POLYEMBRYOXY AXD THE DETERMIXATIOX OF SEX. 311 



of the division of the genu into numerous distinct individuals (hiring 

 the nioruhi stage. 



4. One egg of Hyponomenta ordinarily does not receive more 

 than one egg of Encyrtus. However. Avhen it happens that one egg 

 of Hyponomenta is stnng two or three tunes (perhaps by ditferent 

 individuals) a corresponding number of chains of embryos are formed 

 within the caterpillar. 



5. The albumino-fatty (nutritive) mass in which the embryos are 

 immersed results from the proliferation of the amniotic cells fur- 

 nished by the germ of the Encyrtus (derived from the paranucleus). 



6. The enveloping membrane, as well as the epitheloid cells which 

 clothe its interior, is formed at the expense of mesenchymatous ele- 

 ments furnished by the organism of the host (Hyponomenta). These 

 formations can be compared to an adventitious cyst destined to isolate 

 the parasites. 



It is on the eggs of Ilyponomeuta maJineUa that the act of oviposi- 

 tion of Encyrtus was observed for the first time ( 1897). 



Marchal, having inclosed an apple branch in a cage of gauze, 

 placed within the cage some of the cocoons of the moth. The moths 

 issued on the last days of June and the first days of July. On the 

 4th of July a goodly number of copulations were achieved: on 

 the 6th one could already observe many freshly deposited egg masses 

 on the twigs. On the ISth a large number of Encyrtus having 

 emerged fi'om the parasitized caterpillars deposited in the cage, 

 Marchal noticed that at 1.50 in the afternoon (at the moment when 

 the sun sends forth its hottest rays) an Encyrtus. which, perched on 

 an egg mass of Hyponomenta, appeared to be in the act of oviposition. 

 Taking advantage of so favorable an occasion, he could, during four 

 consecutive hours, follow with the lens the minute parasite, which 

 completely absorbed in its work, passed from one egg mass to another, 

 piercing the eggs with its sting. The operation lasted each time a 

 little more than half a minute (two minutes toward the end of the 

 journey). 



Other observations were carried on with the parasites of H. 

 mahaJehdeUa. As this moth emerges later than the others. Marchal 

 could, thanks to this circumstance, obtain new ovipositions of E. 

 fuscicoUis during the period comprised between the 12th and the '2-2d 

 of August, and complete at the same time the necessary material for 

 his work. The author concludes from his last observations that 

 Encyrtus does not live more than ten days in the imago state. 



The investigation of the egg of Encyrtus within the egg of the 

 moth is extremely difficult, if one limits one's self to dissociating the 

 vitellus. Marchal employs most especially the method of section 

 cutting. Having collected, on September 10, 1901, some parasitized 



