LARViE, ETC., OE OPHIOCEPHALUS STRIATUS. 
121 
time, for although the little fishes called “millions” in the West 
Indies, top-minnows elsewhere, are not such a universal panacea for 
malaria as it has been suggested they would prove to be, yet it is 
useful that renewed attention should be given to the part played by 
fishes in general, by the fry of the larger fishes, and by top-minnows 
in particular, in checking the natural increase of mosquitoes by 
feeding upon their larvae. 
Mosquito larvae too often thrive best in places to which fishes 
can never have access, as in the small accumulations of chocolate- 
coloured water in the crevices of jungle trees ; and even in the tanks 
they favour the scum-covered water at the edge, where the vegetation 
is rotting, though this of course will vary with the water-level in the 
tank at a given time. An analogous case is afforded by the Colombo 
Lake fly, an undetermined species of Chironomus , whose numbers 
have not appreciably diminished in spite of all the precautionary 
steps which have been taken during the past five or six years, due, 
it must be supposed, to the fact that although their larvae are 
naturally useful as fish-fodder, yet they thrive best in places which 
are not otherwise attractive to insectivorous fishes. In such a case 
as this the normal balance cannot be restored without the applica¬ 
tion of herculean methods. 
Next to the microscopic Crustacea which compose the most 
important source of food-supply for the young of freshwater 
fishes, the larvae of mosquitoes are looked upon as the best friends 
of the fish-culturist, although the mature insects are amongst our 
worst enemies.* We are thus placed, often at one and the same 
time, in perplexity as to how we may procure the larva3 and how 
to avoid the flies. In any case, it is expedient to recognize 
as beneficial those species of fishes which prey upon mosquito 
larvae. 
Wishing to test the selectivity of mosquitoes for different waters, 
I placed two chatties side by side, one (referred to as A) containing 
water rendered turbid by decaying animal matter, the other (B) 
containing clear water. Two days later I took 20 culicine egg-rafts 
from A, as against 1 from B. On the following day 11 more egg- 
rafts appeared on A ; none on B. On the day after this, again 14 
rafts were found on A, only 2 on B, the water not having been 
changed in the meantime. At the same time another species was 
observed to lay single eggs which adhered to the surface of chatty A 
below the surface of the water close to the water’s edge. Water 
containing decaying vegetable matter is equally attractive to egg¬ 
laden mosquitoes. Such a pronounced preference for a putrescent 
nidus for the eggs on the part of insects whose feeding habits are so 
highly specialized, amounts to a biochemical reaction to which the 
term saprotaxis might be applied. 
* Compare Dr. Emil Walter. Die Fisherei als Nebenbetrieb des Landwirtes 
und Forstmannes. Neudamm', 1903, see pp. 57-62. 
