EAST AFRICAN REPTILES AND AMPHIBIANS 



81 



this series listed above appear to me to be much nearer ahli than 

 africanus despite Mertens' statement that africanus is the mainland 

 form as shown by Peter's collections. At the same time they cer- 

 tainly are in no way separable from the Dar es Salaam and Pemba 

 skinks, and the specimen of the latter is supposedly topotypic of 

 africanus. Were this not the case one might presume the existence 

 of a southern race from Pemba to Mozambique {aMi) and a northern 

 from Mombasa to Manda (africanus). 



Sternf eld's key is divided into three sections as follows: 



1. Scales in 22 rows. 



2. Scales in 24 rows (sometimes 22 or 26). 



3. Scales in 26 to 28 rows (sometimes 24). 



And it might be said that for all practical purposes these very arti- 

 ficial major divisions are useless. The Lumbo series, for example, 

 is comprised of 5 skinks with 22, 20 with 24, and 5 with 26 rows, giving 

 an exact average of 24, which causes these skinks to fall into the 

 Comoro-Madagascar-Aldabra Section II. 



The minor divisions are based on color and the whole description 

 of africanus appears under Section I, c. and might be translated as 

 follows: "Upper side brown or greenish-brown; dorsolateral stripes 

 very sharply defined, edged with dark border; middle of the back some- 

 what brighter * * * A. h. africanus new subspecies (Manda, 

 Mandabucht, Malindi, Pemba)." 



The tabulation of the material available to me is as follows: 



Num- 

 ber of 

 speci- 

 mens 



Locality 



Scale 

 rows 



Average 

 number 



Labials 

 ante sub- 

 ocular 



Length of 



head and 



body- 



Mombasa, K. C 



Pemba Island 



Dares Salaam, T. T... 

 Lumbo, Mozambique 



22-24 



22 



22-24 



22-26 



4-5 

 4 

 4 



4-5 



Milli- 

 meters 

 39-43 

 46 

 45-47 

 32-46 



I have already recorded specimens from Pemba with 24 mid-body 

 scale rows, so that, though at first sight there does appear to be an 

 increase in scale rows as one proceeds southwards, the average num- 

 ber at Pemba would be 24 if I included all the material I have ex- 

 amined. These series are too small for one to form definite 

 conclusions. 



Mertens has said that though his type of ahli has 5 labials anterior 

 to the subocular, 4 is the number in all his paratypes. This is in 

 keeping with my conclusions that 4 is the normal number and 5 

 occurs only as a deviation from it. Tv/o of the Mombasa skinks 

 have 5 on one side of the head and 4 on the other. 



