ABSTRACTS OF PAPERS 137 



This discrete and profuse subfamily arrangement would be shocking to a 

 "lumper" like our late colleague and honored friend, Dr. Richard Lydekker, 

 who combined" all the mastodons and elephants into two genera, namely, 

 Mastodon and Elephas. The application of subfamily names to these mono- 

 phyletic, or similar polyphyletic ascending series, is considered preferable to 

 the coining of a new taxonomic term. 



The propriety of thus applying subfamily terms is disputed by some pale- 

 ontologists, notably by my colleagues, W. D. Matthew and W. K. Gregory. 

 The subfamily termination, inw, may, in the author's opinion, be adopted with- 

 out any real exaggeration to express the fact that many of these phyla have 

 Jyeen distinct and separate from each other for enormous periods of geologic 

 time. This is real hereditary relationship in the family or subfamily sense. 

 For example, it may be shown that the longirostral bunomastodont phylum 

 began with Palwomastodon of the Upper Oligocene, and that this animal was 

 already too specialized as a longirostral bunomastodont to constitute the an- 

 cestor of any other phylum than its own. This main longirostral phylum is 

 geologically the oldest and phylogenetically the most complete. It illustrates 

 one general law of mammalian evolution, namely, that a phylum having spe- 

 cialized in a certain character usually tends to evolve this character to an 

 extreme; the long jaw of Palceomas'fodon goes on lengthening until in Lower 

 Pliocene time it attains the great length observed in the forms recently de- 

 scribed by Barbour^^ as Eubelodon morrilli, Megahelodon lulli. 



In this longirostral phylum, as well as in the brevirostral bunomastodonts, 

 the question of the application of the generic nomenclature of Linnaeus is cer- 

 tainly a most puzzling one. Thirteen distinct generic names have been pro- 

 posed for the longirostral bunomastodonts and six distinct generic names for 

 the brevirostral bunomastodonts. 



Several puzzling questions arise: first, how many generic names can con- 

 sistently be applied within each of these phyla; second, which generic names 

 in the long list shall be given precedence ; third, shall the law of the technical 

 priority of a name prevail, or shall we recognize only the priority of the first 

 clear definition and conception of a genus which is based on one or more defi- 

 nite and clearly described characters of its genotypic species? 



This whole question has been raised in the previous communication to the 

 Paleontological Society by Doctor Matthew." I am disposed to recommend 

 that certain well defined generic names may, after due consideration, be 

 adopted by the Paleontological Society as nomind servantur. The selection 

 of these names will be greatly facilitated by a true phylogenetic classification 

 of the Proboscidea, to which the present outline is preliminary. 



At 12.30 the Society adjourned for luncheon. 



CONTINUATION OF SYMPOSIUM 



At 2 p. m. the symposium was resumed, with the reading of a paper 

 on the Mesozoic history l)y Doctor Stanton entitled 



11 Richard Lydekker: Catalogue of the Fossil Mammalia in the British Museum (Nat- 

 ural History). Pt. IV, The Order TJngulata, Suborder Proboscidea. 8vo. liondon. 1886, 

 pp. xxiv, 233 (1). 



^ E. H. Barbour : Op. cit. 



" W. D. Matthew : Op. cit. 



