3d. On the other hand, it is pretty evident that in Labrador and Newfoundland all the mountains were completely covered with glacial ice, which extended far out over the bordering continental plateau. But this was at that time probably elevated above the sea-level, so that it is doubtful if the ice ever extended far into the sea. The facts point to considerable preglacial elevations of land followed in Labrador, at least, by a period of extensive depression below the present level, and subsequent partial elevation. 4th. The freshness of the glacial striæ in exposed places and the small amount of modification which has taken place in the topography since the retreat of the ice sustains the abundant. evidence elsewhere found of the recent date of the glacial period; while the indications of recent changes of level point to terrestrial rather than astronomical causes to account for the vicissitudes of the glacial period. ART. VIII.-On the Recurrence of Devonian Fossils in strata of Carboniferous Age; by HENRY S. WILLIAMS. [Read before the Geological Society of America at Baltimore, December, 1894.] THE fossils which form the subject of the following remarks were brought to my notice by Dr. J. C. Branner during the progress of the Geological Survey of Arkansas. They were collected by various members of the survey from a dark colored limestone and associated calcareous shales at Spring Creek, a few miles west of Batesville. The first batch of fos sils sent me were in a rotten-stone, originally a sandy limestone, and seemed to be identical with the Leiorhynchus quadricostatum of the Devonian rocks of New York, which led to my reporting them to be of Devonian age. Afterwards fossils from the same ledge were sent which were reported in my correspondence to be of undoubted Carboniferous age. As the presence of Devonian rocks was looked for but had not been definitely proven the find was of considerable interest, but the confusion in my identifications led naturally to suspicion of either mixing of the evidence or error in the identifications. McChesney had previously described a few fossils as com ing from dark shales near Batesville, Ark., and probably of Hamilton (Devonian) age (Nucula Vaseyana, Nucula? rectangula and Pleurotomaria nodomarginata). And the black shales met with in the same part of the State and farther west, have *Descriptions of fossils, Chicago Acad. Sci., vol. i, pp. 37, 40 and 47. been supposed to be equivalent to the "black shales" of Tennessee and of Devonian age. More fossils were collected and the Arkansas geologists examined the locality with special care but with the result of confirming the singleness of the horizon from which the fossils came and the certainty of the duplicity of the testimony of the fossils themselves. Finally, Drs. Branner, Penrose and I went together and examined the locality with special care and sent in a collector, Mr. Weller, to make full collections of the fossils of the neighborhood, and the materials are now being elaborated for a full report of the fauna. On account of the importance of the facts this preliminary announcement is made. The geological age of the Spring Creek limestone is established to be younger than the Batesville sandstone and older. than the Boone chert, of the Arkansas survey nomenclature, which makes it equivalent to the Warsaw or St. Louis limestone of Missouri and the Mississippi Valley sections in general. Three kinds of evidence confirm this determination: The stratigraphy of the immediate neighborhood of Spring Creek, and second the correlation of the fauna with faunas of the same general region of higher and lower horizons, and third the comparison of the species of the fauna with those of a different geological province whose age is established on independent evidence. The stratigraphical evidence is as follows: The locality is in the northern part of Arkansas, geologically on the southern slope of the Ozark uplift, which centers in southeastern Missouri, where the upper paleozoic terranes lie with a general dip southward and southwestward, with thin upper edges graded toward the north so that outcrops are of older and older rocks on passing northeastward from Spring Creek, a point a mile or so west of Batesville. There is a fault near the point where the fossils come from running northeasterly, the southeasterly mass has fallen below the northwesterly mass. The Spring Creek limestone is on the northwest side of the fault hence the occurrence of the Devonian types cannot be explained as having been caught in a fault, since the other side of the faulted rock has been dragged down, leaving more recent and not older rocks at the same horizon on the opposite side. The strata underlying the Spring Creek limestone was shown to be the Boone chert (= Keokuk-Burlington of Missouri, Illinois, Iowa, etc.). The Batesville sandstone above it contains a fauna closely like the fauna of the St. Louis limestones in some of its species. At Mountain view some of the same species are reported from a similar stratigraphical position. In its more western exposures, according to the interpretation of the Arkansas geologists, the same interval is occupied by the Fayett ville shales. There is common agreement on the part of all the geologists of the survey who have studied the region that the stratigraphical position of the Spring Creek limestone is between the Batesville sandstone (= Chester-St. Louis horizon) and the Boone chert (= Keokuk-Burlington horizon). The second method of determination is by identification of the fauna with other faunas of known age. The general Carboniferous age is clearly indicated by the presence of Productus of the types of cora and semireticulatus, of Spirifers of the S. bisulcatus type of the Carboniferous limestone of England and elsewhere, and further by the identity of several of the species with those in the neighboring formation containing only faunas of the upper formations of the Mississippian series ("Subcarboniferous"). The species are not in general strictly identical with species of any of the typical divisions of the Mississippi Valley Carboniferous, and it is necessary to use the third method of correlation to reach greater precision than a general correlation with faunas of neo Carboniferous age. According to the third method we compare the species with faunas of other regions whose age is determined, and this reveals some of the more interesting features of the case. Upon making close identification of the fauna it is found that one of the most characteristic and abundant species in the fauna of the Spring Creek limestone is identical with Walcott's Rhynchonella Eurekensis of Lower Carboniferous limestone of Secret cañon road Cañon and the Cañon of Pinto Park of Eureka District, Nevada. This is a unique species, no Rhynchonella like it is reported from North America, but it is represented by several European species from Devonian horizons, or more ancient (cf. R. Dumonti Gosselet, Devonian of Ardennes; R. princeps var. Barrande; R. livonica v. Buch, Wenukoff, Tab. V, fig. 3. Russian Devonian). resembles also the striated varieties of the recent Rhy. psittacea. Pleurotomaria nodomarginata McChesney, referred to above, or a very closely allied form is among the Spring Creek species. Walcott identified the same species in the Secret Cañon road Cañon locality in Nevada. It When the fauna is compared with the fauna of Eureka district three-fifths of the genera are the same, and many of the species closely allied. The same close generic resemblance is seen on comparing the former with the Baird shale described by J. P. Smitht from the U. S. Fisheries in Shasta County, California, and both these western faunas are peculiar in holding species which are of markedly Devonian type, though the * Monograph, Paleontology of the Eureka District, p. 223. Journal of Geology, vol, ii, Sept.-Oct., 1894, p. 594. majority of the species are so typically Carboniferous as to leave no doubt of their Carboniferous age. In describing the Lower Carboniferous fauna of the Eureka district Mr. Walcott says: There is also a certain commingling of upper Devonian species with the lower Carboniferous fauna. We find Discina Newberryi, Macrodon Hamiltonia, Grammysia Hannibalensis, G. arcuata, Sanguinolites Aeolus, Pleurotomaria nodomarginata associated with common Carboniferous species.-Pal. Eureka District, p. 8. The same commingling of species is noted by J. P. Smith.* He states that 29 of the 84 species of the Baird shales of Shasta County, California, are identical with the forms described by Walcott from the lower Carboniferous of the Eureka district, and in this Baird shale fauna are such Carboniferous species as Productus Burlingtonensis, P. giganteus, P. Nebrascensis, P. punctatus and P. semireticulatus, Spirifer lineatus and S. striutus. But in the California fauna 15 species are found which are known Devonian fossils of Eastern America. In my manuscript report on the Arkansas fauna to Dr. Branner, I had suggested the relationship between the Spring Creek and Eureka faunas, and Mr. Smith who had read it before writing his paper on the Shasta faunas noticed the confirmation his species furnished of this interpretation. The Shasta and Eureka faunas find an unmistakable representative in the fauna of the Spring Creek limestones of Arkansas, and the fact that they differ from the ordinary Carboniferous faunas of the Mississippi Valley in the particulars which associate them with these faunas west of the Rocky Mountains is a strong argument for the theory that this Spring Creek fauna. migrated into the seas over Arkansas from the west, was there only temporarily and was soon withdrawn or destroyed, leaving only this solitary record of its existence in the series of the Mississippi Valley. Thus all the evidence in hand points to the Carboniferous age of the Spring Creek limestone, and there are sufficient reasons for referring it to the lower third of the Carboniferous (the Mississippian); and its correlation with an horizon about equivalent to the Warsaw, St. Louis, or the Spergen Hill formations is strongly suggested. As a general fauna, this Arkansas fauna is more closely allied to those of Eureka District, Nevada and of Shasta County, California, than to any other fauna in the Mississippi Valley or farther east. Both of these western faunas, although presenting species of Devonian type commingled with the Carboniferous species are separated by considerable thickness of strata from * Journal Geol., vol. ii, p. 597. AM. JOUR. SCI.-THIRD SERIES, VOL. XLIX, No. 290.-FEB., 1895. the latest Devonian horizon of the local section; by three thousand feet of conglomerate in the Eureka District, and by some, at present, unknown amount of sediments in Shasta County, California. The second point requiring verification is the presence of actual Devonian species in this limestone of Carboniferous age. There are two species both of which are represented by numerous specimens in the same strata with the Carboniferous species: they are Leiorhynchus quadricostatum Vanuxem, and Productella lachrymosa, varieties stigmata, onusta, etc. Hall. The first of these species was reported by Meek from the White Pine Mountains, Eureka District (= White Pine shale of Walcott) in 1877.* The age of these shales, was, in the same report (p. 201), referred to the Devonian by Hall and Whitfield on the evidence of this species and a Lunulicardium (L. fragosa, Meek sp.) and an Avicula (A. equilatera), while the beds immediately above were called Carboniferous without hesitation. Walcott noted the mixture of the Devonian and Carboniferous species in this White Pine shale, but concluded from study of the section that the beds in question covered a fauna. uniting the two systems but of pre-Carboniferous age (p. 6). The second species in some of its varieties is also reported by Walcott from Upper Devonian limestones of the Eureka District associated with many typical eastern Devonian species. The Rhynchonella Eurekensis Walcott, found in the Lower Carboniferous limestones of the Eureka District above, and separated from the White Pine shales by 3000 feet of quartzite conglomerate, is associated with the two above-mentioned Devonian species in the Spring Creek limestone of Arkansas. The two species L. quadricostatum and P. lachrymosa, have been regarded by Meek, by Hall, Whitfield and Walcott as Devonian. species. They are characteristic of Devonian rocks of New York, although the Leiorhynchus has been reported, with early Carboniferous species in Pennsylvania, as it has in Nevada. Not only the species but the subgeneric types in both cases are Devonian; both Productella and Leiorhynchus are characteristic Devonian modifications of the genera Productus and Rhynchonella respectively. These two species then are not only characteristic Devonian species but Devonian subgenera, and where seen in Nevada they are still below the fauna with which the general fauna of the Spring Creek limestone is correlated in the Nevada sections. Geol. Expl. Fortieth Par., vol. iv, p. 79. + Paleontology of the Eureka District, p. 5. |