Division in the Radiata. 395 



stomttj ice have not to do loith the division of a true ^'' individuaV 

 in the strict acceptation of the term, but with that of a creature 

 which is really very simple, but potentially and in nuce is 

 a multiple creature, which Avould resolve or develop itself into 

 a long series of individuals. I therefore confine myself to refer- 

 ences to those cases in which a trae artificial or spontaneous divi- 

 sion has been observed in the Medusee — namely, by M, Hackel 

 in the species of the group Thaumantias, and by M. Kolliker in 

 young Mesonemata{Stomohrachium) . If, however, Hiickel's ob- 

 servations relate to Medusae of the category which Mr. AUman 

 subsequently termed "Gonochemes" this case would have to be 

 assimilated to those which, for reasons explained above, fall 

 outside the considerations here involved. After mentioning 

 the spontaneous and artificial division of Protohydra, a type 

 which is still but little known, and also the artificial division 

 of Lucernaria and the spontaneous division of Schizocladium 

 (which appear to me to have just as much resemblance to 

 gemmation or sporogony), I take up the examination of the 

 faculty of regeneration and the divisibility of the ActiuijB, 

 commencing with the remarkable investigations of Dicque- 

 mare, which are abeady a century old, and then passing 

 in review the more recent experiments and observations 

 of Gosse, Peach, Bennett, Dalyell, M'Cready, Mrs. Thynne, 

 Van Beneden, Wright, and Hogg. From these it appears 

 not only that the Actiniee may be divided longitudinally and 

 transversely, but also that spontaneous division is by no means 

 a rare phenomenon, at least in some species {A. cereus, dian- 

 thus, cavernosa) ; in an Actmid probably produced from the 

 ova of Caryophyllia Smithii it was even repeated so frequently 

 that, in the course of two years, two individuals were thus in- 

 creased to 278. By a similar longitudinal division the frequent 

 appearance of double Actinise more or less joined together has 

 also been explained ; but their formation may be interpreted 

 in a different way. 



Although in the Actiniaj (except, perhaps, Corynactis) gem- 

 mation cannot be regarded as a normal mode of reproduction, 

 as in the compound Anthozoa, it nevertheless occurs sometimes 

 in them — partly in the same species which are subject to sponta- 

 neous longitudinal division, and (as in the supposed Caryo- 

 phyllia Smithii) in conjunction with this latter operation. 

 Consequently these two modes of agamic reproduction, different 

 as they are in form, evidently replace one another, and cannot 

 differ much in their essence ; moreover, when the division is a 

 little excentrical, and one of the parts is a little smaller than 

 the other, we may ask ourselves whether the division is not 

 really a disguised gemmation. The limit is no less difficult to 



