[33] EMBRYOGRAPltY OF OSSEOUS FISHES 487 



cle<avago being- merely indiciited. When the large cells of the disk com- 

 mence to divide somewhat later and earlier than each other an irregu- 

 larity in the form of the cells ensues which tends to restore the circular 

 torm of the germinal disk, which after the second and up to and in- 

 cluding the third and fourth cleavages had a subquadrate form when 

 seen from above or below. When this stage of development is com- 

 pleted, that is, when the germinal matter of the disk has been split up 

 thronghout its entire thickness into a single layer of cells, the morula or 

 nuilberry stage of development has been completed. 



After the lapse of forty- five hours and a half, or twenty-two and a 

 half hours later, the morula stage has been replaced by another condition 

 of things represented in Fig. 13. Cleavage of the segmentation spheres 

 or cells of the disk also takes jilace now in a jilane parallel to that of 

 the disk itself, so that by this time three layers of cells may be very 

 distinctly made out, which are superimposed upon each other. These 

 layers are the first indication of the development of the germinal layers, 

 and foreshadow the conversion of the germinal disk into a blastoderm 

 in which the epiblast first appears, then the mesoblast and hypoblast. 

 At this stage there is at first, however, no regularity such as might be 

 expected in the disposition of the layers, because the arrangement of 

 the cells is somewhat modified by mutual pressure. 



Later still, or on the fourth day, the advance in the cleavage is very 

 marked, the individual cells being only a fraction of the size which they 

 presented on the second day, and they display, moreover, an arrange- 

 ment into anumber of verj' irregular layers as shown in Fig. 14. Another 

 change has been suffered by the form of the whole disk; it no longer 

 presents the concave inner face shown in Fig. 13, but has become ver}' 

 convex on its inner side and has contracted considerably in transverse 

 diameter and become thicker in the center. 



A little waj^ back we hinted that certain irregularities in cleavage 

 manifested themselves at about the time the germinal disk was divided 

 into twelve to sixteen cells or segments. This phenomenon has been 

 noted in the segmentation of an undetermined fish egg by Prof. W. K. 

 Brooks, of Johns Hopkins University. It is also very strikingly devel- 

 oped during the early stages of cleavage in the eggs of the Clupeoids. 

 It w^ould appear that segmentation is a rhythmical process, and that be- 

 tween the phases of actual segmentation there usually, if not always, 

 intervene periods of rest or quiescence. Within the past year I have 

 had very good opportunities upon several occasions to study the ti^ans- 

 formations of nuclei during the process of segmentation in the early 

 development of the germinal disk and blastoderm of fish ova. These 

 phenomena I have noticed more particularly- in hardened and stained 

 prei)arations, treated with acid carmine, by the use of which the details 

 of the process at various stages may be made palpable. Some of my 

 observations have formed the subject of a short paper published in the 

 Bulletin of the Fish Commission for 1881. That paper, " On the nuclear 



