Annals of the Transvaal Museum 
119 
When undisturbed they perch on the rocks and display their splendid 
colours in the bright blue head and neck, ana in the yellow back. 
Agama aculeata occurred in great abundance on the sandy plains : 
here it was quite an inconspicuous animal, its colours harmonizing with 
those of the ground: it was also found on bushes, but the arboreal 
individuals were usually characterised by the most magnificent colours 
which rendered these creatures as conspicuous as large flowers."' 
We are inclined to suspect that these gay colours are only assumed 
during the breeding season, and that at other times all the individuals 
both $ and $ are protectively coloured. The preference for arboreal life 
exhibited by Agama aculeata when in breeding costume may perhaps be 
due to the fact that a conspicuous lizard is safer in trees than on open 
ground: it is also probable that the colour of the vegetation has a 
stimulating effect on that of the lizard. It is scarcely possible to believe 
that the gayly coloured arboreal individuals of A. aculeata and the equally 
gorgeous rock frequenting ones of A. atra are in any wise concealed from 
their enemies. Such individuals of A. atra are hard to catch and bite 
fiercely when molested: on the other hand protectively coloured 
$ specimens found on the ground are easily caught. 
A few reptiles in Namaqualand would seem to be under no 
circumstances protectively coloured. A striking example of this type is 
found in the small lizard Cordylosauvus trivittatus which has a distinctive 
pattern of strongly contrasted colours of black and light buff and a bright 
blue tail : the latter part of the creature is extremely conspicuous, and 
is very easily broken when the animal is handled. In its natural haunts 
it is certainly one of the most brilliantly coloured animals of Namaqualand 
and its colours are obviously in direct contrast with those of its 
surroundings. This creature which is very swift may be seen running 
about in rocky places during the heat of the day. It may be that this 
species affords an example of warning colouration : but on this point 
there is no real evidence (see however page 140 footnote), and its alertness 
is decidedly against such an explanation of its striking colours. On the 
other hand its colouration may be epigamic though the sexes are alike : 
to this principle Professor Poulton (in lit.) refers the colouration of 
flamingoes. f 
The black variety of the black-necked cobra, Naia mgricollis. is 
another example of a reptile which is not protectively coloured and which 
makes no attempt to conceal itself. In this case it is possible that both 
aposematic and epigamic influences have operated. $ and 5 cobras often 
differ greatly in colour the <J<J being more conspicuous than the $§ : the 
specimen we took was a 
* Agama distantly which, like A. aculeata , is in all probability one of the geographical 
forms of A. Jdspida has according to observations made by Master JB. Penfold and myself at 
Pretoria the power of changing its colours to some extent so as to match its surroundings. 
Master Penfold has written to me that “ A . distantly though the usual colour is sandy brown, 
can develop a dark brown or a light brown body with blue head or cheeks and an orange 
chest or a plain brown body ’with a few dark markings.’ * 1 A dark coloured specimen which 
I caught in the winter of this year near Pretoria on dark soil which it matched very 
closely, was placed upon a light grey rock which was here and there covered with an orange 
coloured lichen. After a short time the animal changed its colours, the body becoming 
light and the head and anterior parts of the body developing in places a chestnut brown 
colouration : the flanks also became lighter and were seen to be spotted with orange. — P. A. M. 
f We wish here to express our thanks to Professor E. B. Poulton for his kindness ip 
writing to us at length on the phenomena here alluded to. 
