McCrady.] 174 [AprU 18, 



The third case, c, is probably only a further advanced stage of 

 case a. The radial pseudopodial extensions already exist by hypoth- 

 esis, and the yolk is included in their interspaces, which would cor- 

 respond to the vacant interspaces in Noctiluca. At the surface they 

 are united to a thin peripheral layer of protoplasm enveloping the 

 whole. 



10. In the second case, b, the utilization of the yolk is supposed 

 to be effected gradually by the mechanical inclusion of small portions 

 of it by inward extensions of the peripheral protembryo, so that the 

 yolk is probably taken up, layer by layer, centripetally, as in Musca, 

 Astacus, Limulus. In this case the blastoderm passes from the homo- 

 geneous simple protembryonic state to that of a cellular tissue, with- 

 out presenting any such appearances as would be regarded seg- 

 mentational. 



11. In partial or meroblastic segmentation, the protembryo is, 

 under this theory, supposed to surround the whole yolk, as in case b : 

 either from its first formation; or by gradually extending itself over 

 the yolk (during development) from a definite point, as an Amceba 

 extends itself over a food particle of considerable size. In all partial 

 segmentation, however, the protembryo is specially massed at a defi- 

 nite point, and is not simultaneously and uniformly extended in equal 

 thickness over the whole yolk mass. The process, then, which leads 

 to cellification here is limited at first to the point where the protem- 

 bryo is massed. The gradual taking up of the yolk by inward sar- 

 codic extensions is probably indicated by the formation of the so- 

 called " formative cells " in the egg of the chick, and the supposed 

 wandering of these cells by which they find their way into the middle 

 layer of the blastoderm, is supposed under this theory to be due not 

 to their own spontaneous movement, but to the movements of a hya- 

 line and entirely diaphanous protoplasm, which is assumed to be part 

 of the protembryo, and to fill the segmentation cavity. The free 

 cells which lie beneath the blastoderm in the egg of the brook trout 

 (Oellacher), are probably formed in the same way. 



12. Wholly diaphanous protoplasm, devoid of granules, is fre- 

 quently seen to extend itself in various forms outward from the 

 peripheral contour of Amoeba. There is no a priori reason why 

 such protoplasm should not constitute part of the protembryo in the 

 egg. The fact that it has not been observed, is not conclusive proof 

 of its non-existence ; for its extreme transparency and freedom from 

 granules would render its detection, even when sought for, exceed- 



