968 MR. A. LOVERIDGE : NOTES ON 



more $ $ which present an interesting link between forma 

 iypica and var. vmierescecce, and nullify my contention that there 

 is little foundation for saying that the prajorbitals are much 

 shorter than the rostral horn in $ $ , but I think still sho^v it is 

 impossible to rely on their relative lengths as a guide to sex as 

 Meek has suggested. 



Miss Procter has giA'en me the note she made on the specimens 

 when they turned up, and I reproduce this below, together with 

 the table which she drew up : — 



Register Number 3809. 3534. 3513. 3545. 3253. 3253. 3253. 3687. 



Sex <?<?<?<? ? ? ?? 



Length of head and body . 113 91 81 96 103 100 94 94 

 Length of rostral horn ... 25 18 14 17 11 11 6 3 

 Length of proeorbital llepresentod by mi- 

 horns 27 17-18 14 14 11 5-0 mitespiiuis ■omni. 



(1 only) long in both speci- 



mens as> in forma- 

 tj/2>ica. 



" Taking the species as a whole, the rostral horn is the one to 

 develop first, and is in all cases stoviter than the prteoi-bitals. 

 The prEEorbitals were evidently first developed in the d . In 

 forma iypica the d has the three horns of equal length; these 

 are represented in the $ by three short sharp spines (usually not 

 more than 1 mm. long). In these Nairobi specimens the horns 

 develop in the $ in a highly variable degree. The rostral always 

 more or less developed ; Avhen the rostral is shoi-t the prajoibitals 

 are ahaent as iu forma iypica ; when the rostral horn is long the 

 praeorbitals may be short, medium, or equal in length to the rostral. 

 In one S of var. vauerescecm the prseorbitals are actually longer, 

 though still more slender, than the rostviil, in another tlioy are 

 shorter as in the 5 . Var. vauerescecm therefore difl'ers from forma 

 typica in the presence of a horn, or horns, of variable degree of 

 development in the female." 



CHAMiELEON MELLEUI Gray, 



Blgr. Cat. Liz. iii. 1887, p. 472. 



Tlu'ee specimens. Two from Kilosa on 8. vi. it 8. ix. 22, of wliicli 

 one was on a low cocoanut-palm. The third was a half-grown 

 individual from Kisanga, t.aken 24.ii.23. 



As I know of no description of the coloration in life, it may 

 be as well to give that of one of these females. Snout a lich 

 dark green ; a vertical band of pale yellow from buccal bordei- 

 below eye extends upwards to postocular legion, but not on to 

 crest. A semicircular patch of pale yellow round forearm 

 anteriorly. Four vertical bands of a rich dark green on body, tirst 

 partly covered by posterior ends of occipital lobes, the second about 

 mid-body bifincates before reaching the dortal crest and each 

 arm of the foik again bifurcates on the crest, the third and 

 fourth are similar to the second; the inter-band areas pale yellow 



