66 GENERATION OF INSECTS 



to move and seemed to be about to fly away ; of the other 

 six, the one which had gone to the bottom, and three 

 others that had floated on top, gave signs of hfe after 

 three minutes by moving their legs and thrusting out their 

 proboscides, but soon after they ceased all motion, and 

 were really dead. A few days after, I made many ex- 

 periments, keeping the flies in the water for longer or 

 shorter periods; the water being sometimes iced and 

 sometimes of ordinary temperature; sometimes I held 

 the flies under the water ; then again I would allow them 

 to float. At last I ascertained that if the flies be really 

 drowned they cannot be restored by the potency of the 

 sun. Hence, I cannot see how it is possible to believe 

 Columella, who reports that bees found lying dead under 

 the beanstalks and which are kept all Winter in a dry 

 place, return to life if, during the mild season of the 

 equinox, they are exposed to the sun and powdered with 

 fig-tree ash. I have not made this experiment, but I 

 esteem it to be beyond belief. 



I return to the flies hatched in the tun-fish : these, like 

 all the others, on escaping from the shell began to ex- 

 crete the natural abdominal impurities that are due, I 

 believe, to the food consumed by them while they were 

 still worms, in which state I noticed that they never ex- 

 crete anything. The flies do not live more than four 

 days after birth, when retained in the closed jars with- 

 out food ; but this is not an unusual rule in Nature. 



What seems more strange is that spiders, born in 

 closed jars from the eggs of the same, can live so many 

 months without apparent food. On the fifth of July I 

 shut up a female spider in a glass jar covered with paper, 

 and I observed that on the twelfth she had spun herself 

 a nest on the lower part of the paper, in the shape of half 



