MUSEUM OF COMPARATIVE ZOOLOGY. 209 



obliquity is more likel}'- to have been produced by an inequality in the 

 constriction, just as in ordinary cleavage, where it is often found that 

 the segmentation furrow advances from one side with much greater rapid- 

 ity than from the opposite. In a view at right angles to this (Fig. 67) 

 the spindle and globule are seen to be quite symmetrical, though not 

 strictly radial to the vitellus. I have so often observed this obliquity, 

 that, although certainly not to be considered as universal, I believe it to 

 be characteristic of this stage. 



Another not less peculiar, though by no means constant phenome- 

 non, affects the inner star. During the formation of the second polar 

 globule, the radiate appearance in the vitellus becomes wider and wider, 

 until it at times is traceable to within a short distance of the periphery. 

 It is at the close of the formation of the second globule that it seems 

 to attain its maximum extent, — to dominate the whole vitellus. Lying 

 as it does, with its centre so near the surface of the yolk, the rays are 

 necessarily of very unequal length. That, however, is neither their most 

 noticeable nor most interesting peculiarity. Seen in certain favorable 

 positions, they are observed to stretch away toward the periphery, not 

 in;.'igid straight lines, but in bold, sweeping curves, which are so related 

 to each other that they present the appearance of extensive, more or 

 less sharply curved spirals. A view upon the animal pole affords a sur- 

 vey of the most extensive curves. Occasionally (Fig. 78) the course of 

 rays may thus be traced in a sweep of nearly 400"^. Tracing them from 

 the centre of the star, these fibres may be seen to curve in a constant 

 manner, so as to be only slightly divergent. Finally, after completing 

 an immense arc, they become invisible near the surface of the yolk. 

 Such prominent fibres, however, do not describe their spirals in a single 

 plane, but, as the focusing of the instrument as well as side views teach, 

 they gradually descend toward the vegetative pole as they recede from 

 the centre of the star. Necessarily not all the rays of a given vitellus 

 are thus extensive, but all show the curvilinear course more or less dis- 

 tinctly. In some cases the rays do not seem to centre in a common 

 point, but to arise along an axis, as in Fig. 66. The latter, however, is 

 not a straight line, nor even a simple curve, but shares in the spiral in- 

 fluence expressed in the rays in such a way as to have approximately 

 the form of a corkscrew. The dotted line a /3, Fig. 66, gives the projec- 

 tion of this corkscrew axis on the plane of the optical section. 



Both the extent of the rays and the degree of their curvature are. 

 subject to great variation in diff'erent eggs. The opposite extremes to 

 those just described are presented by Fig. 63, and all gradations be- 



VOL. VI. — NO. 12. 14 



