48 PROCEEDINGS OF THE AMERICAN ACADEMY 



mann, becomes split off by a horizontal plane of division before the 

 first meridian cleavage appears. Equally fruitless has been our search 

 for this llortinanniau stratum in mouuted preparations and sections. 

 Not only is there no such splitting, but the whole course of events 

 thus far is incompatible with such an operation. 



A series of optical sections obtained by the aid of the camera, at 

 intervals of one to four minutes during the first cleavage, is exceed- 

 ingly instructive. At the time the meniscoidal form is reached, there 

 is still no boundary line between the periblast and the blastodisc. 

 The blastodiscal portion of the protoplasmic envelope thins out 

 gradually, and passes with even outlines into the periblastic portion, 

 which, after passing the equator, becomes very thin. At every step 

 it is very evident that the blastodisc is thickening and broadening at 

 the expense of the periblast. At the time tlie first cleavage begins, 

 the blastodisc has an oval form, with only a broad, shadowy outline, 

 as seen fi'om the inner surface. As the cleavage-groove descends, the 

 margin of the disc thickens and the adjoining periblast becomes thin- 

 ner. At the time the two new nuclei appear, the outlines of the 

 blastodisc begin to clear up ; but they do not attain their sharpest 

 definition until a few moments after the second amphiastral stage is 

 reached. Seen from the surface, the blastodisc now appears to be 

 separated from the periblast by a clear-cut boundary line ; but a pro- 

 file view shows that the continuity is still unbroken. The illusive 

 appearance is due to the abruptness with which the thickened margin 

 of the blastodisc rises above the jieriblast. 



Soon after the second amphiastral phase appears, we notice that 

 the blastodisc begins to expand iu the direction of its shorter horizon- 

 tal axis ; i. e. in a direction at right angles to the coming (second) 

 cleavage-plane. At the same time the margin of the blastodisc begins 

 to lose its sharp outline in surface views, especially at those points iu 

 or near which the second cleavage-groove is to terminate. Optical as 

 well as actual sections show that the blastodisc is thinning out at these 

 points, so that the boundary line between it and the iieriblast is again 

 only a little less obscure than at first. With the appearance of the 

 third generation of nuclei, the outlines begin to clear up again, be- 

 coming sharper and sharper, until a maximum limit is reached soon 

 after the introduction of the third amphiastral i)hase. 



Thus towards the close of each of these earlier amphiastral divis- 

 ions there is an obscuration of boundary lines, especially at the mar- 

 ginal extremities of the incipient cleavage-grooves, followed in each 

 case by a gradual clearing up, which culminates in the early part of 



