XVIII. THE AGATE SPRING FOSSIL QUARRY.'

By O. A. PETERSON.

In 1904 a field party of the Carnegie Museum, the writer in charge, was extremely fortunate in discovering what will undoubtedly prove to be one of the most important quarries of fossils as yet discovered in the Miocene of North America. Geographically the quarry is located on the Upper Niobrara River, locally known as the "Running Water," in Sioux County, Nebraska. It is about twenty-five miles southeast of Harrison, a station on the Fremont, Elkhorn, and Missouri Valley Railroad.

The geological horizon in which the quarry is located is at the base of the Nebraska' Beds, or the top of the Harrison Beds.

In 1890 Mr. James H. Cook, on whose property the Agate Spring Fossil Quarry is located, discovered many small bones and fragments in the talus from the fossil-bearing stratum of the hills, in which the quarry is located. Very naturally he thought that the bones were those of Indians interred together with their horses. Mr. Cook accompanied me to this place in August, 1904. Realizing that this was a discovery of much paleontological promise we immediately began work on the deposits and resumed work early in the season of 1905.

The quarry is about one quarter of a mile south of the river referred to above. It is near the base of two closely connected and rounded buttes, which have been separated by erosion and in part covered by vegetation. These hills are composed of a buff-colored sandstone varying in degrees of hardness. A layer of about three or four feet in thickness, including that of the fossil-bearing stratum, is of a rather light color. The strata in this immediate vicinity, including that of this fossil quarry, have a slight dip northward. Immediately under and overlying the fossil-bearing stratum there are layers of compact Read before the American Society of Vertebrate Paleontologists, December 27, 1905.

2 Scott, Bull. Geol. Soc. Am., Vol. V., pp. 595-596, 1893. Hatcher, Proc. Am. Philos. Soc., Vol. LXI., pp. 117-118, 1902. Peterson, ANN. CARNEGIE MUSEUM, Vol. II., pp. 473-474, 1904. Darton, U. S. Geol. Survey, “Geology and Underground Water Resources," pp. 177-178, 1905.

and quite hard sandstone. The fossils are embedded in softer matrix in a stratum of approximately twenty inches vertical thickness. The bones are sometimes found to be slightly crushed, but on the whole the material is in a very fine state of preservation.

During the season of 1905 the party uncovered an area of 45 x 20 feet in the quarry. This area was plotted out in squares five feet in dimension, and a diagram (see Plate XXI.) was made representing this arrangement. The bones are found mostly disarticulated, much mixed, and thickly distributed through this layer of sandstone. It was soon discovered that the most intelligent way in which to secure this tangled

FIG. 1. located.

mass of material was to take out blocks of sandstone which contain the fossils. On the diagram the more important specimens are always indicated, as are the areas in the quarry where the bones are less abundant, and the numbers correspond to every package. Thus an intelligent idea of the various features of the quarry may be had at any time. The sketch is of importance in connection with the work in the laboratory. By this system of carefully indicating the relative size, shape, and position of each block, the correct association of the parts of any articulated skeleton, which may be found in this great mixture of bones, will be possible.

The irregular line on the diagram of the Agate Spring Fossil Quarry (see Plate XXI.), marked "Outcrop of the Fossil Layer" indicates the contour of the hill before work was commenced. The region marked "Cook Excavation" represents the area, which Mr. Cook excavated in the fall of 1904. Mr. Cook, realizing the importance of having the specimens preserved by the best and most modern styles of collecting, kindly desisted from his operations, after receiving a letter from the writer with an earnest appeal for postponement of effort until the party from the Carnegie Museum should reach the spot in the following spring (1905). From the line of Mr. Cook's explor

FIG. 2. Excavating in the Agate Spring Fossil Quarry.

ation the excavations and work were carried on by Mr. T. F. Olcott from April until July, when the writer accompanied by Dr. J. A. Hermann again joined the party, and continued work with it during the remainder of the season of 1905. Toward the latter part of the season Mr. W. H. Utterback also joined us and together we accomplished quite effective work. The blank area on the sketch indicates the exposed layer where work will be resumed in the coming season of 1906.

The outlines of the blocks are nearly always irregular. This is due to the method which we were sometimes forced to follow in working around an important specimen in order not to break through its

center.

When a block has been outlined, and it is decided where it is to be detached, it is bandaged and undermined to that point, and easily broken off from the main ledge in the quarry. The broken surface of the block and the corresponding face of the quarry are now carefully shellacked to prevent the contacts from crumbling. The different faces of the blocks are lettered and numbered to facilitate work in the laboratory. Plaster of Paris and strips of burlap, together

FIG. 3. feet long.

with flour-paste and delicate strips of muslin, where they are needed in protecting the bones and to firmly hold the block together, are the materials employed in connection with quarries of this kind.

A PROVISIONAL LIST OF THE FAUNA.

Of the more important forms found in the quarry the following may be mentioned.

Diceratherium Marsh, the remains of which are apparently most abundant. Other genera of rhinoceroses, besides Diceratherium, are also present. Some fifty or sixty skulls, lower jaws, and other skeletal material of rhinoceroses have already been taken out of the

quarry. Next in importance is the very large Elotherium which I have provisionally named Dinohyus hollandi, pending a more systematic study of the specimen. Moropus distans Marsh is well represented by limbs, foot-bones, parts of jaws, teeth, etc. Merychyus Leidy and other small forms are also present. Among this general mixture are found the remains of some carnivores, common in these horizons.

From this list it would appear at first sight, that the geological horizon in which this quarry is located, should be referred to the same horizon as the Upper John Day of Oregon. The fact, however, remains that the entire fauna, which is at present known from the upper series of the Miocene section in western Nebraska and eastern Wyoming, shows more highly specialized characters than obtained in the fauna known from the Upper John Day. For example, no such specialized forms among the Oreodontidae as Merychyus elegans, Merycocharus proprius Leidy, and the artiodactyls Merycodus Leidy, and Syndyoceras Cooki, recently described by Barbour, have been reported from the Upper John Day.

Moropus distans is much larger and doubtless differs otherwise from the form represented in the John Day formations. When the material representing Diceratherium from the Agate Spring Quarry is prepared for study, there will, no doubt, be found a number of important specific, if not generic differences, separating it from the material representing the genus in the John Day Beds.

From these recent discoveries it appears that the Miocene section from the Oligocene to the top of the Nebraska Beds, in this general locality may perhaps have to be regarded as belonging to the lower Miocene. From the study of the subdivisions of the Miocene strata in western Nebraska, and eastern Wyoming as distinguished by Hatcher,' we understand that the Harrison Beds represent the hiatus between the Upper and Lower Deep River Formations in Montana. Since Diceratherium and the large Dinohyus have been discovered in the uppermost Harrison or at the base of the Nebraska Beds it would seem to indicate the necessity of regarding the Monroe Creek Beds as of considerably later geologic age than that assigned to them by Hatcher. The Gering Beds which form the lowermost division in the Miocene formation in this locality, would then have to be regarded as equivalent to the Upper John Day and Lower Deep River. However, a 1 Loc. cit., p. 118.