137 
A FISH THAT PREYS ON MOSQUITO LARV&# IN SOUTHERN 
NIGERIA. 
By Dr. W. M. Grananm, 
Director of the Medical Research Institute, Lagos. 
During a seach for mosquito larve at Yaba, near Lagos, my attention was 
drawn to their complete absence from a swamp about a mile from the Medical 
Research Institute. The pools in the swamp appeared to be well suited for the 
growth of mosquito larve, and I caught female mosquitos full of eggs in the 
adjoining fringe of bush. The absence of larve was therefore evidently not due 
to any lack of fertile female mosquitos. An examination of the pools showed 
that they all contained small, active, surface-feeding fish,* belonging to the 
family CypRINODONTID”, and it seemed probable that it was by their agency 
that the pools were kept free from larve. 
To test this hypothesis, a number of the fish were caught and taken to the 
Laboratory in a large jug of swamp water, and the contents of the jug were then 
emptied into a basin and put aside, as it was getting dark. The following 
morning a number of the fish were found dead on the floor, owing to their having 
jumped out of the basin during the night. As these fish can leap a distance of 
from one to two feet, in all future experiments it was found necessary to keep 
them in covered vessels. 
On watching the fish in their native haunts, the importance of this faculty of 
leaping becomes plain, for it is by this means that they are enabled to pass from 
pool to pool. Thus every pool ina swamp is likely to be visited, and the whole 
food supply will be better distributed among the fish. No other species of fish 
were found in the swamp, and the Haplochilus itself was absent from a dug-out 
water-hole near the swamp, in which I found tadpoles. 
At the Laboratory, the Haplochilus were found to eat mosquito larve greedily. 
Even when a hundred larve were introduced into a vessel with a dozen fish, all 
the larve had disappeared in an hour. I was unable to discover that preference 
was shown for any particular species of larve, though I tried the fish with one 
Anopheline, and five species of Culicine larve. They declined, however, to eat 
pup of any species, though they nibbled at them occasionally. Probably these 
fish are unaccustomed to see pupx, for in the pools they inhabit the larve 
would be very unlikely ever to reach that stage of development. 
* Mr. Boulenger has supplied the following notes on the structural characters of the species, 
for which he proposes the name Haplochilus grahami, sp. n. :— 
“ Allied to H. senegalensis, Steindachner. Depth of body, 4 to 44 times in total length, length 
of head, 3 to 3} times. Snout a little shorter than eye, which is 3 to 34 times in length of head, 
and 1} times in interorbital width ; lower jaw projecting. Dorsal with 7 rays, originating about 
twice as far from occiput as from root of caudal, above posterior fourth of anal. Anal with 15 
or 16 rays. Pectoral, { length of head, extending beyond base of ventral ; latter small, equally 
distant from end of snout, and from root of caudal. Caudal rounded-acuminate, longer than 
head. Caudal peduncle slightly longer than deep. 28 or 29 scales in a longitudinal series, 22 
round body in front of ventrals ; lateral line indicated by a series of pits.” 
