South African Agamas allied to A(jama liispida and A. atra. 231 
beyond foiu'th ; tail longer than head and body, feebly compressed and crested 
in the male. 
A vertebral crest or regular series of strongly keele<l scales on the 
back ; foot much shorter than tibia ; ventral scales more or less 
distinctly keeled . . . . . . . .A. anchietae, Bocage. 
No trace of vertebral crest on the back ; foot as long as or slightly 
shorter than tibia ; ventral scales smooth . . . .A. methucni, sp. n. 
II. Dorso-lateral scales with the keels pointing obliquely outwards ; no 
dorsal crest ; ventral scales smooth : ear-opening lai'ge ; fifth toe 
extending as far as first, or farther, third and fourth equal or 
either slightly the longer ; tail longer than head and body, more 
or less compressed and crested in the male . . . .A. atra, Daud. 
Enlarged dorsal scales usually present but small and few; lUo to 
loU scales round middle of body i^'orma lypica. 
No enlarged dorsal scales ; 150 to 1(30 scales round middle of body 
var. uiicrotcrvlepis, BJgr. 
Scattered strongly keeled and mucronate scales on the back; 170 
to 180 scales round middle of body var. holubi, Bocage. 
Numerous strongly enlarged spinose scales on the back ; to 110 
scales round middle of body ....... var. nulis, n. 
As the work of testing and criticizing proceeded it was found that the 
definitions of divisoius I and II of the Synopsis deserved to stand, the 
direction of the dorso-lateral scales - a character which had not been pointed 
out before — being readily applicable in all cases. But many of the characters 
suggested for further subdivision proved to be either utterly worthless, or 
else subject to such frequent exceptions as to require a complete re-casting 
of the key. Thus the size of the ear-opening as compared to that of 
the eye-cleft was found to suffer exceptions amounting to 20 per cent, in 
A. distanti ; the degree of slenderness of the toes and the proportions of 
them varies much more than was supposed, and the same may be said of the 
scaling on the back, on the tibia, and on the belly. 
The following detailed table Avas constructed to show the amoiuit of 
variation in specimens from a given locality, collected in this case from the 
neighbourhood of Kimberley by the junior author. 
1. 
2. 
3. 
4. 
5. 
6. 
7. 
8. 
9. 
10. 
11. 
12. 
Kimberley S 
81 
4 
3 
3rd and 4tli 
20 
20 
21 
\^ 
S 
98 
4 
3 
N 
ith 
X 
21 
21 
23 
A 
? 
110 
4 
4 
0 
3rd 
23 
23 
D 
X 
0 
? 
102 
4 
3 
N 
3rd and 4th on left foot, 
4tli on the right foot 
21 
21 
D 
0 
10-1 
4 
4 
X 
3rd and 4th on right 
foot, 4th on left foot 
0 
23 
24 
24 
N 
0 
$ 
106 
5 
3 
0 
3rd and 4th 
X 
20 
23 
21 
N 
0 
? 
93 
3-5 
3-5 
N 
3rd 
20 
20 
20 
D 
X 
X 
? 
1- 
i 
0 
4th 
0 
D 
0 
0 
$ 
IlO 
4 
4 
0 
4th 
X 
22 
22 
23 
N 
0 
0 
$ 
106 
4 
3 
4tli 
X 
24 
24 
25 
A 
0 
0 
