1896.] Mr Harmer, On Cyclostomatous Polyzoa. 213 



has expanded widely so as to cover the proximal part of the colony 

 and thus to give rise to the discoidal form characteristic of the 

 adult 1 . The mouth of the young funnel is closed by the uncalcified 

 " terminal membrane." This, during the growth of the colony, 

 gives rise to zocecia in such a way that a part of it, capable of 

 forming new zooecia, is left round the entire circumference of the 

 colony, in addition to the part which appears to cover the entire 

 upper surface, and from which the " cancelli " are developed. In 

 Lichenopora the adult colony assumes the form of a somewhat 

 modified funnel, but in Floscidipora pygmaea, MacGill., an Aus- 

 tralian species, the funnel-shaped form persists with practically no 

 modification. 



The peculiarities of Lichenopora are thus partly dependent on 

 the mode of development of the terminal membrane, which forms 

 an uninterrupted, annular margin round the mature zocecia and 

 ovicell. 



In another series of forms, including Idmonea serpens, Tiibuli- 

 pora and Diastopora, the proximal part of the terminal membrane 

 is used up completely in the formation of the proximal zooecia. 

 The terminal membrane thus becomes incomplete on the proximal 

 side of the colony, but it remains as a more or less semicircular 

 margin, capable of giving rise to new zocecia, on the distal side. 

 Growth thus ceases on the proximal side, and the colony assumes 

 a fan-like instead of a circular shape. The terminal membrane is 

 commonly used up in the formation of zocecia at a point opposite 

 the primitive disc of the colony, and the growing margin is thus 

 divided into two. By further growth of these parts, accompanied 

 by a cessation of growth in the middle of the distal side of the 

 colony, the fan becomes bilobed ; and by further subdivisions of 

 the parts of the terminal membrane, each lobe becomes further 

 lobed. This interruption and subdivision of the terminal mem- 

 brane is one of the essential differences between Lichenopora and 

 Idmonea, and it is a noteworthy fact that this generic distinction 

 is already marked at the commencement of the life of the colony. 

 I do not at the present moment include the branching species of 

 Idmonea in this comparison ; and it is possible that the affinities 

 of /. serpens are really with species of Tnbidipora, as many authors 

 have assumed by placing it in that genus, rather than with forms 

 like I. atlantica, I. radians and other similar species. 



In recent species of Diastopora, the colony early acquires a fan- 

 like shape. The terminal membrane does not, however, divide, 

 and the colony does not assume a lobed form. It is well known 

 that in this genus the two lateral edges of the fan grow towards 

 the proximal end of the colony, where they unite secondarily with 



1 Quart. Journ. Micr. Sci. xxxix., pp. 81, 93. 



