420 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



appear at an earlier phase. The cleavage of the yolk is wholly, or almost wholly, confined to the 

 surface furrows, dividing planes rarely extending into the yolk below. My material was not ade- 

 quate for determining whether invagiuatiou was preceded by delamiuation or not, but seems to 

 render it highly probable that a delauiinatiou does not occur. 



My observations on Pontonia domestica, Palcemonetes vulgar fs, and Hippa talpoidcs are very 

 fragmentary. In Poutouia I have one stage (Fig. 27) with three cells, one of which is dividing 

 with no sign of yolk division, and another with sixteen nuclei and corresponding yolk 

 pyramids. Here the conditions are precisely like those in Alpheus saulcyi at a similar stage, 

 and probably the segmentation of the two is sitniliar. In Palcemonetes Faxon ( 17 ), relying 

 wholly upon surface views, states that segmentation of the egg begins in two planes almost 

 simultaneously. These planes are at right angles to each other and pass through the long and 

 short axes of the egg. Whether this follows close upon the first division of the nucleus I have 

 not determined. I have an egg with two nuclei near the center of the undivided yolk, and a stage 

 with thirty-two yolk pyramids. This egg also agrees closely with that of Alplieus saulcyi at the 

 same stage. The nuclei are not quite at the extreme surface of the yolk. In the next phase when 

 sixty-four pyramids are present the protoplasm abuts on the surface (Fig. 24) of the egg. All the 

 protoplasm is distributed to the yolk pyramids and no delamiuation has taken place. 



With reference to Hippa I can only add the note that yolk pyramids normally occur, and in 

 the 64-cell stage (Figs. 1, 4) the lines of cleavage between adjacent segments are very distinct and 

 extend nearly to the center of the egg. At this stage all the protoplasm is apparently concen- 

 trated in these cells. Yolk pyramids similar iu surface views to those of Hippa occur m CalUnectes 

 hastatus, Platyonychus ocellutus, and Libinia canaliculata. 



Crangon vulgaris. Kingsley (31) describes the segmentation of Crangon as follows: 



With tlio first segmentation the protoplasm begins to leave its central position and seek the surface of the egg; 

 before the second division is completed it has reached the surface, leaving the. yolk in the center. * After the 

 second protoplasmic segmentation is effected the first segmentation furrows appear, the one following close upon the 

 other. The iirst to appear corresponds iu its direction to the lirst unclear division; the second is at right angles to 

 it. * * * In Crangon, so far as I have been able to see, amti-.boid cells reach the surface and take part in the 

 formation of the blastoderm before the process of gastrulatiou begins. In that form no yolk pyramids occur. 



Of cell division he says : 



In the process of cell division I have never seen any traces of karyokinesis; the division seems to be direct, and 

 affects first the nucleus and uext the protoplasm. * * In fact I do not recall a single statement of karyokinesis 

 being witnessed iu decapod segmentation excepting in Astacns. 



On PI. I he gives a drawing (Fig. 3) of an egg with "about sixteen segmentation spheres." 

 Fig. 4 of the same plate represents a section of the egg shown iu Fig. 3 and has six nuclei, five of 

 which are even with the surface, while one is near the center of the egg. The yolk spherules appear 

 to be fused, owing possibly to the disturbing effect of the reagents employed. This central cell, 

 according to Kiugsley, represents a portion of the egg protoplasm which is belated in its passage 

 to the surface, but it divides and gives rise to cells which eventually reach the surface at a certain 

 point which marks the germinal area. Thus all the nuclei take part iu the formation of the blasto- 

 derm, and the migration of the belated cells is completed before the "gastrula" iuvagination 

 occurs. I have made no observations on the very earliest phases of segmentation of Crangon, but I 

 have several eggs sectioned at the sixteen-cell stage, which ought therefore to correspond with 

 Fig. 4 of Kiugsley's paper, but, on the contrary, they show a somewhat different condition of 

 things. There are just sixteen nuclei present, all of which are peripheral or nearly so, and each 

 nucleus forms the center of a yolk pyramid, the cleavage planes of which are very marked and 

 extend more than half way to the center of the egg. Fig. 15, PI. xxvn, of the segmented egg of 

 Alpheus, although belonging to a later phase, will fairly represent the condition of things which 

 we find iu Crangou. None of the nuclei are tangent to the surface, but between them and the sin - 

 face there is still a considerable layer of yolk. Each is surrounded by a large mass of protoplasm, 

 which stains lightly with ha3tnotoxylon, and has the characteristic rayed appearance. One of tiiese 

 nuclei is in the equatorial plate or metakinetic stage of division, and may be represented very 

 nearly by Fig. 28, which shows a dividing cell in the egg of Alpheus at the same period. As already 

 stated, the ceutral yolk mass does not contain a single nucleus. The yolk is in the usual form of 



