50 PROCEEDINGS OF THE AMERICAN ACADEMY 



The 16-cell stage marks an epoch in the history of teleostean devel- 

 opment. The median ph\ne of the embryo is given with the first cleav- 

 age-plane, so that a longitudinal and a transverse axis are determined 

 from the outset ; but a more precise orientation is not attainable before 

 the 16-cell stage. At this time we are generally able (in Ctcuolabrus) 

 to recognize antcro-posterior relations, and with these a rtgJit and a 

 left side. From this stage onward we recognize also a difference of 

 constitution between the ce?itral and the marginal cells.* This dis- 

 tinction is brought out clearly in most of our mounted preparations 

 obtained by the use of osmic acid followed by the chrom-platinura 

 solution. As the contrast becomes more pronounced as cleavage 

 advances, and as it eventually ends in a decided histological differen- 

 tiation, its appearance at this stage must be regarded as an early 

 anticipation of events realized at a much later date. 



The relations now estabhshed between the blastodisc and the 

 periblast, although fundamentally the same, are not so simple and un- 

 mistakable as at first. A precise knowledge of these relations is not 

 readily attainable by a study of the living ovum. Actual sections and 

 mounted preparations are required for this purpose. Vertical sections 

 of this stage show that only the twelve marginal cells rest on the 

 yolk. 



The shaded portion of Fig. 1 shows how much of the floor of these 

 cells is in contact with the yolk. The four central cells and an ad- 

 joining zone of the marginal cells form the roof of a shallow cleavage- 

 cavity, of which we shall have more to say farther on. The floor 

 of this cavity is formed by a very thin stratum of the periblast. This 

 siib-germinal stratum becomes apparent in some cases as early as the 

 8-cell staire. In some of our sections it is in contact with the central 

 cells, but seldom shows traces of fusion with them. In Fig. 2 all 

 these relations are accurately shown. The periblast (/)) joins the 

 marginal cells at their inferior outer angle, and from this point to the 

 point (c?) which marks the limit of the cleavage-cavity there is not 

 the slightest trace of a periblastic layer. This region may be desig- 

 nated as the zone of junction (2). No nuclei are to be found in any 

 ])ortion of the peril)last at this time. As to the origin of the sub- 

 germinal periblast seen in this stage, we can state positively that 

 it does not arise by a horizontal cleavage. It is a portion of the 

 periblastic material which works its way under the blastodisc con- 



• Rauber noticed tliis fact in the ovum of Gobius, but gave no explanation 

 of it. Morph. Jahrb., VIII., p. 287. 



