EAST AFRICAN LIZARDS. 945 



wliite, the throat only being mottled or streaked with greyish 

 reticulations." 



There were eight eggs (23 x 11 mm.) in tlie ovary of a Kidai 

 female and apparently quite ready for laying (18. ii. 23). Young 

 ones about two inches long were very numerous between Kidenge 

 and Pwaga (ii. 23). 



In the stomach of Gwao's specimens was a quantity of fruit,' 

 in another the leaves of " upilo." Other indeterminate 

 vegetable matter in many others, also termites and ants of various 

 species, including, I believe, a stink-ant {Faltothyreus sp.). At 

 Izikisia 1 watched one of these lizards snapping up termites 

 after a rain-storm, which was causing them to emerge in large 

 numbers. The stomach of an Ikikuyu specimen was crammed 

 with termites. 



A very high percentnge of these lizards are aflfected with a 

 parasitic nematode worm {Thelandros sp., $ 5 '^on det.). 



I had great diiUculty in hunting these active lizards on the 

 rocks until I offered a. small reward to the delightful small boys 

 of the Wanyaturu tribe, who are experts with bows and arrows, 

 which they all appear to carry. They began shooting them, 

 which I had presently to stop for lack of preservative. They gave 

 me to understand that they hunt and eat them in times of 

 scar cit}^ and this is very possible, as these youngsters regularly 

 ate the bodies of the rats as they were skinned. 



Agama LiONorus var. MWANZ^, var. n. ' " 



The Wasakuma call this Agama " kuli." . 



In Mwanza district forty-one specimens were collected of an 

 entirely different Agama, but which has no specilic characters by 

 which I can distinguish it from the foregoing. They are abundant 

 in Mwanza itself, but were not collected thei-e, my series coming 

 from Mwadira, Shanwa, and Sa^ayo. For this col our- form I 

 propose the varietal name of mwanzcc, with Shanwa, Mwanza, as 

 type locality. The c? and $ types collected on 20. x. 22 have , 

 been presented to the British Museum. 



The longest neck-spines were less than half the diameter of tho 

 ear-ojiening, i.e. 2 mm. and 5 mm. Labials 10-11. Pire-anal 

 pores of the males ranged from 10 to 13 and averaged II, this 

 being based on twenty-one males examined. 



The number of scale-rows at mid-body according to locality 

 and sex are : — 



1 Sagaj'o male 87 gives an averaKe of 87 scale-rows, 



7 Mwadira males range from 71 to 90 „ 79 » . • 

 15 Shanwa „ „ „ 70 to 82 „ 76 „ 



8 Sagayo females „ „ 70 to 78 „ 73 „ 

 11 Shanwa „ „ „ 70 to 85 „ 77 „ 



which gives a range for the variety of 70 to 90 with an average 

 of 78. 



The coloration of the adult male. — " Snout to just beyond mid- 



