According to Sellards the hard phosphate, belonging to the Alachua (Dunnellon) formation (Fla. Geol. Surv., vol. V, p. 37) resulted from a disintegration of underlying Upper Oligocene deposits and probably the Vicksburg limestone. Through chemical action these rocks were partly dissolved and the residual materials were mixed by local subsidence and by action of streams and later modified by chemical changes. The land-pebble phosphate of the Bone Valley formation had, Sellards concluded (Fla. Geol. Surv., vol. VII, p. 55), resulted from underlying phosphate marls of Upper Oligocene age. This occurred during a time of general subsidence of sufficient extent to permit marine waters to reach the area covered by the Bone Valley phosphates. The presence of sea-water is indicated by the occurrence of bones of cetaceans. With regard to the effects of streams and of the chemical action of the water on the rocks, which contributed to the formation of the hard rock phosphate and the production of sinks and caves, it may be remarked that we know of no time when rocks were dissolved and caves formed to the extent that they were during the Pleistocene. As shown on page 15, various deposits of marine marls along the Atlantic coast are referred by the writer to the Nebraskan. Among these marls are the coquina rock found at St. Augustine and the marine marl underlying the bed at Vero, which contained early Pleistocene vertebrate fossils. These marls are known to extend well inland, being found at Kissimmee, 50 miles from the coast. In some places they are met with at depths of 70 feet (Sellards, Fla. Geol. Surv., vol. VIII, pp. 105–106). Marls of probably the same age occur on the western coast of Florida (Dall, Bull. 84, U. S. Geol. Surv., p. 152). The writer believes that some of these marls may yet be connected with the phosphate beds of the Bone Valley formation. A figure taken from Sellards (Geol. Surv. Fla., vol. VII, opp. p. 53) may be found on page 377. This illustrates the relation of the Dunnellon and Bone Valley formations to the underlying deposits. XI. THE FIRST INTERGLACIAL, OR AFTONIAN, STAGE. Mention has been made of collections of fossil vertebrates which long ago were secured at Fossil Lake, Oregon, and of others along Niobrara River, near Grayson, Nebraska. Lists of the species found at each locality were given by Dr. W. D. Matthew in 1902 (Bull. Amer. Mus. Nat. Hist., vol. XVI, pp. 317–320). These deposits and animals were regarded by Cope and Marsh as belonging to the Pliocene, until G. K. Gilbert, in his work on Lake Bonneville (Monogr. I, U. S. Geol. Surv., pp. 393–402) showed that the Oregon fossils must belong to the Glacial epoch, but he referred them to a late time in this epoch, that of the last glaciation. It thus became quite impossible to determine the age of any collection of fossil vertebrates. In 1887 (Univ. Geol. Surv. Kansas vol. II, pp. 299–308), Williston wrote: