THE CARPS IOI 
cages, I sunk a large box in the earth and supplied it 
with a constant stream of water from a natural source. 
As there are several life histories still to be worked 
out, I will describe this sunk box method in some detail 
in the hope that it may be useful to others who feel 
inclined to investigate for themselves. Selecting a ditch 
which remained at a fairly constant level all the year 
round, I led the water from it in a 2-inch pipe to a 
box sunk in the ground. At the head of the box by 
means of a board and clay puddle, I held up the water 
a few inches. The board was pierced by two lead pipes, 
one a half-inch in diameter, which allowed the water to 
go into the box; the other, a two-inch pipe, pierced the 
board at a higher level, and to one side, and carried off 
the excess. By this means the water always ran at full 
bore through the half-inch pipe. The overflow from the 
box had to be guarded by very fine gauze, twenty 
strands to the inch, in order to prevent the roach larvee 
escaping. At first I found the scum on the surface of 
the water blocked up this fine gauze, and the water 
flooded over the side of the box. To overcome this I 
nailed a square tin sieve on to the end of the tank, so 
that two inches of the sides of the sieve were above, 
and two inches were below the surface of the water. An 
overflow hole was cut through the end of the tank and 
through the side of the sieve in contact with it. The 
scum now gathered on the solid sides of the sieve and 
on the end of the tank, and the water welling up through 
the gauze which formed the bottom of the sieve escaped 
