378 



ANNUAL REPORT SMITHSONIAN INSTITUTION, 193 3 



in clumps at each end of it, that will later multiply and again enclose 

 the yolk in an internal sac, which becomes the stomach of the mature 

 insect. These dormant cells associated with the mesoderm, therefore, 

 must be the endodermal cells of the embryo, and, if so, the vitello- 

 phags dispersed from them are also endodermal. 



Here, then, clearly is a striking example of apparent utter disregard 

 on the part of the embryo for the historic method of mesoderm and 

 endoderm formation. Yet, we must admit that the insect embryo is 

 true to its ancestral history insofar as it does develop each of these 

 layers from what ought to be the blastopore area of its body, and in 

 that it turns over the business of digesting the yolk to at least some 

 of the endoderm cells, namely, the vitellophags. It is quite evident 



FiGUBB 10. — Diagrammatic cross-sections of three early stages in the developing insect 

 embryo showing one method of mesoderm formation, with accompanying endoderm 

 cells. 



A, the germ band divided into a middle plate (MP) and lateral plates (LP) ; gastru- 

 lation represented by internal proliferation of endoderm cells (End) from the middle 

 plate. B, the middle plate becoming invaginated, most of the endoderm cells being 

 dispersed in the yollc as vitellophags (Vph). C, the middle plate transformed into 

 mesoderm (Msd) by the closure of the ectodermal lateral plates (Ecd) ; a remnant 

 of the endoderm (MR) left on the inner surface of the mesoderm. 



that if the embryo developed its endoderm by the traditional method 

 of invagination, its stomach would become a sac surrounded by the 

 food material. A stomach, however, is not designed to work from the 

 outside, for in such a case it might digest the animal instead of the 

 food. Stomach cells dispatched into the yolk, however, may digest 

 the latter and render it available to the growing tissues of the embryo. 

 Hence, while the insect embryo adheres to the fundamental principle 

 of gastrulation, it discards of necessity the idea that stomach forma- 

 tion must proceed in one particular way. 



Let us see now just how far the insect embryo has departed from 

 the ancestral path in the matter of gastrulation. We first go back to 

 the typical open gastrula (fig. 11 A) with the mesoderm forming 

 along the sides of the elongate blastopore. If then we assume that 

 the blastopore has closed, we arrive through B at the condition in C, 

 which clearly suggests the developmental process that takes place 



