MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 421 



breezes having' subsided, the waters ai'c quieter and more favorable to tht^ activities of delicate 

 sessile animals. 



Perhaps the activity of the unicellular commensal algte, pi-esent in such great numbers in the 

 endodermal tissue of nearly all species, may be associated with the changes. But fully expanded, 

 more transparent tissues in the daytime, would manifestly be most favoi'able for the functional 

 activity of their chloroplasts. 



TENTACLES. 



The tentacles of Madreporarian polyps exhibit a certain diversity of form and arrangement, 

 though not to the same degree as the corresponding organs in the Actiniaria. They are mostly 

 di.sposed around the margin of the oral disk, in two or more entacma?ous, alternating cj'cles. In 

 Madrejx/ra and Pontes, however, they appear to constitute onh' a single cycle. In living pol\-ps 

 the tentacles can usually be seen to correspond in position with the internal mesenterial chaml)ers 

 of which they are the external prolongations, and further to conform in position, and, as a rule, 

 in number, with the internal skeletal septa. In nearly all cases they correspond exactlj^ with the 

 number of internal mesenterial chambers, both entocoelic and exocoelic. Tentacles arising from 

 the entocoelic mesenterial chambers may be known as entofeyitades, and those from exocoelic 

 chambers as exotentacles. Where present the latter always constitute the outermost cycle, and 

 all the inner cj^cles consist of entotentacles. In Sidernstnpa the exotentacles differ in form from 

 the entoccelic meml)ers (PI. XXII, tig. 151), while in Ayaricia, and the fully developed ajjical 

 polyps of Miidrepora, exoccjelic tentacles are wanting. Gardiner (1900, p. 365) also found the 

 tentacles to be entocoelic onl}' in Cmnopsammla. 



In all coral polj-ps so far described onl}- one tentacle arises from each mesenterial chamber. 

 The number of tentacles therefore represents the actual number of mesenteries present, and, in 

 general, the number of septa also. None of the species examined shows the stichodactylinous 

 condition so prevalent among tropical Actiniiv. " 



During extension the tentacles are usually elongated, broad below and narrow above, the 

 walls thin and somewhat transparent. Most of the species studied are characterized by a 

 white, opaque, knob-like apex, more or less distinct, and constituting a batter\' of nematocysts. 

 The tentacular stems of coral polyps are rarely smooth throughout, but exhibit round, oval, 

 or ii-regular opaque thickened patches, which, like the apical knob, are aggregations of 

 nematoljlasts. These are elevated a little from the genei'al surface, but rarely show any spiral 

 or other regular disposition. Such restrictions of the tentacular nematoblasts are very excep- 

 tional simong Actinians, and in this gi-oup the knobbed condition is also unusual {Corynacfis, 

 Coralli>iiorj>hut<). 



With the excejition of SUlerastra'a rddians and S. niderea the tentacles of all the species here 

 described are simple, while in the genus mentioned the entotentacles become bifurcated toward 

 their free extremity, Init the exotentacles remain simple. 



In the living polyp the tentacles assume varied positions. During retraction they are 

 usually withdrawn within the calice, and completely hidden hx the overfolding column wall; but 

 in some genera, SlderdntrsM and Agaricia, they remain exposed under all conditions of retraction 

 or expansion of the polyps. In Pot ites and other forms the tentai'les may occasionally remain 

 exposed on retraction of the polyps, though more usually hidden under the retracted column wall. 

 Upon expansion of the polyp the organs stand erect or overhang, even to such an extent as to 

 nearly hide the column wall; and on the same polyp different cycles may sometimes assume 

 different attitudes, as where the inner cycle is erect and the outer overhangs. The tentacles of 

 corals rarely display much independent motion when fully extended, compared with the activity 

 exhibited by the long tentacles of anemones. The tentacles of CJadorom and Sidei'dsfrxn. 

 and probably others, possess considerable adhesive power, more especiall}' at the apex; the 

 distal part of the stem may also fold round any small object. When small annelids are placed 



"The term is applied to polyps (e. g., Corynactis, Discosoma) in which the tentacles are arranged in radial rows, 

 so tliat more than one tentacle comrannieates with a xingle mesenterial chamber. The character serves to distinguish 

 the tribe Stichvdacti/linn: from other Actiniaria in which only one tentacle communicates with a mesenterial interspace. 



