354 SALMON AND TROUT. 
it has altered very little in appearance from the new-born 
larva. 
Before treating of the next change, from nymph to sub- 
imago, there are two points requiring consideration, and on 
these two points, unfortunately, very little reliable informa- 
tion can be obtained. They are—/rs¢ly: the length of time 
intervening between the depositing of the eggs and the appear- 
ance of the winged subimago on the water ; and secondly : the 
nature of the food on which the insect subsists during the 
larval and nymph stages. 
When, with the kind assistance of a friend, I first succeeded 
in hatching May-fly eggs in captivity, we entertained strong 
hopes of being able to work out 
these two important questions 
of the life-history. Much time 
was consumed in daily micro- 
scopic examination of the eggs 
as the gradual development of 
the embryo proceeded. To 
provide as far as possible in 
captivity the same circumstances 
and the same surroundings as 
the larve would have experi- 
LARVA. NYMPH. enced in the natural state, mud, 
gravel, and weeds from the river 
were distributed in the troughs in which the eggs were hatch- 
ing. A grave cause of anxiety was that, of course, we could 
only use London water, and very possibly in the filtration which 
it undergoes the most necessary food for the young larve might 
be removed. All our efforts, however, were in vain. In a few 
weeks, out of many hundreds of thousands of eggs hatched 
not a single living specimen could be found. 
Although much disheartened at this first failure, and having 
been in two consecutive ‘seasons since prevented from trying 
the experiment again, I do not confess myself beaten, and fully 
expect some day to succeed in rearing full-grown May-flies from 
