?.'y Traterirrf for the foi^rrtior of the endodprr in 



A 



Storotoca vas scarce and it in rot inrcssi ble to have -ris- 

 intrrrrpto'^ tho ':'ho?^0'^ona, Hov.'ovcr,! a"^ ir^clined to thinl: 

 that the c^doder-a" cells arise by a trarsverse divisior of 

 the ectodernal cells, as Vetschril-'off sho'vs ir the ex'^ert- 

 ioral case of Clytia viridlcr^s. Fipure IB is drav.T frcr 

 the c-l}' secticr I ^-as able to seciire frc rreserved "cterial 

 showing the bp,?ir,rirp; of the for-rtior of the e-doderr, and 

 that^ras cut rlirhtly ^blioue, causirr ?.C'-e doubt. A 

 sectior of a little older stap-e a^d drawn rith hipher r^agni- 

 ficatio" is shOT/n i" Fipure 19. Here there are three cells 

 that an-e-r to have lust divided by trarsversr division, 

 /.nother reason which causes ire to thinl: that the e^^doder'-al 

 cells arise by transverse divisic^ of the criminal ecto- 

 derF cells is "he fact that thr ectcrerrrl cells in this 

 repicn are practical ly as ride as those i^^ other rarts of 

 The blastula. This v^uld liot be the case if the lonri- 

 tudiral division cccured: for necessrrily cell division 



