MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 419 







larvre. On the other hand, the Bahaman variety of Al/iltens licteroHiclis and Alj>hc-iui HI i nor (from 

 Beaufort, North Carolina) agree in having a relatively smaller yolk and in hatching as /,<ua-like 

 forms. They differ, however, in their segmentation, the first species agreeing in this respect with 

 A. saulc'i/i, while in A. minor yolk pyramids are generally absent and the segmentation is irregular. 

 The yolk in A.hcterocliclia of Beaufort is about nine, times ;ts voluminous as that of the Bahainaii 

 heterochelitt. The segmentation, however, has remained unaltered. The peculiarities which we 

 find in the early stages of A. minor can not therefore be laid to the door of the yolk alone, but must 

 be regarded as a comparatively recent modification of the yolk pyramid type. While the type of 

 segmentation may be very persistent and uniform, it is subject to profound change, not only in 

 closely allied species, but, as has been shown in a few instances, within the species itself. 



In the Decapod egg we have, as a result of segmentation, a great central yolk-mass which is 

 either undivided or imperfectly divided, and which completely tills and obliterates the segmenta- 

 tion cavity, and a surrounding layer of cells, the blastoderm. More fully stated the process is as 

 follows: The segmentation nucleus divides first at a point near the center of the egg. The daugh- 

 ter cells separate widely, and a second division follows. With subsequent divisions the cells 

 approach nearer the surface. The yolk may share in these early divisions but often does not, until 

 eight or sixteen cells are formed (third or fourth segmentation stage). When there is no pro- 

 gressive segmentation of the yolk in the early phases, the yolk segments appear either gradually 

 and oil one side of the egg, or make their appearance simultaneously in relation to all the nuclei 

 present. Segmentation is always rythmical. During one phase (period of "rest") the segments 

 shrink away from the surface and flatten out in the usual way, while at the beginning of the period 

 which follows (period of " activity") they swell and stand out prominently. The division of the 

 yolk often appears total in surface views, while in reality it is not. The constrictions marking 

 off the segments may be nearly superficial, or they may extend deep into the yolk. Cell division 

 is usually indirect. The only exception to this rule which I have observed is Alpheiw minor (see 

 p. 397). With each division the protoplasm approaches nearer the surface of the egg, and the 

 segments become more pyramidal in shape. These are the " yolk pyramids" which were first 

 described in the crayfish by Rathke in 1829. The bases of the pyramids form the surface of the 

 egg and their apices fuse ill the central yolk mass. The nucleus, surrounded by a rayed body of 

 protoplasm, lies near the base. By repeated division the pyramids become smaller; the cleavage 

 planes of adjoining segments become less distinct and the protoplasm draws nearer and nearer the 

 surface. Before, however, any of the protoplasm is flush with the surface, certain of the cells 

 divide horizontally, that is delamiuate, and one of the products of each division migrates into the 

 yolk below. The surface of the egg is now covered with a single layer of small polygonal cells, 

 and the pyramidal structure of the peripheral yolk has nearly disappeared. The greater part 

 of the protoplasm of the egg is thus at the extreme outer ends of the reduced yolk pyramids, while 

 the lesser body of it is represented by the migrating yolk cells. 



The invagiuatiou stage soon follows, and the large numbers of cells which now become migra- 

 tory (in Alphens penetrating to all parts of the egg), joining the primary yolk cells, represent 

 mesoblast and eutoblast. The process above described will be found to apply, I believe, to the 

 majority, if not to all Decapods. Differences in detail may be expected in the time of appearance 

 of yolk segmentation and the degree to which this is carried. This account differs from that usually 

 given in recognizing delarni nation, following close upon the latter phases of segmentation, or the 

 origin of cells, from the blastosphere before the iuvagination appears. This regularly occurring in 

 such typical forms as Alpheus and Homarus argues strongly for its presence in allied species 

 where it has possibly been overlooked. 



Stenopws hispidus.I have described and figured the segmentation of Stenopus in a paper on the 

 life history of this form. In one egg which I sectioned before the beginning of yolk segmentation, a 

 single cell, probably the male pronucleus, is seen at the surface, while another cell, the female pro- 

 nucleus, lies near the center. A single polar body was also present, just beneath the shell, near 

 the first nucleus. I saw no evidence of yolk segmentation until the third phase of division was 

 reached. The yolk then became constricted at the surface into eight blastomeres. The superficial 

 furrows are quire deep during active periods, giving to the egg the appearance of total cleavage. 

 This egg now resembles that of Pemeus as figured by Haeckel,butiu the latter form yolk segments 



