143 



family including some of the genera which have been referred by me to the 

 family Bicellariidae. That it must be referred to the Scrupocellariidae and not to 

 the Bicellariidae is sufficiently evident from the stronger calcification and the 

 structure of the distal walls, the auicularia and the owcia. The last mentioned, 

 which are hyperstomial and the basal wall of which is a part of the frontal wall 

 of the zoa-cium, have a mostly membranous ectoooecium, which has only a 

 calcified marginal portion. The frontal gymnocyst is unusually large, whereas 

 there is but a very slightly developed secondary cryptocyst, which in the 

 oldest zocecia terminates in a number of tooth-like processes. The basal wall of 

 the zooecia is acutely arched, transversely striated and each radical fibre takes 

 its origin from a [)roximal pore-chamber. 



Family Membraniporidae. 



This family comprises all the Malacoslegous forms which can neither be refer- 

 red to the Cribrilinidae nor to any of the above-mentioned iiimilies, and which in 

 contrast to these can only be characterized negatively, viz. l)y their not possess- 

 ing the combination of characters peculiar to any of the above families. It 

 shows greater variation and wider contrasts than any of the other Malacostegous 

 families. 



The frontal wall of the zoacia is sometimes quite membranous, sometimes 

 to a greater or smaller extent provided with a calcareous layer, which may be 

 sometimes a gymnocyst (Electra), sometimes a cryptocyst (e. g. Omjchocella) and 

 most often a combination of both. Spines are sometimes wanting, sometimes found 

 in great numbers in the whole periphery of the frontal area. The separate zooe- 

 cia communicate sometimes by uniporous or multiporous rosette-plates, some- 

 times by pore-chambers. The heterozocecia have in some cases a calcified trans- 

 verse bar and may appear both as avicularia and as vibracula. They are 

 sometimes independent (vicarious), sometimes dependent, and sometimes both 

 forms are found together (Callopora craticula). The ocEcia are usually hyper- 

 stomial, in a single genus acanthostegous and in some cases endozoa-cial {Cale- 

 schara Rosseliana), sometimes (Oochiliiui) surrounded by kenozoa>cia. The colonies 

 are most frequently incrusting, but in many cases free and then either laminate 

 or forming richly branched tufts. Within this seclion so rich in species no small 

 number of genera and a few families have subsequently been set up or proposed, 

 e. g. by Busk, Waters, JuUien, Norman and others. Neither time nor my 

 material permit me to give a criticism of all the genera proposed, but I must 

 confine myself to set up a few new ones and to give new diagnoses of some 

 older ones. A grouping of the numerous species described, according to their 



