166 Transactions of the Boyal Society of South Africa. 
Atkactaspis rosteata Giinth. ; B.M. Cat., III., 514 (? = A. dueedeni 
Gough). 
Genesa (J. Hill). 
The specimen agrees in colour with Smith's figure of hihroni except 
that the head is not pale. The rostral shield has a sharper horizontal 
edge than in hihroni, and, moreover, the shape of the chin shields and of 
the third lower labial is in agreement with the figure given by Peters in 
the Eeise nach Mossambique " as hihroni (but referred to by Boulenger 
as rostrata). There are only 184 ventrals — but the specimen had been 
chopped in two, and possibly a portion had been lost. 
A. duerdeni Gough (Eecords Albany Museum, II., 178) is based on 
a juvenile specimen, and the figure is considerably enlarged; the total 
length, not given in the original description, is 250 mm. The snout is 
rounded rather than prominent, and has no sharp horizontal edge ; 
but such a character may not be the same in juveniles as in adults. 
Note on the Phenomenon sometimes referred to as Mimicry 
OR AS Protective Colouration in Snakes. 
Some remarkable instances of the similarity of colour pattern amongst 
snakes which are not genetically allied have been described by Dr. K. 
Sternfeld (Mit. a. d. Zool. Mus. Berlin, 1910, p. 58) in dealing with the 
ophidian fauna of German S.W. Africa. He points out that Dasypeltis 
scahra, Bhamphiophis multimaculattts and Bitis caudalis, which are all 
common in German S.W. Africa, have the same type of colouration, and 
inasmuch as the last mentioned is a poisonous viper, and the two former 
harmless, he believes it probable that this is a case of Batesian mimicry. 
As further support to this theory he states that in Upper Egypt, where 
Bitis caudalis does not occur, Dasypeltis scahra has another type of 
colour pattern, that of Echis carinata, which is also a viper. 
In dealing with the Kalahari collection belonging to the Kimberley 
Museum, we also met with the same phenomenon. In the same localities 
along with Bitis caudalis there were found Bhamphiophis multimacidatus 
and the young of Pseudaspis cana, both of which, especially the former, 
have much the same colour and pattern — roughly speaking, dark blotches 
on a dull pale background — as Bitis caudalis. 
Dasypeltis scahra, though absent from that collection, is common 
in certain parts of the Kalahari, and should be included in this so-called 
mimetic association : perhaps also we might add Boodon guttatus and 
Typhlops schinzi. 
The mimicry explanation will of course imply that in past times 
the harmless snakes were very extensively eaten by birds or other 
foe which would not take vipers ; but such discriminating foes are not 
