforate Highly magnesian (>10 mol % MgC0 3 ) Calcite low in magnesium (<5 mol % MgCO ) Aragonite Single crystal Radial microstructure Granular microstructure Alveolinellidae Fischerinidae Miliolidae Ophthalmidiidae Peneroplidae Spiril- linidae Calcarinidae Camerinidae Discorbidae Heterohelicidae Homotremidae Planorbulinidae Buliminidae - — • Anomalinidae — — Elphidiidae Globigerinidae Globorotaliidae Lagenidae Polymorphinidae Rupertiidae Buliminidae Pegidiidae Anomalinidae Cassidulinidae Chilostomellidae Nonionidae Rotaliidae Ceratobu- liminidae Rober- tinidae *Amphisteginidae *Cymbaloporidae * Indicates intermediate to low range Wheeler, observed that the tests and spines of the same echinoid individuals fre- quently showed different MgC03 contents. Lowenstam (1954b) showed that the percentage of aragonite in serpulid worm tubes increases with increasing water temperature, and Chave pointed out that the total MgC03 content of such tubes should increase with temperature as more Mg is taken into the calcite, but decrease with further temperature rise because of the increasing percent of aragonite. In a still more detailed X-ray diffraction study, Blackmon and Todd (1959) examined shells of 155 species of calcareous foraminifera, belonging to 29 Recent families, and found that the mineralogy with few exceptions fitted the pattern re- produced here as table 3. The calcite of some samples contains as much as 20 to 24 mol percent MgC03- The control of the carbonate mineralogy of these materials is believed to be mainly biologic, although a depression of MgCO^ content, highly variable in amount among different species, is observed in colder waters for all foraminifera with shells of highly magnesian calcite. Pilkey and Hower (1960) found that the percentage of MgC03 in specimens of Dendraster excentricus collected from the coast between British Columbia and Baja California ranges from 8.4 to 9.5 (215 analyses) and that the percentage of SrC03 ranges from 0.296 to 0.354 (117 analyses). These authors conclude that that Mg content is directly related to both water temperature and salinity, and that Sr content is inversely related to water temperature and independent of salinity. 22 ILLINOIS STATE GEOLOGICAL SURVEY CIRCULAR 298 Goldsmith et al. (1955) made additional spectrographic analyses and cell- size measurements on some 25 of Chave's samples, and described one Mg-rich al- gal calcite that contained several mol percent MgCOo in some form other than ideal solid solution, as shown by a measurable cell-size change when homogenization took place at high temperature. In view of the several variables that affect the content of MgCOo in skeletal materials, summaries such as presented here in tables 1, 2, and 3 are of only gen- eral value in indicating skeletal MgCOj contributions to sediments. In apportioning the total MgC03 content of a sediment among the several skeletal contributors, so as to determine whether there has been a gain or loss of MgCOo to sea water, for example, it is necessary to identify the organisms as closely as possible and to refer to detailed information on the MgC03 contents of individual species or small taxonomic groups. Chave (1954b) obtained reasonable agreement between total and apportioned MgCOo in several sediments. He discussed earlier attempts by Vaughan (1918), Bramlette (1926), Goldman (1926), and Thorp (1935), suggesting that the use of average MgCO~ contents from Clarke and Wheeler's data and the difficulty of distinguishing fine-grained fragments of aragonitic and calcitic algal calcites limited the accuracy of their results. Chave (1954b) found that calcitic fos- sils in porous sandstones and oolites of Pleistocene age had lost Mg, that some of the fossils in silts and marls of intermed- iate porosity had lost Mg, but that fossils imbedded in relatively nonporous post- Paleozoic shales had retained their Mg. The Key Largo Reef and the Miami (back- reef) Oolite of Pleistocene age in south Florida have been above sea level most of the time since they were formed, and Mg has been leached by ground water. The Miami Oolite, which contains about 15 per- cent fossils originally high in Mg and mol- lusks and oolites that originally contained smaller amounts of Mg, now has only 0.37 percent MgCOo . 10,000- 8,000- 6,000- 4,000 2,000 (a) 30-| 20- 10- (b) DOLOMITE — CALCITE 100 20 40 60 80 PERCENT 100 Mg - 2.5 5 7.5 10 13.2 DOLOMITE - - 19 38 57 76 100 CALCITE 100 81 62 43 PERCENT 24 Fig. 3 - The frequency of occurrence of various calcite-dolormte mixtures in the Middle and Upper Paleozoic carbonate rocks of the Russian Platform (Ronov, 1956): (a) computed from 23,101 individual analyses, and (b) based on measurements of areas on maps contoured for Mg content. GEOCHEMISTRY OF SEDIMENTARY CARBONATES - II 23 480 420- 360 300- Ll. 240- 180- 120- hJ c/> >■ _l < < o a: 00 DOLOMITE — CALCITE 100 Relative Abundance of Various Calcite-Dolomite Mixtures Plots of the frequency distribution of calcite-dolomite mixtures occurring in carbonate rocks, made for large numbers of samples (Steidtmann, 1917; Ronov, 1956; Nikolaev, 1956), show maxima near both ends of the range (figs. 3, 4, 5) . If dolomitization in nature took place by some one process (for example, sec- ondary replacement by Mg derived from an external source), this bimodal distribution would suggest that a threshold of some sort must be exceeded before dolomitization can proceed, but that once the threshold had been passed, dolomitization has a good chance of going essentially to com- pletion. It appears more likely, however, that the two ends of the range are being reinforced statistically by carbonate rocks of special types, in the one case by fine- grained, virtually pure, supposedly chem- ically precipitated dolomite rocks from evaporite sequences, in the other by the weakly dolomitized rocks in which very local redistribution of the Mg originally in biogenic magnesian calcites is ade- quate to account for the dolomite present. When Ronov 's analyses are used to plot maps of the Mg concentration in various parts of the Paleozoic sequence on the Russian Platform, and the areas of various Mg concentrations on the maps are made the basis of a frequency distribu- tion, the picture shown in figure 3b is obtained and is quite different from that in figure 3a. Many of the intermediate mixtures in figure 3b must result from averag- ing of analyses for sub-layers, some of which are rich in dolomite, others in cal- cite. The frequency of intermediate mixtures in the plots of figures 3a, 4, and 5 probably is increased slightly by an analogous effect. Analyses of samples of finely interlaminated dolomite and calcite would necessarily be an average of a considerable number of such layers. Such an analysis may indeed be a perfectly valid statement of rock composition, but it does not correspond to either of the two depositional environments in which the successive laminations were formed. PERCENT Fig. 4 - Computed percentages of calcite and dolomite for 1148 analyses of car- bonate rocks from North America (Steidt- mann, 1917; reproduced with permission of the "Bulletin of the Geological Society of America") . Dolomitization Throughout Geologic Time Daly (1909) compiled 900 chemical analyses of limestones and dolomites from Belgium, Canada, and the United States and found that the average Ca/Mg ratios were relatively constant for pre-Devonian rocks but rose steadily for Devonian and Carboniferous samples and reached an apparent maximum during Cretaceous time, although the number of samples of Tertiary and younger rocks was rather small. The increase of Ca/Mg .ratios with geologic time for carbonate rocks of the Russian Platform (Vinogradov et al., 195 2) is shown in figure 6. A number of explanations have been put forth to explain this distribution, including the increased time the 24 ILLINOIS STATE GEOLOGICAL SURVEY CIRCULAR 298 400 300 GO Ld CO >- L»- 200 O cr UJ CO 100 fl i-l n 1 J f 1 V y j T. 20 40 60 PERCENT DOLOMITE 80 100 Fig. 5 - The frequency of occurrence of various calcite-dolomite mixtures in Upper Carboniferous rocks of the Samarov Valley. The rocks rarely contain more than 2 percent insoluble residue (Nikolaev, 1956). older rocks have had to undergo post-depositional replacement, the times at which particular groups of lime-secreting organisms appeared, and the possibility that the partial pressure of CO2 in the atmosphere was greater in earlier geologic periods than more recently. Strakhov (1956) pointed out that aridity and concomitant high evaporation rates seem to have played an important part in adjusting water composition for dolo- mite formation, both in present-day lacustrine deposits and in the Paleozoic rocks of the Russian Platform. Until comparisons of carbonate rocks of different geologic ages from similar climates can be made, the possibility remains that the compilations mentioned above are influenced by average past climates in the countries from which the samples came. Some Empirical Generalizations Fairbridge (1957) showed a correlation between dolomite content and the per- centage of insoluble residue in the analyses of "Cambro-Ordovician" rocks from near Harrisburg, Pennsylvania, published by Lesley (1879). However, the Ellenburger GEOCHEMISTRY OF SEDIMENTARY CARBONATES - II 25 36 r 32 28 24 20 16 12 PCm Cm 510 430 S, I S 2 iDjbjl IJCjCglCjl P, IP 2 I T 310 — 275 225 150 1 Cr I T> lol 1 10 - •70- ABSOLUTE TIME IN MILLIONS OF YEARS Fig. 6 - The change with geologic time of the Ca and Mg content of the car- bonate rocks of the Russian Platform (after Vinogradov et al . , 1952). (I) Based on analyses of 97 composites of 3569 samples. (II) Based on 2826 published analyses for Ca, 2583 for Mg. limestones and dolomites analyzed by Goldich and Parmalee (1947), virtually all of which contain less than 7 percent insoluble residue, show no evident correlation between the calculated normative percentage of dolomite and weight percent insol- uble residue. Folk. (1958) found an average of 12.56 percent insoluble residue in the nine dolomite beds of the Axemann Formation of the Beekmantown Group in Centre County, Pennsylvania, but an average of only 6.52 percent in nineteen limestones and dolo- mitic limestones in that formation. The difference lies in the kind and amount of authigenic silica; euhedral quartz crystals and overgrowths in limestones are out- weighed by the large amount of disseminated chert in the dolomites, which also con- tain some quartz overgrowths and aggregates. It is possible that a similar explana- tion applies to Lesley's analyses. 26 ILLINOIS STATE GEOLOGICAL SURVEY CIRCULAR 298 When the Funafuti core was the only one available from the Pacific, the specific distribution of dolomite with depth in that core was made the basis for con- siderable speculation. It is now known that the distribution of dolomite with depth on Pacific islands is extremely variable (Ladd et al . , 1948). On Kita Daito Jima there is strong dolomitization from the surface down to 103.5 meters, and then only a small MgO content except for a few intervals (Hanzawa, 1940); at Funafuti, dolo- mite is found below 637 feet (Cullis, 1904); at Bikini, there is no dolomitization down to the bottom of the hole at 2556 feet depth; two Great Barrier Reef bores to 600 feet and 732 feet showed no dolomite (Richards and Hill, 1942). Appreciable dolomite was found at Eniwetok in only four samples of Eocene age at depths below 4300 feet (Ladd et al . , 1953). The MgCOg values do not appear to have been re- ported in detail for two holes drilled on Maratoea Atoll northeast of Borneo to depths of 856 and 1407 feet (Kuenen, 1947). The dolomite discussed above is inferred from chemical analyses; small percentages of MgO that, particularly in the upper parts of the cores, could be present in magnesian calcites have been ignored. In' the Mg-rich horizon near the top of the Funafuti core, X-ray diffraction measurements (Schmalz, 1956; Graf and Goldsmith, in press) have shown that the Mg occurs in magnesian calcite and not in dolomite. IMPURE CARBONATE ROCKS Arenaceous and Cherty Carbonate Rocks The silica in carbonate rocks occurs as detrital and authigenic quartz, chert, and the so-called amorphous silica. The last of these forms the shells of siliceous organisms such as diatoms, radiolaria, and siliceous sponges. It may be hydrated and consists of a disordered association of cristobalite crystallites very little larger than one unit cell and therefore not properly called crystalline cristo- balite (Warren and Biscoe, 1938). Although angular chert fragments have been described from the siliceous limestone portion of the Zesch Formation (Barnes et al., 1947), it is generally quartz that is encountered as a detrital contribution. The variation this contribu- tion may show with change in depositional environment is well illustrated by the Ballyshannon Limestone of early Carboniferous age. In the Donegal Syncline this limestone shows rapid facies changes between deltaic sandstone members to the north and shallow-water oolites and fragmental bioclastic limestones with a grad- ually decreasing content of smaller and smaller quartz and feldspar grains to the south (George and Oswald, 1957). Unusual rock types in the Ballyshannon include an oolite made up of angular quartz grains with oolitic skins, and a fine-grained limestone containing an abundance of both fine silt and sponge spicules. Such fine-grained quartz-carbonate mixtures may be difficult to classify in the field. Two samples from the Paradox Member of the Hermosa Formation of Pennsylvanian age in southeast Utah which would routinely have been called calcareous shales were shown by X-ray diffraction analysis to consist of 50 percent calcite and dolo- mite, 40 to 45 percent quartz, and 5 to 10 percent illite, montmorillonite, and chlor- ite (Merrell, 1957). The chert in carbonate rocks commonly occurs as beds and as nodules. The replacement origin of the latter is shown by, among other things, residual particles of organic matter outlining oolites and other replaced textural features (see Biggs, 1957; Folk, 1958). All gradations between pure limestone and pure chert occur. Ver Wiebe (1946), for example, described several wells in Sedgwick County, Kansas, GEOCHEMISTRY OF SEDIMENTARY CARBONATES - II 27 in which dolomitic layers in the Kinderhook Limestone of Mississippian age contain 40 to 50 percent chert; Dott (1958) found many limestone units in the Pennsylvanian rocks of northeastern Nevada that contain more than 30 percent chert. Porcelanite of the Lower Triassic Thaynes Formation of west-central Utah is seen in thin-section to contain about 50 percent disseminated quartz present in a fine mosaic in calcite and dolomite (Hose and Repenning, 1959). Chert generally does not, however, re- place collophane (Carozzi, 1960, p. 323) or dolomite, which exists as euhedral crystals in chert nodules and passes through them as unreplaced laminae (Folk, 1958; Biggs, 1957; Hatch et al . , 1938). The chert in most dolomites is a uniform- ly impalpable powder, which in the Beekmantown Group of Centre County, Pennsyl- vania, makes up 5 to 10 percent of the rock (Folk, 1958). Dolomites of the Ellen- burger Group of early Ordovician age in central Texas at many places contain highly disseminated chert which seldom is visible on fresh rock surfaces but which on weathering yields spongy masses containing abundant holes of rhombic shape (Cloud et al., 1946). Attempts have been made to distinguish the relative amounts of chert found in limestones of different types and origins. Vishnyakov (1953) claimed that chert is more common in fine- to medium-grained deeper water limestones and deep-water foraminiferal limestones than in shallower water brachiopod facies, and Teodoro- vitch (1950) stated that oolitic limestones are poor in chert. Hattin (1957) noted that chert usually is not found in purely molluscan or purely algal limestones. Reef cores and reef-like accumulations of skeletal remains are not entirely free of silica. There are, for example, small amounts of secondary chert and a little drusy quartz in the cores from the Horseshoe Atoll, of Pennsylvanian and early Permian age, in west Texas (Myers et al . , 1956). However, there is typical- ly a much greater amount of silica in the interreef limestones (see Pray, 1958). These estimates seem to indicate that limestones formed in waters agitated enough to winnow out the finer grained detrital silica and the invariably small and relative- ly buoyant siliceous shells, if indeed these materials enter the environment in sig- nificant amounts, are less likely to contain chert. Nature and Origin of Chert in Carbonate Rocks Jensen et al. (1957) made an X-ray diffraction study of 42 Danish "flints" and other siliceous rocks (all but one derived from chalk deposits of Cretaceous age) and found they contained from 4 to 95 percent fine-grained a- quartz, typical- o ly of 300 to 500 A particle size. The remainder consisted principally of flaky, crypto- crystalline material having a two-dimensional cristobalite arrangement. These authors were able to correlate with fair to moderate accuracy estimates made by X-ray and petrographic methods of the amount of cristobalite-like material, which would be called opal by a petrographer. Chert in lacustrine dolomite beds of Tertiary age near Ipswich, Queensland, gives an X-ray diffraction pattern "similar but not iden- tical to" that of (3 -cristobalite (Rogers et al . , 1954). Significantly, Folk (1958) failed to detect by petrographic means any opal in the chert contained in the much older carbonate rocks of the Beekmantown Group of early Ordovician age, and Biggs (195 7) detected no "amorphous silica" or opal, only microcrystalline quartz, on microscopic examination of chert from 18 Illinois limestone and dolomite formations ranging from Cambrian to Mississippian in age (see also Rubey, 1952, on the chert of the Burlington Limestone of Mississippian age in Illinois). Fibrous quartz ("chal- cedony"; see Pelto, 1956) generally is restricted to places where a free surface is available as a starting point for its growth (Folk and Weaver, 1952; Biggs, 1957). 28 ILLINOIS STATE GEOLOGICAL SURVEY CIRCULAR 298 The "amorphous silica" in the siliceous tests of organisms is metastable, c-quartz being the stable form of SiO? at room temperature, and it is known from experimental studies of reactions in concrete to be highly reactive toward alkalis. Jensen et al. (195 7) suggested that weakly alkaline waters saturated with CaCOg in the chalks apparently are sufficient to induce concentration and transformation of disseminated "amorphous silica" (see also Correns, 1950). CaC03~saturated water in chalk, not in contact with air, would have a pH of about 9.9; in contact with air, it would have a pH of about 8.3 (R. M. Garrels, personal communication). The first of these figures is significantly above the value of 9, which Krauskopf (1956) found to be the upper limit of the range in which silica solubility is essen- tially independent of pH. The water and residual CaCOg found in many chert nod- ules, as well as the very small particle size of the c-quartz therein, indicate the difficulty of obtaining the truly stable end product, macrocrystalline a-quartz. There is additional support for the origin of at least some chert by reorgan- ization of organic silica. Extensive solution of the silica of diatom shells in the Monterey Formation and its reprecipitation as a cement and as chert masses and beds are shown by lateral transitions within short distances between the cherty or porcelaneous beds and soft diatomites. Significantly, Bramlette (1946) sug- gested that the unaltered diatomites may be poorer in carbonate than the porcelane- ous and cherty parts. Cayeux {in Newell et al. , 1953) and others believed that the principal source of silica for chert in European chalks is the siliceous spicules of sponges. Nodular chert in the Helderberg Limestone of West Virginia contains sponge spicules and is clearly of replacement origin. All stages, from normally calcareous oolites and fossils to their completely silicified equivalents, can be observed. The sponge spicules in the limestone outside the nodules have been largely calcified, indicating at least a partial source for the silica of the nodules (Heald, 1952). Black, silica-rich concretions in a sponge-bearing Onondaga Lime- stone, found by Schwartz and Mathiasen (1934) to contain 0.38 percent finely sub- divided free carbon, are believed by them to be recrystallized and calcite-enriched concentrations of siliceous sponge remains (analysis 55, Part IV). Limestone of the Rauracian Formation of Upper Jurassic age in southern Poland contains very abundant calcified sponge spicules, whereas many of the spicules in adjoining chert-rich beds seem to have served as nuclei for chert growth (Sujkowski, 1958). A similar situation exists for the Turonian white chalk of Polish Volhynia. In the Phosphoria Formation of northwestern Wyoming there are all gradations between chert farily rich in organic matter, which outlines relict sponge spicules, and chert recrystallized to spherulites that contain no carbonaceous matter and no relict sponge spicules (Sheldon, 1957). At several localities chert is restricted to interreef facies rich in siliceous sponge remains, from which it apparently formed. For example, an unusually large amount of biogenic silica is contained in the sooty, black, thin-bedded lime- stone of the Marble Falls Formation of Pennsylvanian age, Lampasas County, Texas, which has from 20 to 45 percent siliceous residue consisting largely of sponge spicules (Damon, 1943; Barnes, 1952). It can be followed laterally into spiculite that contains some 88 percent siliceous residue, is hard, and appears under the microscope to have been partly cemented by silica. This altered spiculite in places passes into chert that is interbedded with the limestone. Nearby reef rock is of very high purity (analysis 44, Part IV) and the small amount of chert it contains is the porous, friable kind that replaces fossils. In the post-Joliet formations of the Niagaran Series in northeastern Illinois, chert is confined principally to the biostromes, composed mainly of siliceous sponges, GEOCHEMISTRY OF SEDIMENTARY CARBONATES - II 29 that occur in the dolomites of the interreef facies (Lowenstam, 1942, 1948). The nearby dolomitized reefs are extremely pure (analyses 79 and 80, Part IV). Chert in zones now devoid of sponge remains apparently also was sponge-derived, for a gradation exists between spiculeless chert in such zones and nearby chert contain- ing completely preserved sponge skeletons, and the specialized fauna of the sponge biostromes also is found in the spiculeless chert zones but not in chert-free areas. Lowenstam found that well shaped chert nodules typically contained a complete siliceous sponge skeleton, whereas bedded cherts and incompletely chertified areas with poorly defined boundaries contained abundantly distributed flesh and tuft spicules that had served as centers for chertification. The relative rates of sedi- mentation, decomposition of soft parts, and reorganization of silica were presumably critical in determining whether recognizable sponge remains persisted or were dis- aggregated. Localized solution of isolated spicules and margins of already formed chert nodules has resulted in the silicification of fossils other than sponges. The remarkable preservation of morphologic details in chertified fossils and the poor de- tail in dolomitized and silicified fossils were taken by Lowenstam to indicate that chertification is pre-dolomitization in this area. Apparently assuming that the preservation of detail in silica replacements is relatively constant, he further con- cluded the silicified fossils must be replacements of already dolomitized ones. Sponge spicules are absent in the core of the Wabash Reef, Wabash, Indi- ana (Carozzi and Zadnik, 1959), very scarce in the adjoining reef detritus, and abundant in the interreef area. The relatively early formation of at least some chert is indicated by its occurrence as nodules in some of the desiccation polygons in the Hardy Creek Limestone of middle Ordovician age in southwest Virginia (Harris, 1958). The size and shape distribution of the polygons is distorted according to the size and shape of the contained nodules. An early age also is indicated by occurrences like those cited by Poor (1925) and Rubey (1952) in which chert pebbles of Mississippian age are found in the basal conglomerate of the Pennsylvanian in Illinois (see also Cayeux, 1941). Dolomite-free calcareous inclusions frequently found in cherts contained in dolomitized rocks indicate that chert formation usually precedes sec- ondary dolomitization (Carozzi, 1960). Silicification of exposed surfaces (Moore, 1955, Crystal River Formation of northwest Florida; Dott, 1958, Pennsylvanian limestones of northeast Nevada) and of the exposed portions of fossils (Howell, 1931; Lowenstam, 1948; Dott, 1958) generally is taken as evidence for continuing present-day redistribution of silica in carbonate rocks. However, Newell et al . (195 3) noted that frequently only parts of fossils are silicified, whetherthey are collected from the surface or well within large blocks, and suggested that a portion of such a partly silicified fossil which stands in relief on the weathered surface of a carbonate rock would naturally be the silici- fied portion . Available data on the solubility of "amorphous silica" and quartz and on the silica content of natural waters suggest that most natural waters are undersaturated with respect to "amorphous silica, " but that solutions supersaturated with regard to quartz are not rare and may persist in that condition for years (Siever, 1957; Kraus- kopf, 1956). Synthetic crystalline quartz has not been formed directly from solu- tion at anywhere near room temperature and pressure, and the mechanism of forma- tion of authigenic quartz crystals in limestones and elsewhere is not yet under- stood . The amount of silica dissolved in alkaline lakes with a pH above 9 increases markedly because of the formation of silicate ions, so that a subsequent pH drop 30 ILLINOIS STATE GEOLOGICAL SURVEY CIRCULAR 298 might be expected to result in inorganic precipitation of "amorphous silica" (Siever, 1957). The chert in the dolomite-rich, lacustrine Green River Formation of Eocene age in Wyoming, Colorado, and Utah might represent such a product. However, diatoms flourished in the Green River lake during at least part of its history, and they may have succeeded in keeping the silica concentration well below saturation. Argillaceous and Glauconitic Limestones Robbins and Keller (195 2) examined the less than 2\± fraction of insoluble residues from 271 limestones and dolomites and found illite to be the dominant clay mineral in both marine and nonmarine carbonate rocks. However, they used 6N HC1, which could have deteriorated a good deal of chlorite. Weaver (1958) con- cluded from several thousand clay mineral determinations that illite is the predom- inant clay mineral in carbonate rocks, as Grim et al . (1937) and Millot (1953) had said earlier. However, he also found clay suites commonly composed entirely of montmorillonite and of mixed-layer materials. Occurrences of chlorite and kaolinite in carbonate rocks are not uncommon, but these minerals are seldom dominant in the clay mineral suite. Weaver found no particular clay mineral restricted to a particular environment. In another study, Degens et al . (1958) found that the illite/kaolinite ratio of 18 samples of marine and freshwater limestones of Pennsylvanian age did not reflect differences in depositional environment. The relative importance of the original character of detrital clay minerals and of diagenetic changes in depositional environments is at present under intensive study by many workers. Myers et al . (1956) described a zone at the top of the reef limestone of the Pennsylvanian and early Permian Horsehoe Atoll of west Texas that lies under dark pyritiferous shale. The pores of the limestone have been filled with dark gray to black clay and, in a few places, with pyrite thought to have been precipitated from connate water originally contained in the shale. Alternating laminae of relatively pure limestone and more argillaceous lime- stone were reported by Brill (1956) in the Berriedale Limestone of Tasmania. Hadding (1958, p. 37-40) described a similar alternation in which the individual layers are about 100 mm thick. The small green pellets called glauconite are shown by X-ray examination to consist of a variety of layer-lattice materials (Burst, 1958a, b) in addition to the carefully defined glauconite of Gruner (1935) and Hendricks and Ross (1941). Burst believed that degraded layer-lattice silicates surrounded by local environments of reduction are nucleation centers to which Fe and K migrate, and he noted that glau- conite formation in fecal pellets and inside foraminiferal shells fulfills these con- ditions. The argillaceous minerals of Recent age in Gulf Coast fecal and cavity-fill material are considerably degraded. Hadding (1932) concluded from his study of Swedish glauconite rocks that glauconite formation is favored in agitated waters under conditions of limited dep- osition, an opinion strengthened by the glauconite deposition actually observed off southern California (Emery, 1960). Some mica-booklet type of glauconite occurs in the Senonian age limestones of Sweden, in addition to that which impregnates sponges, porous fossil fragments, and gastropod shells (Hadding, 1932). Herzog et al . (1958) mentioned a richly fossiliferous, pyritic limestone of lowermost Ordovician age from a quarry at Sten- brottet, Sweden, from which chips containing more than 50 percent glauconite in 1 to 2 mm pellets can be selected. Many of the Eocene marls and chalks of the Gulf Coast are glauconitic and are likely to contain phosphatic concretions and internal molds of mollusks as well (Monroe, 1941; Burst, 1958a, b) . The glauconitic chalks GEOCHEMISTRY OF SEDIMENTARY CARBONATES - II 31 of the Paris Basin all carry variable amounts of phosphate granules (Cayeux, 1935). A dense, nodular, black limestone described by Cayeux from the boundary zone be- tween the Upper and Lower Cretaceous at Interlaken, Switzerland, contains abun- dant quartz, glauconite, and phosphate. The lower twenty feet of the Colon Lime- stone in Venezuela (Smith, 1951) is glauconitic, asphaltic, and highly pyritiferous. Pfefferkorn and Urban (1957) noted that the "Grunsandstein" of Westphalia is actual- ly a limestone containing about 10 percent quartz and 10 percent glauconite. Tuffaceous Limestones The Tertiary limestones of Saipan include relatively pure reef and detrital bioclastic limestones, calcite-cemented conglomerates consisting of limestone pebbles and cobbles and material of volcanic origin, and rocks containing from 20 to 80 percent bioclastic carbonate fragments set in a matrix of tuffaceous material, quartz sand, and crystalline calcite (Cloud et al . , 1956; Schmidt et al . , 1957). Volcanic material typically makes up 3 to 5 percent of the Saipan limestones, but may reach 12 to 15 percent. When these rocks are exposed to weathering, recrys- tallization starts in the groundmass and the volcanic ash is altered to clay. Ladd and Hoffmeister (1945) found at Lau, Fiji, all gradations between lime- stone, limestone containing water-worn grains of volcanic detritus and clay derived from weathering of volcanic material, and pyroclastic rocks. Pyritic Limestones Although many argillaceous and bituminous limestones contain some pyrite, in other less common limestones it is a substantial constituent. The Top Jet Dogger of the Upper Lias succession in Yorkshire is a thin, platy, argillaceous limestone with coarsely crystalline lenses of calcite separated by irregular but laminated layers of pyritic shale. Hemingway (1951) considered the bed to be a diagenetically altered analog of the calcareous mud now forming in the western basin of the Black Sea (see Archanguelsky, 1927). The basal pyritic limestone bed of the Greenhorn Formation of- Upper Creta- ceous age in the Black Hills, South Dakota (Rubey, 1930; analysis 72, Part IV), contains, principally, Inoceramus shells and fine-grained calcite, pyrite (some 25 percent), and secondary gypsum and iron oxides probably derived from the pyrite. The pyrite is minutely crystalline, forms nodules and coatings, and is intimately as- sociated with unreplaced shell fragments and fine-grained CaC03. Rubey suggested that water escaping upward from immediately underlying black shale during compac- tion may have formed the pyrite. He noted that present-day oceanic blue muds con- tain up to 35 percent CaCO^ but are reducing in character and contain sulfides and H2S below the sediment-water interface (Murray and Renard, 1891; see also Murray and Irvine, 1895, concerning the processes involved), so that the calcite-pyrite association is not inconsistent. The Pumpherston Shell Bed and the Pumpherston Oil Shales of Lower Carbon- iferous age in Scotland contain large numbers (~1 x 10^/cm^) of pyrite grains 2 to 30 microns in size (Love, 1958). Microfossils, found only in pyrite grains, are obtained by treating a concentrate of the pyrite grains with nitric acid. The size distribution of pyrite grains virtually coincides with that of the microfossils, which are believed to be actinomycetes rather than bacteria. The even lamination of the oil shales indicates an anaerobic depositional environment free of bottom scaven- gers, and the carbonate cementation that has occurred suggests a late-diagenetic rise in pH . 32 ILLINOIS STATE GEOLOGICAL SURVEY CIRCULAR 298 Phosphatic Limestones The small percentages of phosphate found in most limestones are contributed by detrital apatite, by either replaced or originally phosphatic fossils, and by detri- tal phosphate grains and pebbles. The weakly phosphatic limestones are composi- tionally quite distinct from phosphorites, for highly phosphatic limestones are rare. This rarity is believed to result from the fact that carbonate can precipitate in the ocean (with its present composition) only above a pH of 7.8, whereas apatite is stable above a pH of 7 . (see discussion by Sheldon, 1959, p. 93). Carbonate- poor phosphorites thus presumably form at a lower pH than phosphatic carbonate rocks. Furthermore, the ratio of carbonate to apatite should be high when these two materials coprecipitate (Krumbein and Garrels, 1952). Fish remains and the shells of certain inarticulate brachiopods {Liniula sp., Discina sp., Glottidia sp.) are the most important sources of organically deposited phosphate, with lesser contributions from vertebrate bones, alcyonarians, crusta- ceans, and conodonts (see also table 1). C. W. Collinson (personal communication) estimated that there is a 50 percent chance of recovering 15 or more conodonts in any 1000 grams of marine limestone of Ordovician through Pennsylvanian age in the midwestern United States. Fifteen conodonts would correspond to about 1 ppm Ca 3 (P0 4 ) 2 . An original ecological control of phosphate content appears probable in some limestones. The number of conodonts is greater in sponge-crowded habitats of the interreef beds of Silurian age near Chicago than it is in sponge-poor areas (Lowen- stam, 1948), and Liniula are restricted to the sponge-crowded areas. Lowenstam suggested that putrification of inter-sponge passages by metabolism products of sponges may have kept the number of sessile benthos associates lower than the number of anchorage sites, allowing Liniula to move in. Willman (1943) stated that the PoOc content °f the nearby reefs is often so slight as to be undetectable in a 5-gram sample by conventional analytical methods. Phosphate in the Cambrian and Ordovician limestones of Sweden occurs almost always in beds that contain glauconite (Hadding, 1958). Hadding contrasted these limestones with others rich in organic matter and clay that usually do not contain phosphate. He supposed that relatively agitated water in the depositional environment had removed bituminous muds from the phosphatic-glauconitic lime- stone sediment. This concept has been substantiated by the recent oceanographic work of Emery (1960) and others (see discussion of carbonate sediments). Determinations of phosphorus (0.005 to 0.136 percent) and iron (0.15 to 3.22 percent) in some 380 argillaceous limestone samples from the Lower Malm of southern Germany (Seibold, 1955) show a pronounced negative correlation between P and CaC03 and between Fe and CaC03. There is a distinct positive correlation between P and Fe. The P/Fe ratio has a mean value of 0.046 but shifts progressive- ly to higher values as CaC03 content increases. Part of the iron is present as sulfide, but an iron content in the clay minerals and the presence of iron carbonate, silicate, or oxide are suggested to account for the remainder. Erosion remnants of the phosphatic, glauconitic chalk of Senonian age at Taplow, England, contain phosphate as bones, teeth, fish scales, and phosphatized coprolites and foraminiferal tests, and as a veneer coating nodules and rock bands (Willcox, 1952). The fossils are like those in normal chalk but much more abundant. Petrographic evidence argues against the deposit's being merely current- sorted phosphatic debris. A locally flourishing community of organisms is postulated in a nutrient-rich zone where the bottom currents were deflected upward by a sub- marine ridge, and where a trough provided tranquillity for phosphate accumulation. GEOCHEMISTRY OF SEDIMENTARY CARBONATES - II 33 Two thin, areally extensive beds in the limestone of the Delaware Forma- tion of Middle Devonian age in central Ohio contain abundant crinoid fragments and fish teeth and plates, often worn and rounded (Westgate and Fischer, 1933). An analysis of one of the beds gave 16.80 percent Ca^{PO^)2> 73.27 percent CaC03, 4.97 percent MgCOo, 2.46 percent iron oxides, and 2.14 percent SiO^, for a total of 99.61 percent (Orton, 1878). Both bone beds and phosphate pebble concentrations often occur as basal conglomerates on old erosion surfaces, and are accompanied by pyrite and Mn- oxide stain (Pettijohn, 1926). Phosphate pebbles occur in carbonate rocks, above and below the phosphate zone, near the Quadrant-Phosphoria boundary in southern Montana . Many are found in the Noix Oolite Member of the Edgewood Formation of Silurian age in Illinois (Rubey, 1930). Phosphate pebbles and grains occur scat- tered throughout the Middle Miocene Hawthorn Formation of Alachua County, Florida (Pirkle, 1957), which is a variable mixture of sand, clay, and carbonate, any one of which may predominate locally. The phosphate particles include replaced fossils and fragments of impure limestone, aggregated masses and concretionary oolites, and phosphate filling organic cavities, such as bryozoan chambers. A Cretaceous chalk near Mons and Liege in south Belgium contains numerous small phosphatic concretions "no larger than a mustard seed" that in many places form 10 percent of the rock (McCallie, 1896). The phosphatic limestone of the Bigby Formation of Tennessee contains worn and rounded fragments of small bryozoa and perhaps crinoid stems, in all stages of replacement by phosphate, set in fine- grained calcite (Whitlatch and Smith, 1940). Partial replacement of the fossils in a phosphate-rich environment before consolidation of the rock is indicated. In two areas the Bigby Limestone and part of the Leepers Limestone average 15 to 20 percent Ca3(PO^)2; darker streaks have 40 percent or more. Thin beds of dark, greenish gray, silty dolomite in the Plympton Formation of Permian age in west- central Utah contain as much as 20 percent collophane as minute pellets and inter- stitial fillings (Hose and Repenning, 1959). In cherty argillaceous limestone layers in the La Caja Formation of Jurassic age in Zacatecas, Mexico, containing 10 to 15 percent P2O5, the chert appears mainly to be replacing the phosphate (see also analysis 69, Part IV). This forma- tion shows marked variation in thickness and contains terrestrial reptile remains, conglomerate units, and, in some units, cup-shaped gastropods and abundant ammonites; benthonic mollusks are rather scarce. A muddy environment and not too great a depth of water are suggested (Rogers et al., 1956) . Unconsolidated lime sands and gravels in the northern Marshall Islands are overlain locally by a consolidated 5 to 15 cm bed containing from 2.7 to 14 percent P (Fosberg, 1957). X-ray diagrams show the presence of an apatite that both fills interstices and replaces CaC03. The phosphatic limestone occurs only under humus-producing Pisonia grandis trees, which are a favorable nesting site for sea birds. Finely divided calcium phosphate in the guano is dissolved by rain- water enriched by humic acids and then precipitated when the solution is neutral- ized by the limestone below. McConnell (1943) described replacement of bio- clastic limestone at Malpelo Island, Colombia, by phosphatic solutions derived from the leaching of superficial guano deposits. Hutchinson (1950) concluded after a broad survey of guano deposits that rainfall must be less than 1500 mm a year to allow a consolidated phosphatic zone to form in underlying rock or soil, and less than 1000 mm a year to permit the formation of a major deposit. 34 ILLINOIS STATE GEOLOGICAL SURVEY CIRCULAR 298 Carbonate Rocks Containing Organic Material The Avon Park Limestone of Eocene age in Citrus and Levy Counties, Flori- da, is a virtually unaltered sediment of unconsolidated calcareous mud with lamina- tions of carbonaceous matter, much of which appears to be branches of land plants (Fischer, 1949). Sediments like this are found today in shallow, protected waters around mangrove islands in Florida Bay. The rather common, dense, sparingly fossiliferous limestones rich in organic matter and often in pyrite are interpreted by Sloss (1947) as having formed at depth in areas where impeded circulation inhibited bacterial decay and was inhospitable to the existence of benthonic organisms. These rocks, according to Sloss, contain abundant chert but are rarely dolomitized. Pennsylvanian and Permian limestones of this type, such as the Bone Spring Limestone of the Permian age Leonard Series, occur interbedded with dark shales and silts in basinal off-reef areas in west Texas (King, 1942; Adams, 1951). The fauna in the Bone Spring Limestone is a character- istic pelagic one consisting mainly of ammonites, which could float at the surface and whose shells could float and drift after death. These basins were surrounded by epicontinental shelves furnishing but little sediment. Newell et al. (195 3) presented a monographic study of the Permian reef complex of Texas and New Mexico, with references to other fossil reefs in England, Greenland, and the Italian Tyrol that lie along the borders of euxinic basins. Some 95 percent of the black, pyritic, bituminous limestones of the Delaware Basin, such as the Bone Spring, are divided into light and dark laminae 0.1 to 3.0 mm thick. Some of the finer grained dark laminae contain up to 30 percent by volume of siliceous sponge spicules, many of which have been replaced by calcite . Spicule sorting is indicated by parallel alignment of spicules and by the fact that adjacent laminae show differences in mean spiculite size. Organic matter is tightly sealed in voids and is believed to have been emplaced before the matrix lost its connate water. The content of organic matter insoluble in organic solvents ranges from 0.10 to 1.76 percent. Newell et al. believed that fluids driven out of the basin sediments by cal- cite cementation played a role in the several marginward diagenetic effects observed. Macroscopic fossils toward the basin margin, near the outer edge of the reef talus, are replaced by silica. Chert is very frequently associated with spiculites and is localized along noses of primary fold structures. The preferential replacement by silica of fossils rather than mud, and of certain types of fossils rather than others, does not appear to be a simple function of porosity or of the original content of aragonite. Dolomitization is significant only toward the basin margins, and in some horizons there is a marked replacement of foraminifera and bryozoa in prefer- ence to the matrix. A flood of secondary calcite cement also occurs toward the basin margin. Another carefully studied suite of bituminous carbonate rocks is that con- tained in the Green River Formation, a series of lacustrine beds of Eocene age some 1500 to 2000 feet thick that today occupy several basins in Colorado, Wyoming, and Utah. Deltaic and fluvial sands and shales are followed basinward in turn (1) by a freshwater shore facies of reefs and algal limestones containing caddis fly larval cases, freshwater mollusks, and an abundance of ostracodes (Bradley, 1925); (2) by muddy limestones and limy mudstones; (3) by the so-called oil shale and its somewhat more saline facies. All of these rocks have high carbonate contents; Hunt et al. (1954) estimated that the limestones typically contain 82 percent car- bonate as calcite, the saline "oil shales" and the marlstones of the "oil shales" have 52 to 55 percent as dolomite (analyses 73, 74, Part IV), and the platy and so-called Mahogany oil shales, 42 to 43 percent as dolomite. GEOCHEMISTRY OF SEDIMENTARY CARBONATES - II 35 The marl stones of the Green River Formation are varved and contain light colored layers rich in microgranular carbonate and dark colored layers with abundant organic matter. Studies of modern varved sediments from Lake Zurich (Nipkow, 1927) and other localities described in Circular 297, the first of this series, indicated that the Green River varves probably are annual, with carbonate precipitated at the sur- face by summer warming and photosynthesis. The presence of perfectly preserved fossil fish in the varved sediments, contrasted with the broken and chewed-up bones in the nonvarved sediments, indicates that the former were deposited in a stagnant bottom zone free of scavengers or bottom feeders. The remarkable suite of authigenic minerals in the "oil shale, " described in Part III of this series, indicates that the lake waters at times must have been highly mineralized. The organic matter in these beds, which at places is as much as 80 percent, is relatively rich in fatty and waxy matter, high in C and H and low in O. Bradley (1925) believed that it was derived almost entirely from plankton, for peat bogs are inconspicuous around the shore and there are no remains of large plants. The summer temperature maximum indicated by the larger land plants, insects, and animals would have been high enough to support a bloom. "Oil shale" has been described from various other localities. That from Um Barek, Israel, is of Senonian age and contains 50 to 60 percent carbonates, mostly calcite, 15 to 20 percent silica and silicates, 1 to 4 percent P„0 , 2 to 7 percent R2O0 (mainly AUO3), 1 percent pyrite, 0.002 to 0.003 percent uranium, and from 4 to 25 percent total organic matter, mostly kerogen. The ultimate composition of the kerogen, estimated from total organic content and amount and elemental analysis of the assay products is C, 69.1; H, 8.6; N, 1.9; S, 11.0; O, 9 . 4 percent by differ- ence. The more calcareous beds (Stinkschiefer) of the Posidonia bituminous shales of Liassic age in Swabia, Germany, consist of alternate light and dark units 1 mm thick, and contain 70 to 86 percent CaCO~ and as much as 7 percent organic matter (Einsele and Mosebach, 1955; Wunnenberg, 1950). The occurrence of free bitumens in carbonate rocks has been recorded from widely separated localities. Petrographic descriptions of chemically analyzed Scottish carbonate rocks (Geol. Survey Great Britain, 1956) include numerous cases in which there is observable bituminous matter interstitial to carbonate grains or mixed with accessory clay. Henson (1950) mentioned two fossil reefs of the Lower, Middle, and Upper Cretaceous in the Middle East that are highly bituminous at their outcrops. Reefs in this area pass through transition beds of reef detritus, which are bituminous if porous, into basinal anhydrites and Globiierina facies that are in part bituminous . The organic material in Upper Jurassic bituminous limestones in the southern Jura region of eastern France consists principally of pseudoasphaltenes and resins and has high acidity and sulfur content (Gubler and Louis, 1956). The Seefelder marls of the Tyrol region are a sapropel facies with some layers containing up to 20 percent by weight of organic substances (Fischer, 1957). Swain (1958) gave detail- ed information about extracts, obtained by leaching of Middle Devonian limestones of the Mount Union area, Pennsylvania, with various solvents. Viscous to solid organic matter has been observed coating cavities of var- ious sorts in Mississippian and Pennsylvanian limestones of western Missouri (Searight, 1957), Niagaran age limestone at Monon, Indiana, and the South Canyon Creek Dolomite Member of the Pennsylvanian and Permian (?) Maroon Formation at Glenwood Springs, Colorado (Bass, 1950). Schoch (1918) gave partial chemical analyses of a porous coquina, the Anacacho Limestone of Uvalde County, Texas, and of a Comanchean age limestone of Burnet County, Texas, which were impregnated 36 ILLINOIS STATE GEOLOGICAL SURVEY CIRCULAR 298 with 14.00 and 10.30 percent bitumens, respectively. Anthraxolite, a carbonaceous material similar to anthracite, occurs disseminated or in vugs, fractures, and within quartz crystals in the Upper Cambrian and Lower Ordovician dolomites of the Mohawk Valley, New York (Dunn and Fisher, 1954). A typical ultimate analysis is C, 90.42 percent; H, 3.94 percent; O, 3.42 percent; N, 1.30 percent; S, 0.57 percent; and ash, 0.35 percent. The occurrence of asphalt in porous areas of the reefs of the Chicago area is common (Willman et al . , 1950). Table 4. -" Organic Content of Some Carbonate Rocks [After Hunt and Jamieson, 1956. The values of barrels per acre foot originally reported have been converted to ap- proximate values of parts per million by multiplying by the factor 41.7. ) Stratigraphic Hydrocarbons Asphalt Kerogen Age and unit (ppm) (ppm) (ppm) location Amsden (limestone) 8.3 8.3 290 Mississippian-Penn- sylvanian, Wyoming Smackover (limestone) 87. 83. 460 Jurassic, Arkansas Dundee (dolomitic lime- 87. 100. 1900 Devonian, Michigan stone) Greenfield (argillaceous 120. 150. 1200 Silurian, Michigan dolomite) Reed City (dolomite) 280. 170. 960 Devonian, Michigan Phosphoria (carbonaceous 290. 580. 2400 Permian, Wyoming dolomite) More finely distributed organic matter, however, has been found in carbon- ate rocks. Hunt and Jamieson (1956) found that organic matter and, in most cases, elemental sulfur could be extracted from nonreservoir rocks showing no evidence of oil under close microscopic examination, by micronizing the samples to a particle size of 15 microns and refluxing with selected mixtures of solvents. The results for carbonate rocks are given here as table 4, in which hydrocarbons are defined as that part of the total refluxed extract eluted from an activated alumina column with hep- tane and benzene, minus free sulfur. The nonhydrocarbons (asphalt) are the total extract minus hydrocarbons, sulfur, and ash. The organic matter insoluble in the reflux mixtures (kerogen) is estimated as the organic carbon in the extracted rock times an empirical factor, 1.22. The hydrocarbons are similar in physical and chem- ical properties to natural crude oils, differing only in that their initial boiling point is 400° to 500°F. The kerogen from the Madison Dolomite has a H/C ratio of 0.68, within the range for coals. Age and depth of burial do not seem to have affected the distribution of types of organic matter present. Further work by Forsman and Hunt (1958) showed that there are two kinds of kerogen, one coal-like and the other more like the material from oil shale, as well as gradations between these types (see table 5). These kerogens are found in marine shales and limestones of various ages. The kerogen type appears to be controlled by depositional environment and by sub- sequent metamorphism. Philippi (1957) found small amounts of hydrocarbons, generally ranging from 5 to 5000 ppm, in dense sediments, including marls and argillaceous limestones. He stated that the indigenous hydrocarbon content can be recognized by the pro- portionality between it and the total organic content and that apparently only a small part of the oil generated by the source beds is released to reservoirs. GEOCHEMISTRY OF SEDIMENTARY CARBONATES - II 37 o o rr. ■' ■" 0) X: +■> o CO ^ c * +> Q ^ U u m X U T) e ai 1 r-H •fH _ in = o fH > a> a> i-i ■•-» — i o 4-. XI 10 ro 3 -l >■ -* o .a o en o U f-\ M-f H) o +j 4-. +J O C nj -t-> ^ "5 ^ < ^ ^ ro ^-» -a ^ « 00 r— T \T> r~ O vO r- o 00 •sr •^ o r- co o CM 00 ro r- CM CN r- 00 in [- CM CO ■* o o o o o i-H o O o o 00 rH ■ H *"" *"■ •-H 1-1 CM _, r- CM lT> o in in o o ro f-H ^ O 00 a* a> CM N* CM .— i CM CO o *" 1 1-1 ^ 1-1 CM i CM rH 00 n . 00 CM r- vO v£> ■q- CM CO o O CM vO 00 CM CM vO • • • • • • 0> Q. t- CM 00 «— 1 o o O •^ en +J o 00 T3 • H t+H f-H (H Ol o ■(-> o fO f-H >- c 1 c j*: -h s to ro =J Ol rO +J M- to OJ ro c e Z) -fH IO (H C c (0 OJ c i c e .rH • rH O O -Q Cn ro u H-> c ^ to 0J a • M -rH a. Q. >- C ai E to -C IO 01 M 21 ^ — 1 o a> e Q. J Q- — s ro o o Ol IO H-> =1 > (1) CT> fH ai +J • H • rH ai • fH OJ 0) ro ^h ■o • fH O 3 X ^-* c T n c I/) JS .-j CO C OJ 10 0) - c U C -H O 3 to 01 Q f-H (h •#H 0) E fH £ 0J O 10 ■H > UT +J C o -2 (0 O -H O i-H f-H tO o < ro s E E • rH OJ •H ^H OJ •H ■iH ro +J E ro 4-> tH O to oj E o ro (H o CD •H s > to f-H .rH tn E -fH to ro (0 to CT> » c +j —i •-< £ •» >- JZ 01 >- p-H -J 0J 10 •rH > irt o a. o U c 01 (h to • fH o — 1 H-> (D h 3 0) > s 10 • iH ■r-i H-f at • rH -H Q. E to E ro D -t-> to •fH C S >- c s >- O -rH .^i »*v ^ •H o CO • H CD 4h o > •« >.a: M (0 01 ro 01 -X. OJ (1) ^; Q to — 1 (H J - 0) 01 (h o +J •H OJ c 0J (H •rH c ►** u H-> O w o to ^ri - 0J C OJ o 3 • fH 0J =) C -H M o ro O to ro +-> ro in (U 0) o> o o c H-> f-H • c +> o to 0J IO M C 10 H-> 4-> +> M TD E c o o Qj O 0J ro > o c O 01 o c to to ■ r^ to 3h « O Hh o o; u .m OJ o £ u O a (H 0J ^J 01 fH +J ■M 0J • H -rH XI T3 J J M H l-H +J s xi >-) (U UJ to o LU ro r- a OJ to ~^. >.co l^ ^ T3 a 1-4 — — CJ +J U r 3 n M T) £J >- (H Q.TI to U o i/i N ro •rH o •rH tn »H c H-> m o to "T •rH i — | (H c o o - > -p rH Ol c ro u ■ r^ c »H u l/l a. +J Ol c -^ 0J o (H (H o 01 3 -X 4-. n CJ •• 38 ILLINOIS STATE GEOLOGICAL SURVEY CIRCULAR 298 Lucas (1952) isolated from fetid limestones a volatile, water-soluble sub- stance containing phosphate and ammonia, but no sulfur, which he believed to be similar to the phosphenates . The principal carbohydrate that has been isolated from fossils is cellulose (Abelson, 1959b). Abderhalden and Heyns (1933) also identified chitin in a fossil coleopteron from the Middle Eocene. It appears that relatively impermeable bio- genic carbonate rocks that have had a sufficiently mild thermal history may still contain measurable amounts of the amino acids originally present in skeletal ma- terials. The various amino acids, following breakdown of the protein structure, undergo thermal degradation even at earth-surface temperatures. The process is accelerated by exposure to bacterial action, water, oxygen, and the aragonite-to- calcite transformation or other recrystallization. Conway and Libby (1958) measured the half-life for decarboxylation of alanine, one of the thermally most stable amino acids, at temperatures down to 373 °K, using radioactive labeling and low-level counting techniques. The indicated half-life at room temperature is about 10 10 years. Abelson (1954) obtained a value of about 2 x 1()9 years by extrapolation of the decomposition rate of alanine in aqueous solutions at a series of higher temperatures. If oxygen is present, carboxyl carbon is released at a rate corresponding to a half-life of about 2x10^ years at the same temperature . The thermally more stable amino acids have been found in suitably preserved fossils as old as Ordovician. An Ordovician brachiopod, Plaesiomys sub quadrat a , from Wayne sville, Ohio, had 0.005 percent amino acids, and a plate from the armor of the Devonian fish Dinichthys terellihad 0.03 percent (Abelson, 1957). Although there are amino acids in ground water and it has been shown by Abelson that aspartic and glutamic acids in particular are adsorbed by CaC03, differences in amino acid content of different fossils from the same locality indicate that adsorption is prob- ably only a minor contributing factor. The organic matter, found in some stylolites mixed with clay (analysis 75, Part IV) has been considered a concentrate of material originally disseminated in the limestone, in view of the difficulty of introducing it later into such an impene- trable host (Myers et al., 1956). Intimately distributed organic matter is expelled into intercrystalline spaces during calcium carbonate recrystallization and during dolomitization. Contacts be- tween dolomitic rock and embedded algal nodules in the Lockport Dolomite of Silur- ian age in Orleans County, New York, are marked by a zone of dolomite grains two or three times the normal size, as well as by a band of organic matter expelled during crystallization of the large grains (Cannon, 1955). Organic matter in dolo- mitic rocks of the Phosphoria Formation in northwestern Wyoming is concentrated into inclusions in the dolomite grains and crusts around these grains (Sheldon, 1957). Folk (1958) observed 0.002-inch rims of organic matter around dolomite crystals from rocks of Ordovician age in Pennsylvania. During the dolomitization of the Metaline Limestone in northeastern Washington, the organic coloring matter was expelled and, in places, it accumulated in vugs as rounded globules of anthracite composition less than half an inch in diameter (Park, 1938). Some dolomites have light and dark bands, generally less than half an inch thick. Analyses by R. C. Wells of fine-grained black dolomites versus relatively coarser grained, lighter colored dolomites are almost identical except for a small amount of organic matter in the darker bands. Rejection of bituminous and argillaceous material during the secondary growth of macroscopic calcite crystals in the Swedish anthraconites (bituminous argillaceous limestones) has resulted in the interstitial concentration of these materials (Hadding, 1958). GEOCHEMISTRY OF SEDIMENTARY CARBONATES - II 39 Oolites enclosed in a layer of dolomitic calcilutite described by Wherry (1916) have light upper portions and dark lower portions rich in organic matter, with nuclei occurring toward the bottom. Post-lithification solution of the aragonite of the oolites apparently left a solution residue at the bottom of the voids, which were sub- sequently filled by fine-grained secondary dolomite. Chemical analyses show that the organic matter in siliceous Cretaceous limestones from two localities in the French Pyrenees was not expelled or destroyed during metamorphism of the limestones to hornstones. Small granules of organic matter are enclosed in wernerite and diopside crystals. Louis and Ravier (1952) found 1.43 percent organic carbon and 0.036 percent total nitrogen in a hornstone, compared with 0.98 percent organic matter and 0.065 percent total nitrogen in the unmetamorphosed equivalent. Hunt (1953) attempted to correlate differences in crude oil types in Wyoming with differences in depositional environment and thus of organisms contributing or- ganic compounds to the sediment, but the problem in that area may be complicated by migration, by differences in depth of burial, and by other post-depositional changes. More positive results were obtained for the lacustrine carbonate rocks of Eocene age in the Uinta Basin. The change in mineralogy of the basinward facies in the Uinta Basin from calcite through dolomite to the sodium carbonates, proceed- ing through the increasingly younger Wasatch, Green River, and Uinta Groups, was taken by Hunt et al. (1954) to indicate a progressive change in salinity. The pro- gressive change in hydrocarbons in these beds, from ozocerite to albertite to gilson- ite to wurtzilite, represents an increase in N and S and in condensed ring structures as opposed to chains. The parallel changes in the composition of organic and in- organic materials indicate that the hydrocarbons have not migrated appreciably and that progressive changes in the composition of the lake, and perhaps of the kinds of organisms inhabiting it, are reflected in the hydrocarbon compositions now ob- served, in spite of whatever post-depositional alteration may have occurred. The median organic carbon content of 461 limestones from the United States examined by Trask and Patnode (1942) was 0.49 percent, compared with 0.75 percent for 2245 clastic rocks. The median C/N value for 492 limestones was 19.3, and for 2105 clastic rocks was 15.5. Readily observable changes take place in the amount of organic matter and FeO in limestones during weathering. In eastern South Dakota, black unweathered chalk of the Niobrara Formation analyzed 10 to 11-j percent volatile matter, ex- cluding CO2, and about 3 percent FeO. For the weathered white equivalent, the comparable figures are about 1.5 percent and about 1 percent (Rothrock, 1931). Partial analyses by W. A. Noyes (in Loughlin, 1930) of several samples of the well known oolitic limestone at different quarries near Bedford, Indiana, gave for the unweathered gray material 0.24, 0.21, and 0.22 percent organic matter; 0.067, 0.063, and 0.055 percent FeO; and 0.196, 0.044, and 0.089 percent Fe 2 3 . For buff weathered equivalents at the same localities, the comparable values were 0.12, 0.11, and 0.13 percent organic matter; 0.050, 0.055, and 0.050 percent FeO; and 0.126, 0.150, and 0.119 percent Fe 2 3 . Feldspathic Carbonate Rocks Tester and Atwater (1934) found authigenic feldspar throughout all of the Paleozoic, in shell-rich limestones as well as in recrystallized and dolomitized rocks. Adularia, albite, and microcline were observed in various combinations, either completely authigenic without cores, or overgrown upon detrital feldspar 40 ILLINOIS STATE GEOLOGICAL SURVEY CIRCULAR 298 cores of the same or different minerals. Tiny secondary crystals of both albite and potash feldspar were described by Trumpy (1916) from even younger rocks, the Ter- tiary (Flysch) limestones of the western Rhatikon, Switzerland. Authigenic feldspars from 40 localities, ranging in age from Precambrian to Triassic, were studied by Baskin (1956), who found that these feldspars were ran- domly distributed irrespective of fractures (see also Honess and Jeffries, 1940) and rarely exceeded 2 percent of the rock. All carbonate rocks containing them had been partially or completely recrystallized or dolomitized. Authigenic Na-feldspars were found only in carbonate rocks (see also Flichtbauer, 1950), rarely had detrital cores, unlike the K-feldspars, and were all low-temperature albite except for crystals structurally intermediate between albite and analbite, which were found at one lo- cality. The majority of the potash feldspars were monoclinic . Authigenic feldspars have idiomorphic crystal shapes, are nonperthitic, and show fourling twinning never observed in nonauthigenic microcline or in albite crys- tals formed at higher temperatures than those at which authigenic feldspars crystal- lize. They are remarkably pure, as would be expectable from the K-feldspar — Na- feldspar immiscibility gap at low temperatures. Authigenic potash feldspars rarely contain more than 0.3 percent Na20 (2 mol percent albite), and the K2O and CaO contents of Na-feldspars rarely exceed 0.4 percent (3 mol percent K-feldspar) (see also Honess and Jeffries, 1940) and 0.2 percent (1 mol percent anorthite), respec- tively. These feldspar crystals cut single crystal outlines of the carbonates and contain carbonate inclusions, usually rounded. At least two localities have been reported where authigenic potash feldspar is a major constituent of carbonate rocks. The so-called Kristalltuff of Monte San Giorgio (Tessin) contains about 40 percent feldspar and 60 percent limestone (Flicht- bauer, 1950). Daly (1917) described an apparently unmetamorphosed Precambrian dolomite from theWaterton Formation of Alberta that contained about 40 percent by volume of disseminated, glassy, clear crystals (analysis 85, Part IV). Fuchtbauer (1956) noted a correlation between the amount of authigenic K- feldspar in the Muschelkalk of GOttingen and the amount of illite that is available to furnish K^O for the reaction. There is no correlation between amount of clay minerals and amount of authigenic albite, which Fuchtbauer (1950) believed formed before complete lithification at a time when sea water was still available to furnish sodium. Carozzi (1953) and Topkaya (1950) described similar relations. Arkosic limestone of the Baum Member of the Paluxy Formation of early Cretaceous age in southern Oklahoma consists of soft, fine-grained limestone cementing the quartz and feldspar that overlie Precambrian granite (Wayland, 1954). SEDIMENTARY SIDERITES AND FERROAN DOLOMITES A number of relatively thin but laterally extensive Pennsylvanian limestones have been observed to contain ferroan dolomite in samples taken over large areas. Siever and Glass (1957) observed this phenomenon in the Cutler (Piasa), Galum, and Bankston Fork Limestones of the Illinois Basin in Illinois, using a differential thermal analysis technique by which the ferrous iron substitution in the dolomite could be determined as less than or greater than about l\ mol percent. No correla- tion with depositional facies could be established. Miles (1958) made some 350 X-ray diffraction runs to study further the car- bonate composition of the Bankston Fork, which is a thin, argillaceous, relatively discontinuous, marine limestone of Pennsylvanian age in southern Illinois. The ferroan dolomite in most cases makes up from 20 to 60 percent of the rock, calcite is responsible for 50 to 70 percent, and siderite is relatively uncommon, constituting GEOCHEMISTRY OF SEDIMENTARY CARBONATES - II 41 only 10 to 20 percent of portions of the cores examined. The siderite is present as disseminated crystals, spherulitic masses 10 to 20 microns in diameter, and patches of very fine mosaic. Thin beds and thin nodular zones rich in siderite are in general not uncommon in Carboniferous rocks (Lowe, 1914; Edwards and Stubblefield, 1948; Crane, 1912; Singewald, 1909, 1911; Scheere, 1955). Hunt et al. (1954) found minor amounts of siderite in the bituminous lacustrine beds of the Green River Formation in the Uinta Basin, and commented that the dolomites of this formation are ferroan. Rubey (1930) found abundant iron- stained concretions and thin beds of siderite in the Pierre Shale of late Cretaceous age in the Black Hills region. He also described (1952) nodular sideritic beds a few inches thick interlaminated with calcareous shales in the lower and middle portions of the Maquoketa Formation of Ordovician age in Illinois. Re- fractive index determinations show that the dolomite in a number of analyzed Scot- tish carbonate rocks is ferroan (Geol. Survey Great Britain, 1956). Spherules of kaolinite about 0.5 mm in diameter and containing small lenticular apatite crystals occur in a microgranular siderite rock described by van Tassel (1955). Lenticular and interbedded masses of siderite in irregularly shaped bodies of bauxite in the Eocene age Ackerman Formation are thought (Burchard, 1924) to have formed in a peat swamp environment (see also Mead, 1915). The iron carbonate ore described by Burchard (1915) in northeast Texas, apparently derived from glauco- nite in Eocene rocks, is one of a number of Tertiary glauconite- siderite ores on the Gulf Coast. Mansfield (1922) described a 6-foot bed of hardpan, composed largely of iron carbonate and containing scattered grains of glauconite, from the Hornerstown Formation of Upper Cretaceous age in New Jersey. Rolshausen (1934) encountered a zone of siderite with minor impurities in the caprock at Carlos Dome, Grimes County, Texas (analysis 93, Part IV), associated with quartz, sphalerite, and galena . The Mesozoic sedimentary iron ores of England and Western Europe include a number of unusual rock types made up of varying proportions of chamosite, calcite, siderite, and iron oxides formed by oxidation of the chamosite and siderite (analyses 94-96, Part IV). Well defined beds in which oolites and shell fragments are heavily pyritized also occur there. The existence of current bedding, pebble beds, oolites, and a normal shallow-water marine fauna, together with lateral transitions to sand- stone, indicate marine deposition (the Yorkshire Lias, for example) in a basin or lagoon of limited extent (Rastall and Hemingway, 1940; Hemingway, 1951; Rastall and Hemingway, 1941; see, more generally, Hallimond, 1925; Cayeux, 1909, 1922). The chamosite in such sedimentary ironstones has been shown to be a member of the kaolin group having a virtually trioctahedral structure, typically with 2.9 octahedral sites out of 3 occupied by Fe, Mg, or Al (see references in Youell, 1958). The siderite widely distributed in Devonian rocks in the Bashkirian A.S.S.R. (Florenskij and Bal'shina, 1948) also is mixed with chamosite oolites, in addition to detrital minerals and lesser amounts of pyrite, marcasite, and sphalerite. Rocks containing from 50 to as much as 70 percent siderite have been described (analysis 92, Part IV). The preservation of fine bedding and slump structures, constancy of chem- ical composition, and presence of fragments of each rock type in intraformational breccias and clastic dikes argues for a primary rather than a replacement origin for the interlaminated chert, siderite, and other rock types in the Precambrian of the Iron River District, Michigan (James, 1951). The rocks have been intensely deformed structurally but only slightly metamorphosed. 42 ILLINOIS STATE GEOLOGICAL SURVEY CIRCULAR 298 SEDIMENTARY MAGNESIUM CARBONATE DEPOSITS Longwell (1928) and Rubey and Callaghan (1936) described magnesium car- bonate beds of limited lateral extent in the almost fossil-free Horse Springs Forma- tion of Miocene age in the Muddy Mountains of Nevada. The 1- to 2-micron particles of magnesium carbonate make up a dense, white rock that is interbedded with almost indistinguishable dense, clayey dolomite, and is associated in the formation with beds of gypsum and other salines, volcanic ash, colemanite-bearing clays, and argillaceous sandstones. The exceedingly fine grain size of the carbonates, excel- lent preservation of all details of depositional bedding, concordance and diverse composition of successive beds, and uniform composition of individual beds traced laterally argue for a sedimentary origin in an inland lake enriched in Mg. Similar magnesium carbonate beds from Bissel, California (Gale, 1914; Rubey and Callaghan, 1936) are limited to certain basinal areas and associated with thin, hard beds of dolomite and dark, silty clay. Small surface deposits up to 18 acres in extent at Atlin, British Columbia (Gwillim, 1899; Young, 1916; Bain, 1924), and Bannock County, Idaho (Yale and Stone, 1923), are made up of white, powdery magnesium carbonate. Bedded magnesium carbonate at Needles, California, makes up a thin lens enclosed in Tertiary volcanic and sedimentary rocks. These Tertiary rocks include beds of dolomite, some of which contain uniform, continuous, cherty layers up to a tenth of an inch thick (Vitaliano, 1950). These several materials are surely sedimentary magnesium carbonates (anal- yses 22, 23, 85, Part IV), although the possible effect of associated volcanic ac- tivity on the composition of the waters in some of the lakes involved is not known and the mineralogy remains to be worked out. The general failure of (MgO + CaO) to balance CO2 in some of these analyses has led, depending upon the amount of water present, to two assumptions: the magnesium carbonate is hydromagnesite, or hydrous magnesium silicate is present. Magnesite, identified by partial chemical analyses, indices of refraction, and a positive reaction to the diphenylcarbazide test reaction, occurs as micro- layers and isolated well defined masses in a number of Lower Permian anhydrites of the Volga region (Frolova, 1955). The highest concentrations of magnesite are restricted to anhydrite rocks containing chlorides. A thin-bedded, anhydrite-bear- ing magnesite horizon having regional distribution contains up to 75 wt percent magnesite, as computed by Frolova from the analytical data shown as analysis 88 in Part IV. The rock described by Hartwig (1955) from the Rhon region of West Germany (analysis 91, Part IV) is, on the basis of chemical analysis and X-ray diffraction data, about one-third halite and two-thirds magnesite, the latter apparent- ly containing as much as 15 mol percent FeC03 in solid solution. Brown layers a few millimeters thick and consisting largely of magnesite are frequent in the Upper Evaporite Bed, Eksdale No. 2 boring, east Yorkshire (Stewart, 1951b). A carbonate, identified as magnesite by its refractive indices and difficult solubility in acid, occurs in anhydrite and polyhalite of the Permian potash field of New Mexico and Texas (Schaller and Henderson, 1932) as disseminated crystals and layers up to an inch thick. All gradations between almost pure magnesite and almost pure clay are present. Clear, glassy, authigenic crystals of magnesite several millimeters long have been described from a dolomitic outcrop sample of the Eocene age Green River Formation in Utah (Milton and Fahey, 1960). Magnesite crystals 1 by 5 mm in size, often stained with petroleum or asphalt, were described by Lonsdale (1930) from a cherty dolomitic limestone of Permian age encountered in drill cores. An analysis of the crystals by P. J. A. Zeller gave MgO, 47.24; CaO, 1.47; FeO, 1.67; CO2, 49.49; total, 99 . 87 percent . GEOCHEMISTRY OF SEDIMENTARY CARBONATES - II 43 Beds of magnesite several feet thick, interbedded with dolomitic shale, chert, and tillite in the Lower Adelaide System of Precambrian age, are consistent in thickness and composition over many miles of lateral distance in South Australia (King, 1956). Aside from variable amounts of chert and minor quartz veins, impurities in a typical case might be MnO, 0.06 percent; Fe^O.,, 0.8 percent; Al„Oo, 0.9 per- cent; CaO, less than 1 percent. There is disseminated talc in some samples. Well crystallized magnesite from the Pyrenees (Gomez de Llarena, 1950) is now interbedded with limestone, dolomite, schist, and graywacke, but ripple- marked surfaces are preserved and an original lacustrine precipitation of magnesium carbonate is postulated. Donath (1957) described bitumen-containing magnesite of the Bela Stena deposit, Serbia, which is interbedded with sandstones, marls, and clays in a faulted basin. The magnesite at Ajani is associated with impure dolomite in an essentially undisturbed young Tertiary sedimentary succession. There is serpentine in both regions from which Mg-rich solutions in the depositional basins could have been derived. SEDIMENTARY MANGANESE CARBONATE DEPOSITS The examples of sedimentary manganese carbonate that can be cited all occur in beds that are now somewhat metamorphosed, but the authors who have described the deposits do not believe that the metamorphism was severe enough to mask the fundamentally sedimentary nature of the MnCO~ formation. The MnCOo-rich beds in the nonfossiliferous Lower Cambrian of North Wales (Mohr, 1956) are intercalated in blue-gray shales and have been subjected to low- grade metamorphism. They typically contain subordinate quartz and 50 to 55 percent spessartite, believed to have formed by reaction of clay and siliceous material with fine-grained MnCOo (analysis 97, Part IV). On the Avalon Peninsula of southeast Newfoundland, interbeds of mangan- iferous carbonate, manganese oxide, and associated barite, hematite, and calcium phosphate are found in Cambrian shales (analysis 98, Part IV) (Dale, 1915; Hanson, 1956). Similar thin layers are found in southeast New Brunswick, near Woodstock, interbanded with sharply folded slate, clastic sedimentary rocks, and hematite of Silurian age (Anderson, 1954). Thinly laminated beds of Silurian (?) age including argillite, slate, hematitic ironstone, and siliceous carbonate rocks containing ferrous chlorite, ferroan rhodo- chrosite, and some calcic manganiferous carbonate are found in a belt some 65 miles long in Aroostook County, Maine (Crittenden, 1956; White, 1943). Radugin (1940) described a bed in the Cambrian section on the Usa River in the Kuznetsk Alatau (Siberia) that is a white to gray, fine-grained, rhodochrosite marble containing 32 to 34 percent Mn. Crushed diabase dikes and massive porphyrite are present in the section. Illinois State Geological Survey Circular 298 43 p., 6 figs., 4 tables, 1960 nnmni CIRCULAR 298 ILLINOIS STATE GEOLOGICAL SURVEY URBANA