EAST AFRICAN LIZARDS. ' 951 



All markings vanisli at mid-body, the posterior Jialf being pinkish 

 shading to sand3^-hu(r on the. tail. Four other specimens have 

 the upper surface much speckled with black, approximately in ten 

 longitudinal rows. Tliough showing nluch varLition in colouring, 

 these lizards are very <li(Ierent from lY. kiloste and readily separated' 

 itt ;i. glance. 



A fetnalo ((JO 4-5)5) from Shanwa, 20. x, 22, had ovules rather 

 large; another (67 -f-r) from Ikikuyu, 10. ii. 2.S, held six ovules 

 inoMsnring 7x7 mm. The testes of a male killed on the same 

 (Iny were eulnrged. 



Stomach contents of eight specimens examined were : — (1) cock- 

 vooches and termites, (2) a brown shield-bug. (3) grasshopper 

 and cricket, (4) spider, (.5) cricket, many termites, and a spider, 

 (G), (7), (8) full of termites. 



NucRAS KiLos.E Loveridge, 



Love. P.Z.8. 1922, p. 313. 



Three specimens from Gwao's and Mtalis should be referred to 

 kilosce on colour grounds and possibly on the basis of keeled 

 dorsals. As pointed out under JV. emini, there appears to be 

 intergrading in specimens collected mid-way between the type 

 localities of N. emini and N. kilosce, and the latter is probably 

 oidy a. subspecies of iV. emini. 



Since describing this lizard I have obtained eighteen further 

 rspeciinens from the type locality. Twelve were inales, four females, 

 find two very young. Tlie disproportion in the sexes is probably 

 to be attributed to a greater agility on the part of the females 

 in eluding capture. 



The largest perfect male measured 8| inches (65 -f 147), whilst 

 another exceeded previous maximum measurements by being 

 67 mm. from snout to vent ; the tail was reproduced in this 

 specimen. As the tail was missing in the type female, it 

 may be well to record the measurements of the three perfect 

 specimens in the present collection, (a) 50 -f- 11 7, (6)50-f-113 

 <c) 43 + 88. 



A careful study of the scale characters of these specimens 

 enlarges our knowledge in several particulars, viz. : — 



" Ventral plates in 27 to 31 transverse series " should read 25 

 to 31. The average of these fifteen is 27. 



" Dorsal scales 42 to 56 across mid-body " should read 40 

 to 56. 



"Femoral pores 11 to 13" should nov/ be 10 to 13, the 

 average of the fifteen specimens being 11"6, 



In one the frontal and fronto-nasal were bi-oadly in conta^ct.' - 



Four of these specimens had grasshopper remains in their 

 stomachs. 



One specimen had a number of parasitic nem.'itodes in the 

 stomach which were identified by Dr. Baylis as Pliysalo'piera 

 ahhreviata llud. 



62* 



