CRANGON VULGARIS. 121 



toderm, some remain behind in the yolk to form the cen- 

 tre of the yolk spheres ; and though he has not carefully 

 traced the history of these he believes they form the ento- 

 derm. The views of Tichomiroff ['79], though differing 

 much from those of Bobretzky, are still capable of being 

 reconciled with them in their broader features. 



Balfour, usually so prolific in explanations, does not ap- 

 pear to have expressed any very definite reasons for his 

 ideas of the morphology of gastrulation in the higher ar- 

 thropods. In his studies on spiders ['80a] he does not 

 consider the segmentation, but regards the yolk spheres 

 (each of which is nucleated) which fill the egg after the 

 formation of the blastoderm as constituting the entoderm. 

 In his Comparative Embryology ('80, pp. 336, 378 ; '81, 

 p. 278) he extends the same view to the hexapods; claims 

 that the primitive groove is not a gastrula ; regards the 

 yolk cells as endoderm, and while stating that the mode of 

 formation of the endoderm in the 'tracheates' reminds one 

 of delamination, "there are strong grounds for thinking 



O w O 



that the tracheate type of formation of the epiblast and 

 hypoblast is a secondary modification of an invaginate 

 type", and further, that the primitive groove may be a 

 modified blastopore. 



The Brothers Hertwig ['81], recognizing the difficulties 

 which surrounded the interpretation of the gastrulatiou in 

 the hexapods, studied the early development of Noctua, 

 and for the first time gave a clear interpretation of the phe- 

 nomena in accordance with the gastrula theory. Accord- 

 ing to them, the primitive groove is an actual blastopore, 

 and it must be considered that both the nucleated yolk and 

 the mesoderm are potentially invaginated ; but that the abun- 

 dance of yolk has prevented the entoderm (yolk) cells 

 from reaching the surface and taking part in the formation 

 of the blastoderm, and also that the same substance has 



