122 



very distinctly marked transverse belt almost midway across the frontal surface 

 of the ooecium. The part of the ocecium lying proximally to this is furnished 

 along the middle with a narrow ridge. This cryptocyst-bridge must undoubtedly 

 have arisen from a fusion of two triangular laminae like those we have described 

 in Columnaria borealis. 



Family Flustridae. 



.The zocecia slightly calcified, with an aperture which occupies the whole 

 frontal surface, or at any rate its largest part. Occasionally there is found a 

 secondary cryptocyst. The distal wall is always provided with a varying number 

 (1 — 13) of small, uniporous rosette-plates, and such also appear as a rule on the 

 side walls, which only in a few cases are furnished with multiporous rosette- 

 plates. Vicarious or independent avicularia. The ocecia are endozooecial and immer- 

 sed, generally in ordinary zocecia, occasionally in avicularia or kenozocecia. The 

 colonies are in a few cases incrusting, in most cases free frondose, more or less 

 richly branched, and with the free margin consisting of kenozocecia. 



As the family is defined here, the main weight is laid on the possession of 

 immersed ocecia and vicarious avicularia, as well as on the slight calcification 

 and the large frontal aperture, and I have therefore also referred ^ Membranipora^ 

 flustroides Hincks and M. serrata M. Gill, to this family; the latter species has 

 been considered by Waters also as a Flustra. In conformity to the above defini- 

 tion of this family, I have been obliged to separate out a number of species, 

 which partly have external ooecia, partly dependent avicularia. yFlustra<!^ militaris, 

 ^Fl.i- crassa, ^Fl.^ dissimilis and >Fl.< nobilis are thus referred to the J3ice//an'idae and 

 »FZ.« armata to the Scrupocellariidae. Since however the ooecia and avicularia are 

 lacking in a number of species of this fanjily as in most other families, and as 

 a number of Membranipora species can have vicarious avicularia as well as a 

 quite uncalcified frontal wall, it is difficult to draw a sharp line between this 

 and the family Membraniporidae. Membranipora serrulata Busk is a species which 

 has been regarded both as belonging to Membranipora and to Flustra. According 

 to Busk's original description it possesses immersed ocecia, and if this were cor- 

 rect, it would have to be regarded as a Flustra, but I have not succeeded in 

 finding ooecia in any of the specimens of this species, which our Museum has 

 from the Kara Sea or from Greenland, nor are they found on Busk's original 

 specimens in the British Museum. The species appears incrusting as well as in 

 free, bilaminate growths, but it differs from the Flustra species, known to mee in 

 having multiporous rosette-plates on the distal wall, as well as fully developed 

 marginal zocecia, and I therefore find it more natural to look upon it as a Mem- 



