234 EMBRYOLOGY [CH. 



either side, into the smaller flattened cells of the serosa. When 

 the amnion is completed, the serosa is cut off from the ectoderm, 

 which then becomes continuous, on either side, with the cells 

 of the amnion. It should be borne in mind that, during this 

 process, vitellophags are at work in the yolk dorsally above the 

 ectoderm, so that a continuous supply of nourishment is furnished 

 to the cells of that region. 



The Mesoderm. The main portion of the mesoderm in Odonata 

 seems undoubtedly to be formed by the ingrowth of the primitive 

 groove; though it is very probable that the lateral parts of the 

 mesoderm are formed from isolated cells wandering in from the 

 blastoderm. The cells which form the sides and floor of all the 

 middle portion of the primitive groove become enlarged, so as to 

 differ gradually more and more from the cells continuing the edge 

 of the groove above, which become the ectoderm. Finally, the 

 groove becomes flattened and closed in by overgrowth of the 

 ectoderm above it. The lumen of the groove is lost, and its cells 

 spread out as a slightly convex layer, several cells deep, beneath 

 the ectoderm. This layer is appropriately termed the lower layer, 

 since, in many cases, both mesoderm and endoderm are derived 

 from it. If, however, Heymons' observations be correct 1 , the 

 endoderm has a different origin in the Odonata ; so that we may 

 speak of the lower layer cells in this case as forming the mesoderm. 



The mesoderm soon shews signs of splitting into two layers, an 

 upper and a lower 2 . The upper or ventral layer tends to become 

 closely applied to the ectoderm, and is termed the somatic or 

 parietal layer. From it are developed the segmental muscles. 

 The lower or dorsal layer tends to become closely applied to the 

 yolk-cells lying dorsal to it, and is termed the splanchnic or visceral 

 layer. From it are developed the muscles of the mid-gut. The 



1 Heymons studied chiefly the Odonata, Orthoptera and Aptera. The more 

 recent work of Hirschler, on the embryology of the Coleoptera (Zeit. wiss. Zool., 

 xcn, 1909), challenges many of Heymons' main conclusions, and appears to offer 

 more reasonable explanations of certain difficult points. But researches made 

 upon the Odonata cannot be disproved by work done on the Coleoptera, however 

 excellent it may be, since the embryology of these two Orders differs in many 

 important points. What is required is a thorough re-investigation of the 

 embryology of the Order Odonata itself. 



2 In considering the germinal layers, the ectoderm is taken as the uppermost, 

 although actually the most ventral layer. 



