South African Agamas allied to Acjama Jiispida and A. aira. 251 
The scaling reaches its maximum roughness in specimens from Otto's 
Hoop (which are almost as rough as typical specimens of A. hlspida) ; 
furtlier north these give place to a variation with a lepidosis not nearly as 
rough, and on the north-west the scaling becomes almost typical of A. 
aculeata (Plumtree and Bulawayo). These again give place on the north- 
east to a group which merges gradually into A. armafa. 
The material examined by the junior author shows conclusively that not 
one of the characters adduced for the specific distinction of Agama distanti 
is constant. Yet by using the following characters in comlnnation it may 
be possible to distinguish the form as a variety. 
From A. hispida, l\v the larger ear-opening (with about 20 per cent, 
exceptions), ])y the less tubercular head-scales, with or without a shallow 
pit in the pineal scale, fewer scales round the body (85 to 115, usually at 
least 90, instead of 70 to 85), smooth or feebly keeled ventral scales. 
Whenever the ventral scales are strongly keeled (1 specimen Pienaar's 
Eiver, 1 Potchefstroom, 1 Pretoria, 1 Witpoort, 2 Doornkop), either the ear- 
opening is larger than in A. hispida or the head-scales are less tubercular. 
The shorter fourth toe in conjunction with the larger ear-opening will in 
most cases suffice for the distinction from A. hrachyura, which has 80 to 90 
scales round the body. 
From A. aculeata by the shorter fourth toe, but when, as rarely happens 
in A. distanti, the third find fourth toes are equal, it will be necessary to 
resort to the generally unequal size of the scales on the tibia and the shorter 
tail of the latter. But uncertainty will result in many cases, and it must 
be admitted that the identification of certain specimens such as a few from 
Kimberley will be arbitrary. 
From A. arniata the distinction will be more difficult and more arbitrary 
still. We have based it on combinations of characters: shorter hand, less 
slender toes, more irregular lepidosis of the back and tibia (85 to 115 scales 
round body versus 73 to 100), but with unsatisfactory results, and it seems 
hopeless to attempt a comparative definition of the two forms except as 
applying to the majority of the specimens from a given district. 
Such being the state of things, we are driven to the conclusion that 
A. distanti cannot be regarded as a species, and that its claims to rank as a 
variety can only be accepted in so far as that conception expresses an ill- 
defined grade in evolution. 
According to P. A. Methuen this form is generally found on the plains, 
but sometimes also in valleys or at the bases of kopjes, where stones may 
be present. It occasionally climbs the thorn trees at a height of six feet 
or more. 
The food consists largely of termites, for which it burrows into the 
nests, 
