REVISION OF THE KING SNAKES. 157 



meets that of polyzona. It is probable that it intergrades with 

 annvlata somewhere on the Mexican plateau, perhaps in the region of 

 the divide between the Gulf and Pacific drainage. Its range north- 

 ward can only be surmised, but it should be expected in Sonora 

 west of the Sierra Madre. 



Published records for localities not included in the list of specimens 

 examined are as follows (Boulenger, 1894, 204): Mazatlan; Presidio, 

 near Mazatlan; Mezquital del Oro. 



Variation and ajjinities. — The material at hand indicates that this 

 is a fairly homogeneous form, characterized by a low number of 

 pairs of black annuli separated by broad red interspaces, absence 

 of black tips on the red scales, and by a light colored snout mottled 

 with darker. These characters apply equally well to the Tres 

 Marias Islands and to the adjacent mainland, but the numbers of 

 ventrals and scale rows is distinctly higher on the islands. Using 

 Boulenger's figures (1894, 204) for Forrer's specimens from the Tres 

 Marias, the average for 4 specimens is 231, the extremes 229 to 232; 

 this contrasts rather strongly with the average of 214 for 19 specimens 

 from the mainland. The extremes for these, 200 to 221, do not even 

 reach the numbers for the islands. All of the Tres Marias specimens 

 attain 23 rows of scales, while from the mainland this number is 

 possessed by only three, most of the others having the formula 

 21-19-17. Since the pattern and other structural features seem to 

 be the same in the island as in the mainland forms, it is not desirable 

 to make a specific distinction, on the basis of the specimens now on 

 hand. 



While the color pattern seems to be constant on the west coast, 

 it is very noticeable that toward the interior of Mexico some speci- 

 mens have the spaces on the belly opposite the dorsal red areas 

 partially or completely filled with black, presenting in this a striking 

 approach to annulata. The snout, too, may be blacker and the red 

 interspaces between the pairs of black rings may be much narrower 

 and strongly encroached upon by the latter. 



Since, in other respects, this form is like annulata, it is believed 

 that these similarities in pattern toward the interior of Mexico are 

 evidence of close affinity between those forms, and, in fact, are 

 sufficient evidence of intergradation. 



If the writer's conclusions as to the stem-character of annulata, 

 stated further on, with respect to the forms north of it be accepted 

 then the most natural inference with respect to nelsoni is that it, too, 

 is a derivativ^e of annulata. 



