EAST AFRICAN REPTILES AND AMPHIBIANS 55 



This is the best characterized of all the East African races of Agama 

 agama. It is distinguished by the great development of the canine 

 teeth in both upper and lower jaw, larger than in any of the other 

 races; the greatly depressed and flattened tail whose breadth at the 

 base is included in the body breadth from twice <'old males) to two and 

 one-third times (younger adults) ; the almost absent nuchal crest and 

 the extremely faint keeling of the dorsal scales which appear almost 

 smooth. The females of this species agree with the males in these 

 characters and can thus be readily distinguished from the females 

 of the preceding races. 



The following additions or corrections to Meek's excellent descrip- 

 tion are worth recording: The nasal opening is usually on, but some- 

 times below (41522, 42030, 49099), the canthus rostralis; the narrow 

 elongate scale on the middle of the snout is only present in 8 of the 

 14 specimens; in the others it is transversely divided or so reduced as 

 to be indistinguishable from the surrounding scales; the occipital 

 (and this point is of particular interest in connection with Havicauda), 

 while generally noticeably enlarged, is sometimes the center of a group 

 of scales alm.ost equally enlarged; in other specimens portions of the 

 occipital scale appear to have split off, thus reducing it in size until 

 it is no bigger than adjacent scales (probably some such condition 

 caused Werner to place his flavicav da in the third section of Boulenger's 

 key (1885) instead of in the second; it then ran down to the Persian 

 Agama 7iuj)ta, the species with which he compares it); 9 to 11 upper 

 labials with an average of 10; 9 to 11 lower labials mth an average 

 of 10 ; longest neck spines usually less than, but sometimes equal to half 

 the diameter of the tympanum; lateral scales, more particularly in old 

 males, with denticulated edges, sometimes six toothed; enlarged scales 

 on the median-nuchal line sometimes very strongly keeled; 69 to 93 

 scales around mid-body (70 to 103 in the Museum of Comparative 

 Zoology series) ; there is really no appreciable difference between the 

 length of the third and fourth finger, though there is between the 

 third and fourth toes as in the type; as stated by Angel, there may be 

 a double row of praeanal pores, for No. 42026 has 11 in the anterior 

 and 12 in the posterior row, while M. C. Z. No. 8200 has 2 in the an- 

 terior and 10 in the posterior (stranger still, an undoubted female 

 (U.S.N.M. 40919). holding large eggs, has a row of 10 pores), omitting 

 the supernumerary counts, 11 males (including M. C. Z. specimens) 

 show a range of from 7 to 12, with an average of 10 praeanal pores. 



From traces of color it may be said that the heads of males are red 

 in life, their bodies blue. In alcohol both sexes have the throat light, 

 vermiculated with the usual longitudinal dark lines; the sexes, how- 

 ever, may be distinguished in this series by every male being dark 

 (blue ?) beneath, while each of the seven females is light yellow or 

 white; the most characteristic feature of the coloring of this species 



