576 WILLIAM PATTEN. 



importance to observe the complete disappearance of the 

 gastrula invagination and the existence of a stage follow- 

 ing it where there is no median furrow. The neural 

 invagination^ as we shall soon see, is a second and later process, 

 and although it occurs exactly in the place where the gastrula 

 mouth disappeared, it has no connection with it whatever. 

 The definite comprehension of this fact will, I believe, prove 

 quite as important a step towards the comprehension of the 

 phenomena of insect development, as has the similar distinc- 

 tion between primitive groove and medullary furrow in the 

 case of Vertebrates. This distinction between the neural furrow 

 and the gastrula has not heretofore been clearly recognised in 

 insect embryos. Zaddach surely confounded them, and I find 

 no evidence that any observer has called attention to the fact 

 that there are two fundamentally distinct invaginations occur- 

 ing successively in the middle of the ventral plate. 



The mesoderm at the beginning of the present period is 

 unsegmented, and forms a continuous layer of cells beneath the 

 ectoderm ; it soon separates along the median longitudinal 

 line, thus forming a pair of lateral mesodermic bands, each of 

 which at the same time divides into segments or somites, 

 which, however, do not contain a lumen or body cavity as 

 described by Hatschek and Kowalevsky, but are simply formed 

 by the apposition of two cell layers, which later split into the 

 splanchnic and somatic mesoderm. 



The ventral plate, as seen in the living egg at this time, 

 shows a gently undulating surface, the external evidence of the 

 segmentation of the body, the depressed areas alternating with 

 the mesoblastic somites. The formation of the external depres- 

 sions and the segmentation of the mesoderm occur at the same 

 time. 



In many'places in the mesoblastic somites, cells may be seen 

 which present all stages of migration from the mesoderm into 

 the yolk. Some of these mesoderm cells can be distinguished 

 from those surrounding them by having larger and more 

 granular nuclei; others are partly separated from ths meso- 

 derm being still united with the latter by a delicate filament of 



