48 BULLETIN 151, UNITED STATES NATIONAL MUSEUM 



nasals, which may well be the case with some examples, though not 

 so with those before me. Nor do either of them bear out the alleged 

 difference between West and East African geckos in the proportion 

 of the height to the breadth of the nostril which is alil^e in both. 



There are 15 lamellae under the first and fourth digits, as against 

 12 to 13 and 14 to 15, respectively, recorded for West African 

 epJiippiata. It may be that an East African race can be differen- 

 tiated, but until we have more material, in view of the well-known 

 variability of species of this genus, it appears advisable to exarcise 

 caution. In passing I might say that the chin shields of the Cameroon 

 gecko are markedly different on the right and left sides. Both 

 Boulenger and Anderson have noted azygous arrangements and sub- 

 divisions of the chin shields of this species. 



Family AGAMIDAE 



Genus AGAMA Daudin 



The agamas of the colonorum-planiceps-lionotus group are badly in 

 need of revision, but long series of topotypic material of some of the 

 undermeationed races is necessary before an3^thing like a satisfactory 

 statement can be made as to the taxonomic position of some of these. 



In studying the United States National Museum series from Kenya 

 Colony, together with the material from Tanganyika Territor}^ in the 

 Museum of Comparative Zoology, I have been impressed with a 

 number of facts which may tend to throw light upon the situation. 

 When writing the Catalogue of Lizards in the British Museum in 

 1885 BoTjlenger sought to differentiate colonorum and planiceps as 

 follows : 



a'. Body moderately depressed; dorsal scales mucronate colonorum, 



b'. Body much depressed; dorsal scales not or but very slightly 



mucronate planiceps. 



A perusal of the full descriptions given shows them to be almost 

 identical and that the essential differences are expressed in the key. 



Andersson '^ has shown that the Linnean types of Lacerta agama, 

 still in existence, can be identified with Agama colonorum and Uranis- 

 codon plica. The name agama must replace colonorum on grounds of 

 priority. Though the type locality was given as America the proba- 

 bility is that the two specimens of Agama agama (Linnaeus) came 

 from West Africa, which is the headquarters of the species. 



For convenience sake at first I attempted to treat agama and 

 planiceps as distinct species, each with its several races, but on 

 studying the material so much intergradation occurs between the 

 races of the two groups that one was forced to the conclusion that the 



» Andersson, 1900, Bih. Svenska Vet-Akad, Hand., vol. 26, p. 11. 



