400 ME. J. B. GATENBY ON THE BIONOMICS, EMBKYOLOG?, 



nature has been elucidated (see page 411, where I have mentioned a possible 

 explanation of their nature). In PI. 26. fig. 25 these glands are drawn in 

 a very young lurva, and their position of insertion into the vesicle wall and 

 not into the gut was very clear. In PI. 26. fig. 22 these glands are cut across 

 at Tx, below, or to the sides of the hinder region of the gut. In PI. 26. 

 fig. 23 is a high power drawing of the upper part of one of the tubes in 

 fio-. 22 Tx. There are about three flattened cells in the section, their cyto- 

 plasm is somewhat striated, and in the lumen is an irregular fringe or 

 intima (C). Whether it is chitinous or not I cannot say. The main point 

 to notice is that these glands in Microgaster do not join on to the hind region 

 of the gut, as claimed by Seurat. 



Embryonic Membranes of Microgaster. 



In PI. 26. fio". 16 is drawn the youngest larva I found. The larva is 

 externally completely segmented, but it is covered by an embryonic mem- - 

 brane consisting of large stretched cells. Likewise the larva in PI. 26. 

 fig. 25 is covered by a cellular membrane (S). In the larva in PL 26. fig. 16 

 there are also at A a number of hypertrophied loose cells. Not having any 

 stages earlier than that in PI. 26. fig. 16, it is not possible to say for certain 

 what these loose cells represent ; possibly the outer membrane is the serosa, 

 the hypertrophied cells at A, the amnion. In certain Platygasters (15) the 

 eo-o- divides into parts, one of which gives rise to the embryo, the other to 

 hypertrophied cells. In Microgaster the outer membrane, which is probably 

 a serosa, may be found in larv?e up to 2i mm. in length, and may act ns 

 a medium for nourishing the embryo up to time when it is ready to begin 

 feeding on the fat-body of the caterpillar. 



The proper identity of the membrane or membranes and their method of 

 orio-in could not be worked out because my material was all too far advanced. 

 However, it is important to recognize that the embryonic membrane persists 

 for a remarkably long time during which the larva does not feed, all nourish- 

 ment beino- derived through the instrumentality of the membrane. This 

 is undoubtedly a specialization due to the larva's mode of life. It should also 

 be noticed that though the membrane envelops the larva's body much later, 

 this does not say that food is not being taken in during later stages, for the 

 larva could bite a hole in the membrane. I think food is not taken in till the 

 larva is 2 mm. in length. Full-grown larvge are some 5 mm. in length. 



Tlie Segmentation of the Larval Microgaster conuexus. 



The question of the segmentation is one of the greatest difficulty to 

 elucidate, and special care has been taken by me to ascertain the number 

 of segments. 



If the adult Apanteles larva after emergence from the Pieris caterpillar 



