Mai, 1893.] KEPTILES OF THE DEATH VALLEY EXPEDITION. 207 



southwestern deserts, which agrees with true P. eatenifer in having a 

 broad and low rostral. That Baird and Girard later referred specimens 

 of this form to P. bellona can not, of course, justify the shifting of this 

 name to another type. 



As a general rule this form lias a more pronounced carination of the 

 scales and a less number of smooth scales on the sides, but this char- 

 acter can not be relied upon at all, and whether a specimen shall be 

 referred to either typical P. eatenifer or to this desert form must be 

 decided upon the totality of the characters, as a reliance upon the car- 

 ination leads to very erroneous results. This will be plain at once to 

 any one who will take the trouble to examine and compare the descrip- 

 tions of the various species described by Baird and Girard in their 

 Catalogue of North American Serpents, and as I have examined a 

 number of their specimens I am able to state that the descriptions are 

 generally correct. It will then be found that these Pacific coast speci- 

 mens have only three to five outer rows perfectly smooth, while as 

 synonyms of P. eatenifer, the types of P. iciUccsii, etc., ' ought ' to 

 have nine rows of smooth lateral scales. Again, both types of P. mc- 

 clellanii which 'ought' to have only five smooth rows, because being- 

 true P. sdyi, have at least seven smooth rows. Furthermore, it has 

 been asserted that the typical P. eatenifer occurs as far east as 

 Pyrmont,* Nev., upon the strength of U. S. National Museum No. 

 8139. This number contains two specimens so alike otherwise as to 

 preclude the possibility of their belonging to two different species. 

 Why they should be referred to P. eatenifer I can not discover, for one 

 has only three perfectly smooth scale rows, while in the other the num- 

 ber is four or five. On the other hand, of two specimens in the present 

 collection, both from the Panamint Mountains, Calif. (Nos. 18065 and 

 18066), one has only four rows of smooth scales on each side, while the 

 other has ten. In every other respect the two are practically alike and 

 no one could reasonably refer them to two different species. Yet that 

 would have to be done were we to use the number of smooth scale rows 

 as a character. 



[This subspecies, according to Mr. Stejneger, is the form inhabiting 

 the Great Basin, while, as pointed out above, typical -P. eatenifer is 

 restricted to the coastal slope of California. 



On the east side of Pahrump Valley, Nevada, one of these snakes 

 measuring 5 feet in length was killed April 29, among the tree yuccas 

 along the upper edge of the Larrca belt, at an altitude of 1,310 meters 



* The name ' Pyrmont 'appears in the Eept. Wheeler Survey, Y, 1875, Zoology, p. 541, 

 the specimens referred to having been taken there by the Wheeler Expedition of 

 1872. This is probably the same place as Piermont, which is given on map sheet No. 

 49 of the Wheeler Survey, and on the 'Map of California and Novada with Parts of 

 Utah and Arizona/ published by the Chief of Engineers, U. S. Army, 1879. Pier- 

 mont is on the west side of Spring Valley and on the east slope of the Shell Creek 

 Pange. It is in White Pine County, Nev., about 75 miles due east of the town of 

 Eureka. 



