132 



and fïnally expanding to form an everted lip which surroünds the ooeciopore. In the majority 

 of cases the ooeciopore is transversely elongated and looks frontally and distally; but rarely 

 it faces in a proximal direction. It is usually symmetrically situated, near the middle line of 

 the ovicell. 



The development of the lip of the ooeciostome is variable. In fig. i it is wider than in 

 fig. 4; and in some cases it is wider still. In other colonies, on the contrary, the lip is hardly, 

 or only slightly developed; and even the basal dilatation may be absent. But the same colony 

 may show one or two ooeciostomes of the typical form, even thongh other ovicells have this 

 structure in what may be considered the reduced form above described. 



In nearly all cases, each ovicell has a single ooeciostome, even when it bifurcates with 

 the branch. But in specimen 350. C, where the ovicell may bifurcate once or even twice, 

 several accessory ooeciostomes are present. One or more of these. may have the typical form; 

 but they are usually simpler in character. A case of this kind is represented in fig. 2, where 

 the ovicell has two main branches, each of which shows the indication of a second bifurcation. 

 Each of the principal branches has three ooeciostomes (t>) facing in various directions. The most 

 proximal ooeciostome in the branch of the left side is more or less of the typical form, although 

 not fully developed and without an everted lip-, and this is perhaps the normal ooeciostome 

 of the ovicell. The other five are less like what I have described as the typical ooeciostome 

 of the species. 



The form described by Kirkpatrick as Idmonca pulcherrima (see synonomy) agrees 

 closely with 'Siboga' specimens (e. g. 99. A.), as I have convinced myself by an examination 

 of the type-specimen (89. 8. 21. y^i) m tne British Museum, in the form of the colony, in its 

 reticulate arrangement, in its zooecial characters, and in its ovicells and ooeciostomes. The 

 description given by that author makes no reference to the occurrence of more than one 

 ooeciostome in an ovicell; but his fig. 6a shows an ovicell with two ooeciostomes of the typical 

 form. I have previously called attention to the occurrence of accessory ooeciostomes in Tubu- 

 lipora aperta 1 ), from Norway. 



The polypides have the loop of their alimentary canal unbent during retraction : — a 

 character which seems to be general in Cyclostomes,. and is probably associated with the long 

 and narrow shape of their zooecia. The distal ends of the tentacles include a considerable 

 number of the "excretory vesicles" which I have described in other species of Tubulipora~)\ 

 and similar structures appear to occur in the embryos. The retractor muscles diverge from a 

 single point on the wall of the zooecium, which is situated at some distance on the distal side 

 of the apex of the caecum of the stomach (in the retracted polypide). 



The primitive disc of the colony has no marginal denticulations. 



The present species appears to be closely related to T. milneana D'Orb. 3 ) and T. 

 interjuncta MacGill. 4 ), if indeed it is not identical with one or both of those forms. If these 



1) Hakmeu, S. F., 189S, "Dev. Tubulipora\ Quart. J. Mier. Sci., XI.I, pp. 101, 104. PI. VIII, fig. 2. 



2) t. cit., p. 113. 



3) D'Orbigny, A., 1S46, "Voy. Am. Mérid.", V, 4 e Partie, "Zoophytes", p. 20 (as Idmonea milneana\ PI. IX, figs 17—21. 



4) MacGii.uyray, 1'. II., 1SS6, u Descr. new Pol", IX, Tr. Pr. R. Soc. Vict., XXII, p. 137 (as Idmonea interjuncta). 



132 



