IO 



N. YATSU : OBSERVATIONS AND EXPERIMENTS 



operation, but rather to that of thé initial organization or some other 

 physiological conditions. It may be interesting to add that the 

 stomodaeum fails to develop in all the cases of the above isolation 

 experiment, namely E, 2E and 2M. 1 The sense organ may be formed 

 (Figs. 8, 9. II, 15). 



4. Variation in the formation of tlte 

 comb-rows. 



It cannot de doubted that each blastomere of the 8-cell stage 

 contains the basis for one comb-rows in it as has been ascertained by 

 FlSCHEL ('97 p. 122). I tried several series of experiments and the 

 results agree in almost all cases with his, but exceptionally I met 

 with the cases in which the number of comb-rows was greater or less 

 than that of the original blastomeres. 



The embryos developed from two middle-cells have just been 

 described in the last section. 1E + 2M gave rise to larvae with two 

 comb-rows (six cases) (Figs. 19, 21). Fig. 20 is an embryo with 

 three comb-rows from such combination of blastomeres. 2E+1M 

 yielded larvae with two comb-rows (one case) (Figs. 17, 18) and four 

 rows (four cases). 3E + 2M should give rise five-rowed larvae, but I 

 have three cases with six rows (Figs. 24, 25). From 2E + 3M resulted 

 four-rowed embryo (one case) (Fig. 22, 23). 



In the above exceptional cases one at once notices that the number 

 of the resulting comb-rows is usually smaller than that of the blasto- 

 meres. This may well be explained as the arrested development, but 

 how the resulting embryo has more comb-rows than it might be 

 expected from the number of the original blastomeres is very difficult 

 to interpect. The latter cases may be due to unequal distribution of 

 the ctenoplasm among the blastomeres. 



i E = end-cell, M = middle-cell. 



