ASEXUAL REPRODUCTION IN SAGARTIA 179 



in the total number of incomplete endocoels those mesenteries 

 undoubtedly incomplete but of undetermined grade. The per- 

 centages have been figured on the revised totals. The table 

 shows division planes as follows: complete endocoels, 62 per 

 cent; incomplete endocoels, 34 per cent; exocoels, 4 per cent. 



The significance of these figures, in showing tendencies of the 

 division plane to pass in spaces of a certain character, is af- 

 fected by the relative numbers of spaces of each character in the 

 animals at the time of division. Possibly the relative width of 

 the spaces should also be taken into account. At an early 

 stage in the formation of a cycle of incomplete mesenteries the 

 adjacent exocoels are so reduced that the sum of the widths of 

 endocoels considerably exceeds the sum of the exocoels. This 

 condition gives way to a state of practical equality between 

 endocoels and exocoels so soon that the temporary inequality 

 probably has little if any influence on the relative frequency of 

 passage ofHhe plane of fission in different spaces. Certainly, the 

 inequality between endocoels and exocoels is not sufficient to 

 account for more than a small fraction of the excess of endocoelic 

 divisions recorded. 



The number of exocoels in an anemone is precisely equal to 

 the sum of all endocoels. According to chance, neglecting in- 

 equalities in size, 50 per cent of the divisions should occur in 

 exocoels. As shown in table 10, division planes occur in exocoels 

 in only 4 per cent of cases. It is clear, therefore, that there is a 

 strong tendency of the division plane to pass through endocoels. 



The proportions here found do not agree with those given by 

 Torrey and Mery for S. davisi. Their results, stated by the 

 method I have used, give twenty-nine cases of division planes 

 passing through exocoels to seventy-three passing through en- 

 docoels. These results, as Professor Torrey informs me, were 

 obtained from hand sections of the anemones. Some incomplete 

 mesenteries might be overlooked by that method. This might 

 be responsible for the recording of a somewhat greater proportion 

 of divisions in exocoels than I have found from the study of 

 microtome sections, but it seems scarcely possible that it can ac- 

 count for the whole discrepancy. Even if the divisions in incom- 



