208 TORREY 



this cell toward the apical plate the open anterior end of the 

 blastopore approaches the prototroch and, finally, the position 

 of the adult mouth. The three o^sophagoblasts become grouped 

 about this opening and sink in nearly simultaneously. The one 

 on the right (^^■>,i,i) divides once and nearly equally just before 

 sinking in, but the others do not divide again. The first, which 

 possibly represents the larval mesoblast from the d quadrant, 

 takes the place of the stomatoblast that as a rule arises from the 

 second quartet on that side (2^0 2+). The cesophagoblast on the 

 left (2<rZ2. 2. 1. 2. 1+) i^^^ti^g with less obstruction, sinks in faster 

 than the anterior and that in turn more rapidly than the one on 

 the right. Accordingly, they finally become arranged in a row 

 extending dorso-ventrally (Text-Fig. 6, B and C). When the 

 oesophagus has become differentiated, the one at the inner end 

 of this row (2<^2. 2. 1. 2. 1+) surrounds in large part the opening into 

 the stomach, the median (2<^2. 2. 1. 2) fornis the antero-dorsal 

 wall, and the one at the outer end (3<^2. 2. 1) ^ P^^t of the wall on 

 the right side (Text-Fig; 6, C and D). The posterior wall is 

 formed entirely by the infolding of the columnar cells which lie 

 along the mid-ventral line. There is no doubt that a consider- 

 able number of cells of the ventral plate take part in this 

 process. 



Conn failed to note this invagination of ectoderm cells and 

 so has described the oesophagus as endodermal and as resulting 

 from a division of the digestive tract. Hatschek ('85), however, 

 in Eiipoinatus, has recognized the fact that the entire oesophagus 

 arises from the ectoderm. In this form the oesophagus and 

 archenteron are described as being continuous from the first and 

 as not coming secondarily into connection with one another as in 

 Tlialassema. The intestine, too, arises in this case as a later 

 proliferation of cells from the posterior dorsal wall of the en- 

 teron. The same is the case in Poinatoccros (Von Drache, '84). 



In annelids the stomatoblasts seem to arise, as a rule, from 

 the second quartet, anterior and lateral quadrants (^Nereis, Capi- 

 tclia, Podarke), but in Arenicola, according to Child ('01), they 

 take their origin from the same quadrants of the third quartet. 

 In these forms, also, practically all of the rest of the progeny of 



