POLYEMBRYONY AND THE DETERMINATION 01' SEX. 3l7 



accomplishing the formation of an individual conforming to the 

 specific type. 



Another question which arises is that of determining whether, in 

 the class of insects, polyembryony may be considered as having pre- 

 ceded or followed phylogenetically the other modes of agamic re- 

 production, such as the pedogenesis of the Cecidomyidse or the cyclic 

 parthenogenesis of the Aphides and the Cynipida?. Harmer, from 

 the Bryozoa, arrived at the conclusion that embryonic division may 

 be a consequence of the blastogenic faculty of the adults. Perrier 

 looked in the same way upon all the budding animals. 



Considered from this point of view, the polyembryony of the Chal- 

 cididae appears, not as an initial phenomenon, but as a secondary 

 adaptation due to an acceleration of embryonic j^rocesses {T achy- 

 genesis of Perrier, 1902). The object of this adaptation is to ac- 

 complish the preservation of the species by i^ushing its multiplication 

 to the highest limit possible, since the existence of the adult Encyrtus 

 is short and precarious. 



As to the determining cause of the division of the germ, it is, 

 according to Marchal, in the sudden precipitation of more dilute 

 liquids into the midst of the nourishing medium and in a concomitant 

 modification of osmotic exchanges to the interior of the cells. It is 

 to be noted that in Encyrtus polyembryony actually reaches its great- 

 est intensity at the moment Avhen the caterpillar of the Hyponomeuta 

 begins to feed (first days of April) and in Polygnotus at the period 

 Avhen the young larva of the Cecidomyia gorges itself with sap. Now 

 the production of sudden changes induced by osmotic pressure con- 

 stitutes precisely one of the processes employed to bring about the 

 separation of the blastomeres and their develoi^ment into a number 

 of distinct individuals, as has been demonstrated by the already 

 mentioned experiments of Loeb and Bataillon." 



Furthermore, connected with polyembryony is the question of the 

 determination of sex, and in this particular it offers a special interest. 



I have already observed, in the course of my studies on Encyrtus 

 (1891, p. 527), that all the individuals emerging from the same cater- 

 pillar appertained, in most cases, to one sex only.'' A total of 21 

 carefully controlled observations gave me 5 times males exclusively, 

 9 times females exclusively, 3 times a great majority of males, 1 time 

 a great majority of females, 3 times males and females in nearly equal 

 numbers. 



Marchal has likewise observed that the Polygnotus that issued 

 from a single larva of Cecidomyia nearl}^ always belong to the same 

 sex (1904, p. 314). 



o See on the subject of the influence of pressure, Bataillon, 1900a. 

 & The observations relating to the parthenogenetic reproduction of Pterornalus 

 puparum are recorded in the memoir of Howard. 



