128 THE DEVELOPMENT OF 



to trace their results beyond the point of the recognition 

 of the three layers and hence we have omitted those por- 

 tions which treat of the modification of the primitive en- 

 doderm into the epithelium of the mesenteron. Since 

 Haeckel, most authors have realized that the primitive 

 groove is in some way connected with gastrulation and 

 many are of the idea that the yolk cells and the " Wander- 

 zellen " have a part to play in the formation of the endo- 

 derm. Not so Dohrn. He says ( J 76) that they have no 

 connection with the primitive groove though they may 

 come to lie beneath it. The " Wanderzellen " occur in the 

 adult as well as in the embryo. They form the fat bodies 

 and the blood and they pass out through the dorsal organ 

 into the space between the embryonic envelopes. He also 

 mentions that the neurilemma is derived from similar ap- 

 pearing cells. It is highly probable that Dohrn has taken 

 similar appearance for actual identity and has confused 

 amoeboid mesoderm cells with similar cells derived from 

 yolk cells which are really endodermal. 



Of the early stages of the myriapods we know compar- 

 atively little. Stecker ('77) describes a regular gastrula- 

 tion in four genera of Diplopods (lulus, Craspedosoma, 

 Polydesmus and Strongylosoma) but a glance at his plates 

 convinces one that his statements deserve the criticism to 

 which Balfour subjects them. More recently, Heathcote 

 has investigated the development of lulus and his account, 

 ('86) while confirming that of Metschnikoff, adds other 

 details. The nuclei of segmentation, each surrounded with 

 protoplasm, migrate to the surface to form the blastoderm, 

 the later nuclei uniting with others derived from the blas- 

 toderm to form a keel like that described by Balfour in 

 Ageleua. This keel furnishes the mesoderm, and Heath- 

 cote regards it as homologous with the primitive streak of 

 other arthropods. Other nuclei remain in the yolk and 



