ASEXUAL REPRODUCTION IN SAGARTIA 171 



the plane of division, should also be excluded from a group on 

 which to base a study of the relative frequency of divisions in 

 different spaces. Such a case is illustrated in the newer re- 

 generating region represented in figures 15 and 16. Therefore 

 only such specimens have been included in table 6 as show the 

 characteristic differences between new and old complete mesen- 

 teries. This introduces a slight error through the exclusion of 

 such a specimen as that represented in figure 17, which has no 

 old complete mesenteries. Furthermore, since the specunens of 

 table 6 are to be regarded as a random sample, so far as concerns 

 the position of the division plane, there have been excluded from 

 it all animals selected for sectioning because of the relation of the 

 new area to the directive plane or to the orange stripes which 

 mark the complete endocoels and certain incomplete endocoels. 



It is to be noted that the plane of division is in almost all 

 cases strictly vertical. Occasionally a mesentery is found in an 

 oral or aboral region, but not throughout the length of the 

 column. Such mesenteries, as would be expected, are usually 

 small. That they represent mesenteries torn during fission is 

 probable, in spite of the fact that not a single case has been 

 noted in which both paired individuals show parts of the same 

 original mesentery. Since such partial mesenteries are rare in 

 older parts, it is likely that they either are normally completed 

 by regeneration, or are absorbed. As I have seen nothing what- 

 ever to indicate that a torn mesentery ever grows up or down 

 the column, and as there is some evidence (p. 172) that absorption 

 of mesenteries does occur and that regulative processes correct 

 certain abnormal adhesions, it is likely that the partial mesen- 

 teries referred to are removed by absorption during regeneration. 



Certain other indications of departure from a perfectly 

 simple vertical tear may be mentioned. In pairs 3 and 10 of 

 table 3 there is disagreement between the bounding mesenteries 

 of the related individuals. In no. 3a an incomplete mesentery 

 of the most advanced cycle (I) stands in the position of a mate 

 to a complete mesentery, c, in no. 3b. In no. 10a an incomplete 

 mesentery of doubtful grade, (1) is similarly opposed to a direc- 

 tive, d, in no. 10b. It is entirely improbable that in these cases 



