1896.] Mr Harmer, On Cyclostomatous Polyzoa. 211 



characters of the ovicells is very apparent is Hornera ; but without 

 going further into this case, I may be permitted to compare three 

 recent species of Diastopora (Berenicea in Dr Gregory's sense) 

 which I have examined with the special object of testing Dr 

 Gregory's views. In each case I have succeeded in observing the 

 continuity of the ovicell with a cylindrical portion situated in the 

 ordinary series of zocecia and representing the condition in which 

 the ovicell occurred before it took on the external characters of an 

 ovicell. This condition was in Idmonea serpens shown to be 

 associated with the fact that the ovicell is a modified zocecium, 

 and I have little doubt that the same conclusion may be drawn 

 with regard to Diastopora. I wish to insist on this point because 

 Dr Gregory has figured ovicells, in various species of Berenicea, 

 which from my examination of recent genera I have no hesitation 

 in regarding as modified zocecia (cf. PI. III., fig. 3 ; PI. IV., figs. 1, 2, 

 5, 6 of his Catalogue). These he terms " gonocysts," and on p. 12 

 he defines a gonocyst as " A form of marsupial chamber produced 

 by expansions within the zoarium, and not by the modification 

 of a single zocecium." I believe, on the contrary, that the ovicell 

 is probably a modified zocecium in all Cyclostomata. This is 

 certainly true of Crisia, Idmonea serpens and Lichenopora. 



In Diastopora suborbicularis, Hincks, the fertile zocecium is 

 but slightly expanded distally, as in Dr Gregory's PI. III., fig. 3. 

 In D. obelia, Johnst., the dilated part of the ovicell is very much 

 more developed, while in D. patina, Lamk., the dilated part grows 

 out laterally on each side in a direction concentric with the margin 

 of the colony. The lateral growth may be so marked that the 

 dilated part of the ovicell grows completely round the discoidal 

 colony, and with the secondary union of the two lobes of the fan- 

 shaped colony, the two lobes of the ovicell may also meet one 

 another and form a complete ring-shaped ovicell. Should a second 

 ovicell be developed as a modification of a younger zocecium than 

 that which produced the first one, it is developed concentrically 

 with the first and of course outside it. 



The specific difference between D. suborbicularis and D. patina 

 is further expressed in the slight development and vertical position 

 of the " terminal membrane" 1 in the former, and its large develop- 

 ment, and oblique position (owing to the great extent of the basal 

 lamina) in the latter ; and in the fact that the fan-shaped form of 

 the colony is developed into a secondarily discoidal form much 

 earlier in D. patina than in D. subo?'bicularis. 



These examples will serve to show that recent species of 

 Cyclostomata may be distinguished by means of their ovicells. It 

 will further be apparent that the ovicells themselves belong to 



1 Quart. Journ. Micr. Sci. xxxix., p. 93. 



