THE FORMATION OF THE GERM-LAYERS. 



311 



A gastrula- furrow of this nature has been found in a great variety of Insecta 

 by most of the recent workers on this subject. It must therefore be regarded as 

 occurring universally. The fact that it was missed by Korotneff (No. 47) 

 in Gryllotalpa is of no great importance. In the same way, the negative result 

 obtained by Witlaczil (No. 9S) in the Aphidac is hardly to be accepted when 

 we consider Graber's account (No. 24) of the development of Pyrrhocoris and 

 the recent researches by Will (No. 97) on the Aphidae, in both of which we 

 find a gastrula-furrow described. 



Many variations in detail are found in the process of gastrulation in the 

 Insecta. It is not always accompanied by the development of such a distinct 

 tube as that found in Hydrophilus. The invagination appears in individual 

 cases to be less distinct and variously modified, so that three different types 

 may here be established. 



1. There is actual invagination, accompanied by the formation of a tube 

 (Fig. 158 A, p. 321). The central region of the ventral plate (the middle plate) 

 becomes invaginated, its lateral margins standing up as a couple of folds, each 

 composed of a double layer of cells. These folds now grow towards the middle 

 line and unite, thus forming a tube {Hydrophilus, Musca, Pyrrhocoris, etc.). 

 Finally the cells of this tube lose their epithelial continuity and, becoming 

 slightly separated, assume an irregular polygonal shape. 



2. The middle plate may be overgrown by free ectoderm-folds (Fig. 153). 

 The middle plate does not separate from the ectoderm of the germ-band through 

 the union of folds as in 1 but, at 

 the place where the lateral folds 

 arise in other cases, the connection 

 between the ectoderm and the 

 middle plate becomes broken, and 

 the free edges of the ectoderm 

 grow over the sunken middle 

 plate towards the middle line. 

 Here also the cells of the middle 

 plate only lose their epithelial 

 continuity later. This type 

 occurs in various Hymenoptera 

 and Lepidoptera. It has been 

 observed in Apis by Kowalevsky 

 and Grassi (No. 32), and in 

 Lepidoptera by Kowalevsky 

 (No. 48), whose statements were 

 confirmed by Bobretzky 

 <No. 6). 



3. The lower layer may originate by ingrowth of cells from a median groove. 

 The cells of the future lower layer here lose their epitheloid nature at an early 

 period. A median groove is formed as in the other types, but there is no 

 separation of a distinct median plate or tube, single cell-elements separate from 

 the base of the groove and shift below the ectoderm ; this proliferation of cells 

 goes on until the lower layer is complete, its cells wandering below the ectoderm 

 and the lateral parts of the germ-band. This type appears to occur in the 

 Aphidac, according to Will (No. 97), and in the Phryganeidae, according to 

 Patten (No. 65). 



In the second and third types of formation of the lower layer, a tube with 



Fig. 153. — Transverse sections through the germ- 

 band of Apis in two consecutive stages of gas- 

 trulation (after Grassi). 6, lower layer; ec, 

 ectoderm. 



