34 



THE PHILOSOPHY 



well as of the membranes inclofing the foetus, becomes more firm 

 and tenacious ; and this, perhaps, is the phyfical reafon why they 

 are viviparous in fummer, and oviparous in autumn. Many other 

 flies are known to be viviparous. Upon farther examination, all 

 thefe will probably be dlfcovered to be alfo oviparous *. 



The puceron exhibits another phaenomenon flill more frngular. 

 The maxim, that multiplication prefuppofed impregnation by fex- 

 ual embraces, was formerly thought to be univerfal. Neither Ihould 

 the reception of this maxim be regarded as a matter of wonder ; for 

 it was founded on a very general and ftrong analogy. But the fol- 

 lowing fads fhow, that Nature, though uniform in many fteps of 

 her progrefs, is not invariably limited to the fame mode of opera- 

 tion. 



On the 20th day of May, M. Bonnet took a young puceron, the 

 moment after dropping from the womb of its mother, and fhut it 

 up in a glafs veflel, to prevent all poffibility of communication with 

 any individual of the fpecies. A fprig of the tree on which the 

 animal was produced, fupplied it with nourifhment. The creature 

 changed its fkin four times, namely, on the 23d, 26th, 29th, and 

 31ft days of the fame month. After a minute detail of circum- 

 ftances, M. Bonnet informs us, that his imprifoned puceron grew 

 with rapidity ; that, on the ift day of June, it brought forth ; and 

 that, from this day to the 21ft, it produced no lefs than 95 young, 

 all full of life and vigour f. He frequently repeated this experi- 

 ment, and it was always followed with the fame event. 



M. Bonnet, fufpedting that a fingle impregnation might Influ- 

 ence both the mother and her immediate offspring, refolved to ob- 

 viate 



* See Reaumur, torn. 8. edit, izmo, p. 153. et feq. 



t Bonnet, Traite d'lnfeaologie, torn. i. p. 39.; and Reaumur, torn. 12. p. 353. 



