192 JOHN E. S. MOORE. 



marked concavity (fig. 22) on the side opposite the archo- 

 plasm, now elongated in the form of a spindle. The nucleus 

 also has become elongated, so that it presents an initial form 

 corresponding to the final horseshoe outline of the daughter- 

 nuclei. The plane of this initial horseshoe is conspicuously 

 at right angles to that of the archoplasmic axis, which at this 

 period is also bent, so that the two structures interlock after 

 the manner of one horseshoe hanging in the concavity of 

 another. The chromatin is now looped across the longer axis 

 of the nucleus, as in the final condition (figs. 21, 13) ; the 

 nuclear membrane, becoming wavy and indented, as described 

 in the spermatocytes by Hermann, finally disappears, the 

 chromatin elements remaining free in the cell-mass. 



Prior to this disruption of the nuclear membrane, the radia- 

 tions from the central bodies become connected with the 

 individual chromatic elements. But the question whether 

 they are extra-nuclear structures and push inwards from with- 

 out (as suggested by Strassburger and Watase^), or are con- 

 tinued at the expense of the intra-nuclear substance from 

 within, must remain for the cells here described an open one. 



These astral radiations are completely outside the archo- 

 plasmic spindle (if I may use the term), and, as Hermann 

 says, they invest it like a mantle. 



The free ends of the loops of chromatin stand stiflfly out at 

 right angles to the inner spindle-fibres (the central spindle- 

 fibres of Hermann). They usually, however, do not complete 

 the circle, the space left being probably the representative of 

 the side of the cell destitute of nuclear elements in Hermann^s 

 figures of the spermatocytes. 



All these chromatin bodies are absolutely outside the spindle, 

 and I can obtain no evidence from the cells of the above 

 embryonic ridge that any of the spindle-fibres are formed from 

 the intra-nuclear protoplasm at any time. 



In some cases the spindle is apparently not continuous from 

 end to end, the fibres presenting an appearance somewhat 



* 'Studies on Cepbalopods : Cleavage of the Ovum,' S. Watase, 'Journ. 

 Morpb.' 



