392 MR. J. B. GATENBY ON THE BIONOMICS, EMBRYOLOGY, 



air and carry the latter to nil parts o£ the body. The respiratory system is 

 identical in all the Microgasteridse observed by Seurat. It consists o£ two 

 longitudinal trunks on each side of the body, united in front by a cross tube, 

 dorsally to the digestive tube, but not united in the hind region of the body. 

 Each trunk gives rise to eleven branches, latero-doi'sally and latero-ventrally 

 situated in each segment from number two to number twelve. A branch 

 from each trunk is given off and ramifies into the anal vesicle. The nine 

 pairs of latero-dorsal trunks situated in segments two to eleven present, near 

 their place of origin and just laterally, a short accessory trunk, blind at its 

 extremity ; these accessory trunks, nine pairs in number, are the stigma 

 tubes, but during the whole of the life of the larva inside the host they 

 remain blind, the tracheal system being entirely closed. At the moment of 

 the exit of the parasitic larva from the body of the caterpillar, the stigma 

 trunks, with the exception of the second pair, open to the exterior ; the first 

 pair is in the anterior region of the mesothorux, the seven others on the 

 anterior part of the first seven abdominal segments. 



Seurat did not properly examine the larva by the section method, and 

 since he has written on so many forms in a fairly short paper, his treatment 

 of Apanteles larva is somewhat curtailed and unsystematic, but is neverthe- 

 less the best extant. Seurat's description of the tracheal system is very fine, 

 but I disagree with him in certain ways (page 399). 



' Personal Ohservations. 



a. Material and MetJiod. 



My material for this study consisted of a large number of larvse of Micro- 

 gaster connexus and a certain number of Ajjanteles glomeratus. Porthesia- 

 similis larvae were collected; they were opened, and the parasitic larvse 

 where they occurred were transferred to various fixatives. The smallest 

 caterpillars contained, on the whole, the smallest and least developed parasitic 

 larvae ; the large full-grown caterpillars contained nearly or quite full-grown 

 parasitic larvae. Bouin, Flemming, corrosive-acetic, Petrunkewitsch, Carnoy, 

 and 3 per cent, bichromate of potash were used. Whole mounts were 

 stained in Paracarmine, sections in Ehrlich^s Hsematoxylin, or Iron Hsema- 

 toxylin. For a study of the fat the Flemming-fixed larvae were very useful. 



h. The External Morphology of the Larue. 



In dissecting the caterpillars it is found that the parasites generally lie 

 evenly disposed along the length of the hsemocoel. If they are o-nly half- 

 grown tlie fat-body of the host is still well developed and the parasites are 

 mixed up among the bundles of fat-cells. In full-grown parasitized cater- 

 pillars the fat-body is either poorly developed or not present at all. The 



