EMBRYOLOGY OF THALASSEMA MELLITA 199 



3. History of the Dorsal Gap in the Prototroch and of the Shift- 

 ing of Embryonic Areas. 



Dorsal Gap. — A number of the early observers of the early 

 development of annelids and molluscs have noted the dorsal in- 

 terruption or gap in the prototroch. Among the first of these 

 was Hatschek i^j'^^. Until recent years, however, the cause of 

 this phenomenon has not been known. There are two fac- 

 tors, I think, which we must now taken into consideration. In 

 Amphit7'ite and Clymenclla, as Mead ('97) has shown, the dorsal 

 gap is due to the fact that the cells in the d quadrant of the 

 second quartet, which corresponds to the secondary protochal 

 cells of the other three quadrants, fail to develop cilia. This 

 explanation seems to suffice, also, for a good many other forms, 

 but in Thalassema^ as will be shown below, the dorsal inter- 

 ruption in the prototroch is caused entirely by the migration of 

 cells from the prae-trochal into the post-trochal region. 



In all other types of annelids, whose cell-lineage has been 

 determined, the dorsal gap is present from the beginning. In 

 Thalassema, on the other hand, it does not arise until about the 

 eighth hour (PI. II, Fig. 22). It is caused, in this case, by the 

 divisions and spreading out of the posterior arms of the cross 

 (PI. I, Fig. 7), which, thereby, push the posterior intergirdle 

 cells down along the mid-dorsal line. At the appropriate time 

 (during gastrulation) the prototrochal cells at the juncture 

 of the c and d quadrants are shoved apart (PI. II, Fig. 24). 

 Before they come together again, two large cells and at least 

 eight smaller ones, with dark staining nuclei, are pushed through 

 the gap and finally lie between the X group and the prototroch 

 (Text-Fig. 4, C and D^. The large cells, i<3^i.2. 1. 2. 1 ^^^ 

 1^1.2.1.2.2 (PI- II> Figs. 14 and 24) are designated by the letter 

 '*/," although as has already been indicated, they have not quite 

 the same cell lineage as the "/" cells in Amphitrite, Arenicola 

 and Podarke. The smaller cells probably represent all of the 

 intergirdle cell id^ 2. 1. 1 ^^^ it is also possible that some products 

 of the posterior arms of the cross pass through the break. In 

 Arenicola four descendants of the posterior intergirdle cell 

 (1^1 2) migrate into the post-trochal region, but in this form and 



