ANEMONES 4OI 



that basal fragmentation is accountable for a considerable proportion 

 of them, since basal fragments are not uniform in size and number of 

 included mesenteries and the latter may serve as fundaments of the 

 mesenteries of the new polyp. Why there should be three mono- 

 glyphic types, according to Parker's figures, or one according to mine, 

 is a question that demands a more definite answer. It might be urged 

 that the fra'gments tend to be of a certain size and include such a num- 

 ber of mesenteries as would give rise to six, seven or eight pairs of 

 perfect mesenteries in the adult. In my experience the fragments vary 

 too much in size and in number of mesenteries to support this idea. 

 A more probable view was suggested by Andres in his work on Aiptasia 

 in 1882. Fragments of this species, irrespective of their size and num- 

 ber of mesenteries, took on a hexamerous form and, what is more, 

 showed indications of the two siphonoglyphs of the typical adult. I 

 shall return to the latter point farther on. Even this tendency of the 

 fragment to develop on the plan typical for the species hardly explains 

 the three subtypes so conspicuous in the Newport specimens, miless 

 these types are characteristic of both sexually and non-sexually pro- 

 duced polyps. 



Basal fragments probably develop under the influence of both fac- 

 toids — size and hereditary tendencies. But I do not think it necessary 

 to assume that all irregular polyps arise in this way or by means of 

 soine other non-sexual process. That all i-egular^ /iexa?nerous^ ino7io- 

 glyphic polyps are non-sexually produced seems still less probable. 

 This case will be considered in a succeeding paragraph. What it is 

 important to notice here is, that although we are ignorant of the origin 

 of eveiy polyp with an irregular number of mesenteries, it is per- 

 fectly well known that hexamerous polyps may develop non-sexually. 

 Further, since no one questions the fact that the Sagartids are funda- 

 mentally hexamerous and that the fundamental type is to be found, if 

 anywhere, among the polyps produced from the eg^, we may assert, 

 despite the want of direct embryological evidence, that hexamerous 

 polyps develop sexually also. Consequently, hexamerism is not deter- 

 mined by the method of reproduction. 



(<5) If now wc turn to an examination of the causes for variation 

 in the number of siphonoglyphs, we find that the method of repro- 

 duction is not one of these. I quote from my paper of 1S9S, p. 357, 

 " Both monoglyphic and diglyphic polyps reproduce by fission ; and 

 both result from fission (figs. 3-7). Further, a bud from the oesoph- 

 ageal region may possess either one or two siphonogl3-phs. One 

 bud produced upon the foot disk had, at the time of liberation, but 



