26O 
volution was retarded in the summer season, and, 
instead of being changed into butterflies in July, they 
still retained their pupa state in the month of No- 
vember. In this manner, they were kept more than 
a year in their pupa state, and this mode of existence 
might probably be prolonged for many years, with- 
out affecting their power of reviviscence, especially 
if they were kept in an ice-house, where the tempe- 
rature did not me above zero *. 
549. That air is necessary to the development 
of these animals, may be inferred from the experi- 
ments of the same author. In July, Reaumur con- 
fined the pupae of butterflies and moths in glass 
tubes four or five inches long, which were hermeti- 
cally sealed ; and, in these circumstances, they re- 
mained in their original state for more than five 
months, without indicating any appearance of deve- 
lopment t But when Spallanzani kept the pupas 
of silk-worms in larger tubes, which were likewise 
sealed, their evolution proceeded, and they attained 
their perfect insect state. Similar observations were 
made on the pup^s of flies and butterflies J ; and we 
may therefore conclude, that fresh air is necessary to 
carry forward the change of the insect from its pupa 
to its fly state, By subsequent experiments, Spal- 
lanzani also found, that the oxygen gas of the air 
was consumed in the transformation of the pupa to 
the insect; that carbonic acid was produced, but that 
* Mem. &c. ton), ii. p. 18. 
t Mem. torn. ii. p. 6. 
$ Tracts, &c. p. 200. 248. 
