reduction of a still more complex molar. Most of the evidence for this conclusion is derived from the fact, as he believes, that the Mammalia of Eocene and possibly earlier age, which are found in Argentina generally, have quadritubercular molars. In accordance with this view Cetacea and Edentata with numerous teeth, present a primitive type of dentition which has survived.

The reply which can be made to this fundamental proposition as to time-order, is, that M. Ameghino has probably affixed too great an age to his earlier beds. This is the opinion of Lydekker, and such extinct types as occur in those beds which occur elsewhere confirm this conclusion. Thus the Patagonian, which Ameghino regards as an Eocene formation, containing the Pyrotherium, contains also the primitive monkeys Anthropops, and the cetaceous Prosqualodon, Argyrocetus and Diaphorocetus. Now Diaphorocetus and forms closely allied to Arygocetus and Prosqualodon are characteristic of the middle Miocene in North America and Europe. It is highly improbable that the quadrumanous genera discovered by Ameghino are of Eocene age, since nothing of the kind occurs in Eocene beds in the Northern Hemisphere, where more primitive and ancestral lemuroid families represent them. The presence of supposed Condylarthra (not yet described) however, gives an Eocene character, and if the forms described by Ameghino as Multituberculata are really such, this character would be difficult to deny. However, recently Ameghino has recognized that these forms do do not belong to that order, but are true Marsupialia, and Lydekker assert that they do not belong to the Patagonian formation, but to the overlying Santa Cruz beds. But supposing that the Patagonian formation is upper Eocene, it does not furnish the material for an elucidation of the dental characters of the primitive Mammalia. These are only partly displayed in the lower Eocene, for it is in the Postcretaceous (Puerco and Laramie) that the true ancestral relation of the tritubercular molar is fully seen. These formations may be represented by the lower or dinosaurian beds which lie below the Patagonian formation in Argentina, but no Mammalian remains have been found there thus far by Ameghino. The oldest Mammal is said to be the Pyrotherium of the Patagonian formation, but it has an aspect more modern than Eocene. It is suspected by Ameghino to be a proboscidian, but it has not yet been shown that it is not a marsupial.

Dr. Ameghino misinterprets North American fossils in more than one instance. He cites the Amblypoda in evidence of the proposition that the tritubercular molar is the result of a reduction of the quadri

Geographical History of Mammals, 1896, 115. Ameghino makes the same statement in Enum. Synopt. Mamm. Foss. Eocene de Patagonie, 1894, p. 10.

tubercular. His series is Uintatherium, Coryphodon, Pantolambda; the last the most completely tritubercular. The time order is, however, the reverse, viz.: Pantolambda (Puerco), Coryphodon (Wasatch, and Uintatherium (Bridger); the first the most unmodified tritubercular. In accordance with his general position Dr. Ameghino believes (p. 72) that the quadritubercular genus Procyon is of great antiquity and prior to tritubercular types. This, however, cannot be believed. It has descended from a primitive plantigrade tritubercular, canine type, as have their allies the bears. The same modification is seen in the Mustelidæ in the badgers; and all such are modern forms. He states (p. 26) that in Periptychus and Mioclanus, Phenacodus and Achænodon, the teeth are quadritubercular. The first two genera have tritubercular molars with insignificant rudiments of others both before and behind the protocone (Periptychus) or behind only (Mioclanus), and they belong to the primitive Puerco period. The other two genera are quadritubercular, but belong to later beds, Phenacodus being Wasatch, and Achænodon, Bridger, neither of which formations has any genus in common with the Puerco (except Didymictis of Puerco and Wasatch ages).

Dr. Ameghino believes the Typotherian suborder of the Toxodontia to be related to the Quadrumana. The digits resemble decidedly those of that suborder, but one important difference is overlooked by him. He has pointed out the striking alternation of the two rows of carpal bones in the Typotheria, in which they agree with the Toxodontia proper, and with the Amblypoda. Now in primitive Quadrumana this alternation does not exist, but the bones of the two carpal rows, like those of the tarsus, are directly juxtaposed, or taxeopodous. This characterizes the Condylarthra, which furnish the exact foot characters of the lemuroids, or ancestral Quadrumana.

Finally as to Dr. Ameghinos' views of the origin of the Cetacea, he again inverts the order of succession. He does this by assuming that the Archæoceti are not related to the Cetacea proper, and cannot be ancestral to them. He does not regard the presence of two rooted molars in the foetal Balana as significant in this direction. The opinion of zoologists and paleontologists has been different from this, and I have confirmed the general view in my recent researches on the extinct Balænidæ of the Eastern United States. I have shown that a decided sagittal crest like that of the Zeuglodontidae exists in some of the Miocene whalebone whales. In my estimation the simple teeth of many Cetacea are the result of a process of dental degeneration. 7 Proceeds. Amer. Philos. Soc., 1895, p. 139; 1896, p 141. .

Their increase in number has not been due to subdivision of primitive teeth as supposed by Kükenthal, nor is it a survival of primitive conditions, as supposed by Ameghino, but it is probably a repetition of similar structures due to an extension of the dental groove and dental lamina, following the gradual elongation of the maxillary and premaxillary bones, proceeding contemporaneously with degeneracy of the teeth themselves.-E. D. COPE.

Eozoon canadense.-In recent numbers of the Geological Magazine, Dr. Dawson cites the evidence to date for the animal nature of Eozoon. Briefly summarized, the facts are these: (1) The rocks of the Grenville series, where the fossil in question occurs most abundantly, belong to a sedimentary formation. (2) They form a great calcareous system comparable with the metamorphosed Paleozoic calcareous beds of organic origin in petrological and chemical character. (3) New material showing more plainly the structure of the canal systems and tubes, evidences a definite plan of macroscopical structure. (4) Late discoveries of Archaeospherinae and other objects supposed to be organic in pre-Cambrian rocks in Canada and in Europe afford, to some extent, corroborative evidence in favor of Eozoon. (Geol. Mag. 1895.)

Thickness of the Coal Measures.-According to Mr. J. C. Branner, the total thickness of the Coal Measures (Pennsylvanian) sediments in Arkansas is greater than that of the sediments of the same age in other parts of the country or of the world. He gives the following table of comparison: Arkansas, 23,780 feet; Nova Scotia, 16,000 feet; Utah and Nevada, 16,650 (?) feet: Indian Territory, 10,000 feet. Mr. Branner finds a reason for the great thickness in the drainage of the Continent during Carboniferous times. "The rocks of this series in Arkansas afford fossil evidence that this part of the Continent was probably not much above tide level. The drainage from near the Catskill Mountains in New York flowed south and west. The eastern limit of the basin was somewhere near the Archean belt from New England to central Alabama. This Appalachian water shed crossed the present channel of the Missouri from Central Alabama to the Ouachita uplift, and the drainage flowed westward through what is now the Arkansas Valley, between the Ozark Island on the north and the Arkansas Island on the south." (Amer. Journ. Sci., Sept., 1896.)

Geological News.-PALEOZOIC.-An accumulation of fresh material from the Ichthyologic fauna of the Cleveland Shales, Loraine County, Ohio, has enabled Mr. C. R. Eastman to determine the rela

tions of certain body-plates in the Dinichthyids. These are the median ventral plates of Titanichthys, and a postero-dorso-lateral of Dinichthys. The author further states that "every plate present in the body armor of Coccosteus has its representative in Dinichthys, and that the conditions of overlap and underlap are the same in both forms. (Amer. Journ. Sci., July, 1896.)

CENOZOIC.-Mr. H. W. Fairbanks states that the Lower Cretaceous age is represented in Santa Barbara County, California, by the Knoxville beds, containing the characteristic Aucella fossils. This is the southern point at which the genus Aucella has been found in California. (Bull. Dept. Geol. Univ. Calif., Vol. II, 1896.)

The skull of Orycteropus gaudryi (Ant-Bear or Aard-Vark) from the Lower Pliocene of Samos, described by C. W. Andrews, indicates an animal about one-fifth less than the living species. The close resemblance between the fossil and recent forms is remarkable. Dr. Forsyth Major has pointed out that the former distribution of the genus indicates its northern origin, and that it spread into Africa along with the rest of the Pliocene Mammalia with which it has been found. (Proceeds. Zool. Soc. London, 1896.)

In discussing the geographical distribution of the known Castoroid species, Mr. Merriam notes that the American Castorinae seem to reach their maximum development at or before the beginning of Pliocene time, while the culmination of the Eurasia group appears to occur in the Pliocene. This apparent earlier culmination of the American Castorinae, together with the earlier extinction of certain forms in this country, seem to point to an American rather than to the European origin of the family. (Bull. Dept. Geol. Univ. Calif., 1896.)

According to Dr. Shufeldt, Harpagornis, the fossil bird recently found in New Zealand, represents a more or less aquiline type, that might easily have been the common ancestor to a number of genera of existing modern eagles, as, for example, Haliaetus, Aquila and Thalassaetus. A natural scheme of classification would place it between the genera Aquila and Thalassaetus. (Trans. New Zeal. Inst. [1895], 1896.)

ZOOLOGY.

Fishes in isolated pools.-The occurrence of fishes in pools which have no communication with running streams or large bodies of water has been often noticed, and the explanation of their origin and persistence in such places is in some cases not satisfactory.

In collecting during this month (September) in Camden county, New Jersey, I made the following observations. I fished near Winslow, a pool of about twenty-five feet in diameter, and two feet in depth, with a muddy bottom and a few Nymphæas growing in it. It is distant about a quarter of a mile from an insignificant ditch with a little running water, and is surrounded by higher and sandy ground, offering no possible communication with the ditch. A half mile distant and still more inaccessible is a running stream. From this pool I caught large numbers of the following fishes. Umbra pygmæa, Apomotis obesus and Acantharchus pomotis. The Acantharchi were small, while the others were fully grown.

A quarter of a mile distant from this pool, and at an equal distance from the ditch above mentioned, and not connected with it by any depression of the surface, is another pool of about thirty feet in diameter. The water reaches a depth of three feet in a limited portion of it, and Nymphæas are more numerous, together with Utricularia, etc. Here I obtained the following fishes in considerable numbers. Umbra pygmæa, Amiurus prosthistius, Esox vermiculatus, Aphododerus sayanus, Apomotis obesus, Mesogonistius chaetodon, Acantharchus pomotis. Many of these were fully grown. The turtle Chrysemys picta was also abun

dant.

The mud in the first mentioned pool was light colored, and all the fishes were remarkable for the extreme paleness of their tints. The second pool is situated in better soil and its mud contains much decomposing vegetable matter, and is consequently black. The fishes were all deeply pigmented, including the three species found in the other pool, from which they could be distinguished at a glance. The smaller pool was said to have been dried up during the past summer.

Seven additional specimens of the Amiurus prosthistius Cope enable me to verify the characters already given (Proceeds. Acad. Nat. Sci., Philada., 1883, 133), from an examination of four from the Batstow River, New Jersey. In five of the new specimens where I counted the anal rays, they number 26. Prof. Jordan has recently attempted to