214 University of California Publications. izoologt 



in length, at the same time encroaching' in its lesser diameter 

 more and more upon the column, until, with a tear across the 

 mouth disk, the independence of the two moieties is established. 



Such, iu general, is the process of fission; but there are 

 several facts connected with it which should not be overlooked. 

 The division, which is usually approximately equal, may be very 

 unequal; in rare instances, a polyp is divided into three parts, 

 two large and approximately equal, the third very small. In 

 every case, however, the fission plane passes through the mouth 

 disk, and almost invariably through the mouth also. When the 

 mouth is involved, the fission plane always passes approximately 

 perpendicular to its major axis. If the dividing polyp be di- 

 glyphic, the division (into two) gives one siphonoglyph to each 

 portion.* It has been frequently observed that polyps resulting 

 from fissiou themselves divide, and in every case the second 

 fission plane parallels the first, that is, it also passes perpendicular 

 to the major axis of the mouth. The second division may 

 siicreeil the first before the regeneration of a second siphonoglyph, 

 as sections show, so that not only may division occur in mono- 

 glyphic polyps, but in such cases, may give rise to polyps which 

 have no siphonoglyphs at first Rearrangements of mesenteries 

 foreshadowing both first and second divisions may occur together 

 in the undivided polyp, in rare eases. 



With respect to the relation of the fission plane to the 

 mesenteries, it can be said that among 51 polyps resulting 

 from fission, sections taken before new mesenteries had had time 

 to regenerate and complicate the investigation, showed that ten 

 had resulted from division through exocoels, thirty-two from 

 division through endocoels (in a large but unrecorded majority 

 of cases, between mesenteries which reached the oesophagus), 

 and nine from division through an exocoel on one side and an 

 endocoel on the other. 



The rate of fission varies within rather wide limits. The 

 process may begin and end within twenty-four hours, as in 

 8. luciae also (Davenport, '03), or it may require weeks for 

 completion. Experiments indicate that the food supply may be 



; if. M.. dianthus (Torrey, '98), in which species tin- fission plane is parallel to 

 tlit- greater axis ni' the mouth, and divides the one siphonoglyph in monoglyphic, 

 one or both in diglyphic polyps. 



