202 PROCEEDINGS OF THE AMERICAN ACADEMY. 



Seeliger's next stage (his Fig. 16) is oriented correctly, the poste- 

 rior end being clearly indicated by the smaller marginal cells (left 

 unshaded in his figure). This criterion for the posterior end of the 

 embryo had been previously pointed out by Van Beneden et Julin, 

 has since been recognized by Chabry and by Davidoflf for the forms 

 they studied, and holds good for Cioua also. 



Seeliger speaks of his Figure 16 as representing a stage with four- 

 teen " eudoderm " cells (that is, cells which have been derived from 

 the four dorsal cells of the 8-cell stage). It is highly probable that 

 he should, have represented as *' endodermal " also a pair of lateral 

 marginal cells left unshaded in his figure, thus bringing the number of 

 descendants of the four dorsal cells of the 8-cell stage up to sixteen, 

 the nixmber we should expect as the result of two successive divisions 

 in that hemisphere. The two cells in question correspond in position 

 with the cells c®"^ and d^-^ of Figure 4 of this paper, except that in the 

 egg of Ciona figured the cells named do not reach the margin, but lie 

 a little more central. 



Interpreted thus, Seeligex-'s Figure 16 corresponds closely with a 

 similar view of the 32-cell stage of Van Beneden et Julin (reproduced 

 in Figure 8 of this paper), and with Chabry's Figure 26, and with the 

 same stage of Ciona (Figure 4 of this paper). 



From this stage on, all investigators are agreed as to which is the 

 anterior and which the posterior end of the embryo. If then the ori- 

 entation of Seeliger's later stages is correct (and this will be conceded 

 by every one), that of his 16-cell stage is ?'??correct ; and unless he again 

 reversed the poles m passing from the 4-cell to the 1 6-cell stage, his 

 A:-cell stage is also wrongly oriented. This l)eing so, the two larger 

 cells of the 4-cell stage are really anterior in the Clavelina studied by 

 Seeliger, just as in the one studied by Van Beneden et Julin. 



Another point of difference between the authors mentioned, appears 

 in their discussion of the 8-cell stage. Seeliger states that the larger 

 cells belong to the ventral (ectodermal) half of the egg, whereas Van 

 Beneden et Julin state that they constitute the dorsal (endodermal) 

 half. 



I shall i)resently show that Van Beneden et Julin were mistaken 

 in their determination of the dorsal and ventral faces of the 32-cell 

 and earlier stages. It follows that the larger cells, which they called 

 dorsal, were really ventral, as stated by Seeliger for the form studied by 

 him. If this is so, another point of difference is removed, for in both 

 the species of Clavelina that have been studied the larger cells of the 

 8-cell stage are ventral with reference to the axes of the future larva. 



