20 DR. J. ANDERSON ON MAMMALS, REPTILES, [Jan. 5, 



Genus Ccelopeltis, Wagler. 



7. Ccelopeltts LACERTiNA, Wagler; Boulenger, /oc.e^7. p. 151. 



1 specimen, Duirat, Tunisia. 



Total length 1275 millim., tail 347. Ventrals 170, subcaudals 

 105; 19 rows of scales. 



8. CcELOPELTis PRODUCTA, Gcrvais ; Boulenger, loc. cit. p. 151. 



2 specimens, Duirat, Tunisia. 



No. 1. Total length G63 millim., tail 121. Ventrals Ifil, sub- 

 caudals 62. 



No. 2. Total length 671 millim., tail 115. Ventrals 159, sub- 

 caudals 48. 



The first has 9 upper labials on the right side, and the normal 

 number 8 on the left, the 5th and 6 th entering the right and the 4th and 

 5th the left orbit. The grooving of the scales of these specimens is 

 very feebly indicated and in marked contrast to this character in 

 0. lacertina. 



This appears to be the second record of this snake from Tunisia, 

 the first specimen havuig been obtained by M. Vale'ry-Mayet at 

 Bou-Hfdma near Gafsa. The species was originally based on a 

 specimen fi'om the Sahara. 



Family II. Viperid^e. 

 Genus Vipera, Laurenti, 



9. Vipera lebetina, Linnaeus; Boulenger, Zoc. cit. p. 154. 



Var. nov. deserti. (Plate I. figs. 6 and 7). 



1 5 , Duirat, Tunisia. 



I have no hesitation in referring this specimen to V. lebetina, but, 

 as it is devoid of a canthus rostralis and has the scales on the head 

 from the parietal region forwards to the rostral perfectly smooth, I 

 regard it as a variety which I propose to call deserti. 



All authors who have hitherto had occasion to describe V. lebetina 

 from Algeria have either directly or indirectly referred to the keeled 

 character of the head-scales. Through the kind assistance of Mr. 

 Boulenger, who examined for me the type (No. 4017) of Y. mauri- 

 tunica, Guichenot, in the Paris Museum, I have his authority for 

 stating that the canthus rostralis of that snake is well marked, and 

 that the interorbital scales are feebly but distinctly keeled. It has 

 no large supraoculars, and the rostral is a little higher than broad. 

 It is a female with 163 ventrals and 50 caudals. Another and 

 young specimen. No. 4016 of Guichenot's Collection, is the same as 

 the preceding, but with the canthus rostralis less distinctly marked. 

 It has 166 ventrals and 39 cauJals. 



I have examined all the specimens of V. lebetina in the Brifisli 

 Museum, and I give the leading details regarding them in the 

 following table, and for comparison I have added in the last column 

 those yielded by this variety. 



