OP OECANTHUS AND TELEAS. 265 



in the fresh state no such cells are to be seen. These occur at irregular intervals and 

 are found only when the blastosphere shows traces of having been ruptured by the swell- 

 ing influence of reagents. (PL 23, fig. 35.) In all stages the parasite is surrounded by a 

 mass of clear protoplasmic substance which has been affected by the presence or secretions 

 of the parasite. On sections of the first stage one finds surrounding the blastosphere, but 

 at some distance from it, a thin wall of condensed albuminous matter containing within 

 its substance no indications of cell structure. This layer cannot be made out in the fresh 

 state. It is not found in sections of the two subsequent stages of the parasite. There are 

 no other layers to be found enveloping the parasite in any stage observed. 



This stage 1 agrees perfectly with that observed by Metschnikoff in another species of 

 the genus Teleas, with the single exception of the absence in the present case of any cellu- 

 lar embryonic envelope such as both Ganin and Metschnikoff describe for Teleas and other 

 genera of Pteromalidae. Since the segmentation of the egg was not observed in this case, 

 only negative evidence can be produced to prove that all the resultant cell elements go to 

 form at first a single-layered blastosphere. It is, however, highly probable that such is 

 the case, since at no stage of the blastosphere can any such elements (i. e. embryonic mem- 

 branes) be found. Since this is the stage in which the embryonic membranes (i. e. amnion 

 and serosa) are presumably beneficial to the growing embryo, it must be concluded that 

 all such structures are wanting in this particular case. I hesitate to apply these conclu. 

 sions to any other forms of the Pteromalidae, for the observations of Metschnikoff and 

 Ganin clearly pi*ove that these forms differ greatly among themselves in the particulars of 

 their embryonic development, though bearing very strong resemblances to each other in 

 their older larval and adult states. We have in this form, then, an insect which has no 

 trace of embryonic membranes such as are found in all other insects. Here also is a case 

 in which no invagination takes place to form the mesoderm as is held to be true for most 

 insects. This is the more significant since there is no excess of yolk matter to hinder the 

 process. The facts obtained from the study of this parasite respecting the origin of the 

 mesoderm should not lead to a misinterpretation of the process. For although the conditions 

 favoring its origin by the typical process of invagination (a blastosphere filled with fluid) 

 are realized, yet in this particular instance it is to be viewed not as a primitive but 

 rather as a secondary method acquired in ancestors where an accumulation of nutritive yolk 

 substance induced a modification of the primitive process, — a modification which has been 

 retained although the conditions which led to it have ceased to exist. The key to this inter- 

 pretation is to be found in the tact that the eggs of the non-parasitic Hymenoptera are 

 supplied with an abundance of food material, and in such cases the mesoderm arises in a 

 manner approaching the typical invaginate form. The loss of nutritive yolk in the eggs 

 of the degraded forms is evidently correlated with their parasitic habits. 



The first differentiation of the blastoderm begins as a linear thickening of the embryonic 

 area (germinal band of other insects), and is followed by the formation of a median groove 

 (Primitivfurche) which divides it into lateral ridges (Primitivwulste). The mesoderm 

 and endoderm are derived from the ectoderm by a process of cell budding which resembles 

 typical delamination in that it takes place from all portions of the inner surface of the 



1 In the figures the outer boundary of the protoplasmi mass surrounding the parasite is indicated by a line. In 

 some figures the adjacent yolk globules are sketched in. 



MEMOIRS POST. SOC. NAT. BIST. VOL. III. 34 



