ANATOMY AND EVOLUTION. 
453 
number 12 is reached, the older compartments and corresponding tentacles 
always lying nearest to the 2 primary compartments at both ends of the 
mouth, the youngest nearest to the primary septa. When the number of 
tentacles is 12, they arrange themselves in two cycles of 6, the inner being 
larger (often elevated), and the outer smaller (depressed). [The cnido- 
phorous filaments also appear symmetrically, in pairs, quite independently of 
the supposed sixfold order of appearance.] The next 12 tentacles (making 
24) are not developed alternately with cycles I. and II., but in pairs in six 
of the intervals between them; afterwards a new regulation takes place, 
through the swifter growth of one of the newly formed tentacles of each 
pair, producing, by this substitution, three cycles, G+G-f 12 = 24, and so on, 
the new tentacles always budding forth in pairs in the compartments of the 
smaller tentacles belonging to the highest cycle. The tentacles of the same 
cycle (I., II., III., &c.) are consequently not homologous, those of the 
highest cycle (according to the system of Milne-Edwards and Ilaime, the 
youngest !) containing, in fact, tentacles of every age, from the oldest to 
the youngest. 
Asexual Reproduction. Reproduction through buds has been observed by 
SuMruii (14) in 3 species of RJwdopsammia, in which the buds remain con- 
nected with the nurse polype for a rather long period ; and in RIastotrochuSj 
where 1-G buds are formed in a row along each of the edges of the polypary. 
They are soon detached, leaving a short stem (with septa), from which new 
buds are probably produced. Sometimes also this basal portion is detached, 
leaving a scar, from which 2 buds occasionally spring. In Flahellum and 
Tlaeotrochus the same species occur both in the attached (pedicellate) and free 
(truncate) state ; e. g. F. aculeatum divides transversely into a superior (trun- 
cate) portion (F. oweni^ stokesi) and an inferior (truncate) {F. spinosujn). The 
author supposes that the latter has the faculty of regenerating the lost parts 
and of prodiicing by repeated transversal fission a series of truncate Flabdla, 
as Scyphostoma produces Mcdmm. In Fungia young disk-shaped individuals 
are developed upon branched stems, from the summits of which they are 
severed, following horizontal lines of division. From the alternate dilata- 
tions and constrictions of the stems, the author concludes that new Fungia- 
disks are repeatedly developed on and detached from their terminal surface ; 
and he correctly compares their repeated gemmation with the transversal 
division ” of Scyphostoma. Observations are also made on the regenerative 
power of Fungia^ and the manner in which fracture or other abnormal con- 
ditions (c. g. turning over) will induce the formation of compound or budding 
individuals ; and it is shown that Diaseris (as suggested also by the Recorder) 
is not formed through the coalescence of originally distinct lobes, but divides 
spontaneously or through some extraneous influence into independent indi- 
viduals, producing through regeneration (in the same manner as the muti- 
lated Fungioi) compound (polystomoiu?) Corals. The artificial divisibility, 
natural fission, and occasional gemmiparity of the Anthozoa {Actiniidm and 
Madreporaria) are discussed by the Recorder in his paper on spontaneous 
division in the Radiata (ante, p. 438), where it is pointed out that many phe- 
nomena which (e. g. in Euphyllia, Mussa, Mceandrina, &c.) are described as 
fissiparity, are in fact gemmiparity — the half-separated or half-united poly- 
pites being originally distinct, afterwards more or less coalescing, the sys- 
tematic value of the presumed difference being of course quite illusory. 
