No. 3.] THE CELL-UNEAGE OF NEREIS. 407 



I. The Fust Somatoblast. 



Nothing in the development of Nereis has excited my interest 

 in a higher degree than the history of the first somatoblast, 

 which undergoes a long series of divisions, continually changing 

 in their character, comparable in regularity and precision with 

 the cleavage of the ovum itself. I have followed these changes 

 many times, both in the living embryo and in preserved speci- 

 mens, and am convinced that the order and form of the divisions 

 are almost absolutely constant. Their character is such, more- 

 over, as to bring to the foreground some of the most interest- 

 ing problems in the mechanics of cell-division. 



The first three divisions have already been described, but may 

 briefly be reviewed. The somatoblast yfr^-/ buds off a small cell 

 {x^) at the right side (Fig. 20), second, a similar cell {x^) at the 

 left side (Fig. 29), and third, a third cell {x^) in the median pos- 

 tero-dorsal line (Fig. 31). These three cells are approximately 

 120° apart, and are nearly, but not quite, symmetrical with respect 

 to the median line (cf. Figs. 32 and 48), The foiirth division 

 cuts the somatoblast into equal parts (Figs. 37, 38, PI. XVI ; 

 50, 51, PL XVII), which are nearly symmetrical with respect to 

 the median plane, but always show a slight displacement towards 

 the left. These two cells I shall call the posterior proteloblasts 

 {XX). At the fifth division (Figs. 52-54) each of the pro- 

 teloblasts buds forth a smaller cell (;tr*, x'^) at its anterior extrem- 

 ity. At the sixth division (Figs. 42, 43, PI. XVI) each buds forth 

 a small cell {x^, x^) at its outer margin. At the seventh division 

 (Fig. 54) two small cells (;f^, .r^) are budded forth at the latero- 

 posterior angles.^ At the eighth division (Figs. 46, 56-58) each 

 proteloblast divides longitudinally into two equal parts {XXXX). 

 The four cells thus formed I shall call the posterior teloblasts. 

 Meanwhile x^, x^, x^, and x'^, x^ divide transversely, the result 

 being shown in Fig. 46, PI. XVI. The order of division varies 

 somewhat. As a rule, x^ (Figs. 50, 52) divides first (not x^, as 



1 In regard to this division, there is unfortunately some uncertainty, owing to the 

 fact, as I believe, that I overlooked the sixth division in the living embryos, and only 

 subsequently observed it in the prepared specimens. That such a division takes place 

 at this period in many specimens, there is no doubt. It is possible, however, that 

 this is only a variation of the fifth division, though, in view of the precision that in 

 general characterizes the behavior of the proteloblasts, I must regard this possibility 

 as rather remote. 



