912 



SCIENCE. 



[N. S. Vol. XVII. No. 440. 



numbers, and, I was told, might even decline 

 to let an old subscriber have a copy of the 

 new edition unless the old one was returned 

 — to be destroyed. If this statement was 

 correct, the rarity of the old volumes would 

 be to some extent at least accounted for. 



The discovery of the new volume is inter- 

 esting chiefly from a historical or bibliograph- 

 ical point of view. The only essential change 

 it will entail is the dating back of the first 

 descriptions of seven species, viz., Oroialus 

 oreganus (so spelled), Coluber couperi. Co- 

 luber quadrivittaius. Coluber rhombomacu- 

 lata, Bufo quercicus, Salamandra quadrimacu- 

 lata and Salamandra haldemani. 



In my biographical memoir I did not con- 

 sider it necessary to correct or notice num- 

 erous misstatements respecting Holbrook's 

 works, but perhaps it may be advisable to 

 refer to one here. 



In Engelmann's ' Bibliotheca Historico- 

 naturalis ' (p. 172) and in Carus and Engel- 

 mann's ' Bibliotheca Zoologica ' (p. 134) to 

 ' Holbrook, John Edw.,' is accredited a publi- 

 cation entitled ' Scientific Tracts. 3 Vols, in- 

 12. Boston 1831-33 {London, Wiley and 

 Putnam.) 18s.'* 



John Edwards Holbrook had nothing to do 

 with that serial, the series having been com- 

 menced by one Josiah Holbrook in company 

 with other writers. I have been able to see the 

 volumes, which are in the library of congress. 

 The three volumes are composed each of 24 

 tracts of a monographic nature, the ' terms ' 

 being ' 24 numbers a year, at one dollar and 

 fifty cents, payable in advance.' Volume 1 

 has such contents as ' The Atmosphere ' (num- 

 bers 1 and 3), 'Geology' (2), 'Gravitation' 

 (4), 'Animal Mechanism' (5) and the like; 

 one of the coauthors was J. V. 0. Smith. 



This series was succeeded by a ' new series,' 

 ' conducted by Jerome V. C. Smith, M.D.,' 

 issued in numbers of 32 pages each on the 1st 

 and 15th of each month, miscellaneous in their 

 character, and paged to form two volumes 

 each year. Smith gave up and in 1836 a new 



* The words are quoted from Engelmann (p. 

 172) and differ i~lightly from those in Carus and 

 Engelmann. 



volume (apparently the last) of the ' Scien- 

 tific Tracts ' was published by others in 12 

 semi-monthly numbers of 32 or 24 pages, and, 

 at last, of 16 pages each. Those were the 

 years of tracts, religious, temperance, polit- 

 ical, and even ' scientific' 



My thanks for information respecting the 

 volume in question are due and given to Mr. 

 William J. Fox, as well as to Mr. Witmer 

 Stone. 



Theo. Gill. 



Cosmos Club, 

 April 28. 



REGENT ZOOPALEONTOLOGY. 



CONCERNING THE ANCESTRY OP THE DOGS. 



Mr. J. B. Hatcher, in a recent memoir on 

 Oligoeene Canidae, distinguishes three closely 

 allied genera from the White River formation 

 and proposes some very interesting changes 

 in the phylogeny of the family. His obser- 

 vations are based on the very fine specimens 

 of these rare fossils collected by Mr. O. A. 

 Peterson for the Carnegie Museum. These 

 include one complete and three incomplete 

 skeletons, the skulls all well preserved. The 

 thorough and clear description of the skeleton 

 of Daphcenus felinus is especially valuable as 

 based on a single and very complete specimen. 

 The resemblance of this primitive dog to. the 

 contemporary ancestors of sabre-tooth cats 

 has been strongly urged by Professor Scott in 

 his previous description of Daphcenus; Mr. 

 Hatcher, on the contrary, is impressed less 

 by its feline than by its creodont characters, 

 which he points out at some length. 



He distinguishes three closely allied genera 

 among these specimens : 



1. Daphasnus, with elongate skull, high 

 sagittal crests, robust premolars, etc. 



2. Protemnocyon, gen. nov., with short 

 skull, low sagittal crest and small premolars. 



3. Proamphicyon, gen nov., with elongate 

 skull, high sagittal crest, small premolars and 

 serrate canines. 



(The distinctions between the first two 

 genera are better displayed in the referred 

 species, D. felinus Scott and P. inflatus 

 Hatcher, than in the typical species D. vetus 

 Leidy and P. hartshornianus Cope, which, as 



