EMBRYOLOGY OF THALASSEMA MELLITA 193 



veloped prsetrochal region in every case, and hence the blasto- 

 meres at the 8-cell stage are subequal in size. The above 

 agrees essentially with the explanations that have been given 

 by Lillie ('95), Conklin ('97), Treadwell ('01) and others and 

 without doubt the principles of precocious segregation of forma- 

 tive matter, is an important factor in determining inequality or 

 equality in cleavage. 



Nevertheless, the cleavage of TJialasscma seems to show that 

 another and possibly more fundamental factor must be taken 

 into consideration, and that is the tenacity with which the primi- 

 tive radial cleavage and symmetry of the ancestral prototype of 

 the annelids has persisted. In TJialasseina and in Podarke the 

 bilaterality of the adult is not expressed early in the ontogeny 

 by bilateral cleavages, as in many annelids with unequal cleavage, 

 but is first indicated by modifications in the size relations of 

 the radial cleavages, either directly as a result of a cleavage or 

 by a subsequent increase in size of certain cells or regions. A 

 good example of the first is the fact that in the second division of 

 the third quartet the larger products of both the c and d quad- 

 rants lie on each side of the mid-dorsal line (PI. II, Fig. 13). 

 Of the second, that the anterior cesophagoblast (2<^.2. 2. ].2)> shortly 

 after gastrulation is completed, is the largest cell in the embryo 

 (PI. II, Figs. 23 and 24; Text-Fig. 4, C) ; yet, up to the begin- 

 ning of gastrulation this cell remained of equal size with the 

 corresponding cells in the a and c quadrants. The last point, 

 i. e., the foreshadowing of the bilaterality of the adult by a rapid 

 increase in size of certain regions, is clearly illustrated in the 

 development of the X group (PL II, Figs. 15, 17 to 24; Text- 

 Fig. 4, C and U). The protoblast of this group, when first 

 formed, is not larger than the other members of the second 

 quartet, nor does it divide much more rapidly than they, until 

 the beginning of gastrulation. At this point in the ontogeny, 

 the whole group increases rapidly in size, not only by cell di- 

 visions, but by the growth of the cells themselves, and in the 

 end forms almost as much of the post-trochal ectoderm (Text- 

 Fig. 5, /^) as does the great " somatoblast " of Nereis. The 

 important point to bear in mind is, I think, that even in deter- 



