248 PROPAGATIOX OF FISH. 
be taken to remove the sediment, fungus and minute 
plants as they form, together with all eggs that turn 
white and die. The boxes should be covered with an 
open-work cover, either of lattice-work or wire, to keep 
out leaves. A large species of sieve, of galvanized wire, 
can be obtained at the stores, and answers well for this 
purpose, and also to make doors to the breeding-boxes 
and to be fitted over the outlets or communications be- 
tween tbe breeding-ponds, to keep tlie fish separate. 
The water should be four inches deep in the boxes, 
and its temperature affects the rapidity of development. 
A thorough examination should be made every few days, 
and if the sediment increases beyond control, the eggs 
may be bodily removed iniro a clean box. When the 
young appear they may be left in the boxes or allowed 
to escape into the pond at large, and will take care of 
themselves if there be no larger fish around ; in the lat- 
ter case expect to see them no more. If regularly fed, 
large numbers can be kept without trouble or danger in 
narrow accommodations until they attain a respectable 
size. 
Since writing the above I have had an oj)portunity of 
examining Fry's excellent and thorough little work on 
fi.sh-breeding, which, though principally a translation, 
exhausts the subject in its present stage, and contains all 
the requisite instruction. It is recommended to place 
the fecundated ova on willow hurdles suspended in boxes 
an inch or thereabout beneath the water, so that they could 
be conveniently removed and examined. An excellent 
substitute would probably be, what can be found univer- 
sally throughout our country, a champagne basket. It 
