Soidh A frican Acjamasi allied to Agauia lii.^jnda a7i(l A. afra. 2S5 
Touw's River ? . Dorsal scaling- not so coarse as in specimens from 
Mossel Bay. 
Conclusion. 
South Africa presents many zoogeograpliical problems, luit none is so 
intricate as that of the distribution of the genus Agauia. 
As regards A. Irispida, it seems that a line drawn along the twenty- 
sixth meridian indicates, roughly, the limit between the ranges of the vars. 
armata (e.istern) and acnleafa (western) ; nowhere do these two forms 
overlap. At the extreme east, from the Zambesi to Natal, vai*. armata 
exists alone, Init towards the west, from Rhodesia to the noi tli-eastern 
corner of tlie Cape Colony, it is accompanied or replaced by var. distant i, 
which is the prevailing form in the Transvaal and the Orange Free State. 
In the west, from 25° hit., var. aculeata is accompanied by var. hrachyura , 
which, from 30° hit., passes gradually into tlie typical hinpida in the south 
— the only form occurring at the Cape. 
A. anchietae and its varieties nwthneni and liuolttJi nre N.W. forms. 
A. atra is a more southern species, not recorded within the tropics ; at 
the extreme S.E. of the Cape it is accompanied by var. rtulis, which south 
and east of the great mountain ranges replaces the typical form, extending 
as far to the north as Delagoa Bay. 
It was at first thought that it might be possible to co-relate the distri- 
bution of the several varieties with the various botanical regions as set 
forth by Dr. I. B. Pole-Evans in the Official Year Book for 1917, but when 
a map had been constructed for the Agamas it was found that some of tlie 
varieties, e. g. A. aculeata, had such a wide distri])ution that co-relation was 
impossible. 
In some cases, however, distribution is restricted to comparatively small 
areas, and it may be possible that a systematic survey will show that these 
forms are confined to certain localized botanical areas. In the absence of a 
detailed survey it is not easy to say, and this side of the subject needs 
thorough investigation by reliable collectors on the spot. 
Thus in regions of sandy waste there may be little oases of grass and of 
thorn trees, with var. hrachyura inhabiting the former and var. aculeata in 
possession of the latter. 
We have, for example, recorded var. hrachyura from Aus and from 
Worcester. This at first sight seems to preclude the possibility of a regional 
distribution, but on reference to the above-mentioned map of the botanical 
regions it will be found that there is a fair-sized patch of country, just south 
of Worcester, indicated as being the same as the country round Aus. 
We have recorded varieties aculeata, distanti, hrachyura and hispida 
from Little Namaqualand. 
Dr, L. Peringuey describes this district as follows : " From Port JSToUoth 
