140 CHINESE MODE OF REARING SILKWORMS. 
to be sometimes entirely without food. At the 
end of a week, those which had been well fed had 
acquired about a third more in bulk than them that 
were kept on short allowance. About the twelfth 
day we had carried our starving system so far, that 
we had forgot to put leaves to them for nearly 
twelve hours, after they had been completely ex- 
hausted ; and next morning we found most of the 
caterpillars adhering to the sides of the tumbler, by 
very fine silky filaments. We were apprehensive, 
that, having not been long enough fed, that they 
might die in place of transforming into the chrysalis, 
and loosened them all from the sides of the tumbler, 
and gave them a fresh supply of leaves, which, in 
a few minutes, they began to devour very greedily. 
Next day, while in the act of supplying them with 
leaves, we found one of the larvee again adhering 
by filaments to the piece of paper with which the 
tumbler was covered, and determined to leave it 
undisturbed. It continued in this state for upwards 
of a day ; and when we next examined it we found : 
that it had changed into the chrysalis condition, but 
was of a very small size, and had a thin transparent 
appearance. The caterpillars which had been pro- 
perly supplied with food now refrained from eating, 
and prepared for transforming into chrysalides. 
Those which had been starved still continued to 
eat, and were much less in size—a little more than 
a fourth. When the full fed larvee had all changed, 
