SERPENTES. 135 



Naja naja (Linn^). 



LiNirt, Syst. nat., ed. 10, 1758, 1, p. 221. Boulenger, Cat. .snakes Brit, mus., 1896, 3, p. 380-385. 

 Stejneger, Bull. 58, U. S. nat. mus., 1907, p. 394-397. 



T]ipe locality: — "Habitat in India orientali." 



Stejneger has outlined a scheme for tlie study of the varieties of the cobra 

 which is an advance over that of Boulenger. The latter grouped them accord- 

 ing to color alone, and we had as a result a curious series of subspecies in any 

 one of which might occur specimens from almost any locality. As a matter of 

 fact, by using the scale-counts of Boulenger and those of Stejneger, we may define 

 in a fairly satisfactory manner several apparent geographic races in the East 

 Indian region. 



Naja naja leucodira (Boulenger, loc. cit., p. 384). 



This form seems to be confined to Sumatra, whence specimens have come to 

 the British museum. Examples also exist in the zoological museum of the 

 Department of agriculture at Buitenzorg, Java. 



Body with no marking above; dark brown, almost black. Hood unmarked. 

 Throat whitish, usually spotted with black, and with a black band. Scales on 

 neck 21-25; generally 23; 17 or 19 rows across body. V. 183-193; C. 50-55. 



N. n. sputatrix (Boie). Isis, 1827, p. 557). 



I add here a transcription from a letter from Dr. Stejneger copying Boie's 

 text. Under N. tripudians Boie writes, "Im Pariser Museum Exemplare von 

 Java, Sumatra und den Philipinen, die kaum specifisch verschieden zu seyn 

 scheinen. Bey javanischen Exemplaren die Brille undeutlicher, die Lippen 

 ungefarbt." 



"6. sputatrix Reinw. Russell, Account of Indian serpents, 1796, 1, 2, 

 tab. 36. Elaps fuscus Merr. Coluber castaneus Oppel, cf. Erpet. de Java. 



"Oculis parvis, cauda s, tota e badio aenea, capite dilutiore, cervice in 

 junior e subperspicillato, 173-80 = 50-52." 



Javanese cobras show a remarkable constancy in the number of rows of 

 scales on the neck. This appears to be regularly 25, while the number of rows 

 around the body may be 19 or 21, the latter being most frequent. Boulenger 

 has recorded three specimens from Java belonging to two varieties. Two 

 {loc. cit., p. 382) as var. caeca, and one as var. sputatrix. The scale-counts in 

 all lie within these limits, viz. Sc. If; V. 172-180; C. 44 for the first two men- 

 tioned; and Sc. fi; V. 163; C. 44, for the single N. n. sputatrix. The following 

 counts, taken from seven specimens all from near Buitenzorg, show the range in a 

 restricted locaUty. They can not be separated on a color-basis, and are all N. n. 



