96 THE SNAKES OF SOUTH AFRICA. 
immediately despatched for the District Surgeon. Fortunately, 
the snake was not of a kind whose venom acts promptly, for the 
messenger had to ride four miles, and the doctor four miles 
ere medical assistance could be rendered to the sufferer. 
Dr. Wildish took with him the anti-venom discovery of the 
Pasteur Institute, and its effects have so far been successful.”’ 
Subsequently, water was poured down the hole where the snake 
was seen to disappear, and behold! out crawled a half-drowned, 
harmless Brown House Snake. 
THREE-CORNERED OR FILE SNAKES. 
These snakes, although found in most parts of South Africa, 
seem to be rather rare everywhere. Although large and formid- 
able-looking, they are perfectly harmless. Their average length 
is three to five feet. They are not easily mistaken for any other 
genus of snake owing to their peculiar vertebral or backbone 
scales, which, at first sight, gives these snakes a triangular 
appearance. The row of scales along the backbone are bi- 
carinate, which means double-keeled. These scales have a 
groove longitudinally down the centre. The colour of the skin 
in the groove is yellow, forming a long line from the head to the 
tail of the snake. These backbone scales are rounded and larger 
than those on the rest of the body. The colour of the upper 
parts is usually brown. The edges of the abdominal shields and 
lower half of last row of body scales are blotched, forming 
two rows extending the whole length of the body. The 
head is blackish in colour. Two species are officially 
recognized in South Africa, viz. Simocephalus capensis and 
Simocephalus nyass@. The reason for separating them into 
two species was because S. capensis was supposed to have two 
postocular shields and 203-241 ventrals, and S. nyasse only a 
single postocular and 171-185 ventrals. However, I do not 
agree with the separation of Simocephalus into two species. 
I have examined several specimens and find that the above 
characters which are supposed to distinguish them, are not 
constant. In some there are two postocular shields and 170 to 
180 ventrals. In others a single postocular and 203 to 240 
ventrals, thus proving that these characters cannot be relied 
upon. I have found the same thing to hold good with other 
