312 INSECT A. 



a distinct lumen naturally does not develop. The cell-mass of the lower layer 

 is here, from its origin onwards, solid and gradually widens out helow the lateral 

 plates. The types given are connected by transitionary forms. Thus it appears, 

 according to Graber's recent statements (No. 30), as if, in the Lepidoptera, a 

 type intermediate between the second and third types is occasionally to be- 

 observed. 



It was observed by "Wheeler in DorypJiora, and by Graber in Lina, that 

 the most posterior end of the gastrula-furrow in certain stages appears forked 

 (Fig. 145, p. 291), a condition which we are not in a position to explain. 



The cell-layer derived from the gastrula-invagination (the lower 

 layer) represents the common rudiment of the entoderm and the 

 mesoderm. It has only recently become known in -what way these 

 two germ-layers are separated from one another in the Insecta. 

 With regard to this point we must follow chiefly the statements of 

 Kowalevsky as to the Muscidae (No. 49), of Heider as to Hydrophilus 

 (No. 38), and "Wheeler as to Doryphora (No. 95). Kowalevsky 

 showed first in connection with Musca that the greater part of the 

 lower layer yields mesoderm exclusively, and that only two cell- 

 masses, corresponding respectively to the anterior and the posterior 

 end of the germ-band, are concerned in the formation of entoderm. 

 "We must therefore, in the Insecta, speak of an anterior and a 

 posterior entoderm-rudiment. As the stomodaeal and proctodaeal 

 depressions which appear as ectodermal invaginations develop, the 

 cell-masses of the two entoderm-rudiments are pushed in front of 

 them into the interior, and thus become separated from the meso- 

 derm. The two entoderm-rudiments now represent cell-accumulations- 

 closely applied to the blind ends of the stomodaeum and proctodaeum. 

 They soon broaden out into the shape of watch-glasses, with their 

 concavities directed towards each other and their convex sides turned 

 to the respective poles of the egg. Their shape, however, soon 

 changes, two lateral bands growing out from each rudiment in such 

 a way as to form the letter U (Fig. 154, en). The ends of the two 

 U-shaped rudiments are directed towards each other, and grow out 

 until they meet and fuse. The entoderm-rudiment yielded by the 

 fusion of the two U-shaped rudiments then consists of two bands 

 running longitudinally above the germ-band, mostly dorsal to the 

 primitive segments. Anteriorly and posteriorly these bands pass 

 into one another, and at these points fuse closely with the stomo- 

 daeal and proctodaeal invaginations. As these lateral entoderm- 

 bands gradually widen, they begin to grow round the surface of 

 the food-yolk on which they lie. This circumcrescence as a rule 

 progresses most rapidly on the ventral side, so that the two entoderm- 



