79 



In the genus Electra the median acropetal spine is much branched in E. bellula, 

 while the peculiar unilateral covering spine, which appears in the family Scrupo- 

 cellariidae is branched in a number of species, as in Scrupocellaria reptans and 

 Menipea aculeata. Further, branched oral and marginal spines appear in larger 

 or smaller number in Membraniporina cornigera, M. proteda, in several varieties 

 of Microporella Maliisi, in two species of the genus Chaperia (C. anmilus and C. 

 cervicornis), in Hiantopora radicifera and in some species of the family Cribrilini- 

 dae. The differences mentioned in the shape of the spine have, as will be seen 

 from the given examples, generally a very slight systematic importance, and are 

 not even always an expression for a difference of species. In the same way as 

 there is hardly any family or genus, except those very poor in species, in which 

 all the members have spines, so there are hardly many species, in which the 

 number of spines is constant, and in many cases the variation is very consider- 

 able. We may give here a few instances of the variation in number of the spines 

 in genus and species. In Smittina the number varies between and 8, in Escha- 

 rella between and 10, in Chaperia between and 8 and in Callopora between 

 and 13. In Electra monostachgs the number varies between 1 and 18, and even 

 the one is not always present, in E. pilosa between 4 and 12, in Callopora lineata 

 between 6 and 12, and in Bugula Murrayana between 3 and 8. In discussing the 

 systematic importance of the spines, we must still notice that the family Cribri- 

 linidae is based solely on the mutual relation of the marginal spines, and that a 

 plate-shaped or branched covering spine only appears in the family Scnipocellarii- 

 dae. To this we must still add that this spine is far from being found in all the 

 species of the family, and that the family Cribrilinidae is undoubtedly not a 

 natural one. In contrast to the generally great inconstancy and variation of the 

 spines, it may be mentioned that marginal spines are always lacking in numer- 

 ous families, which are mostly rich in species, e. g. Farciminariidae, Cellulariidae, 

 Steganoporellidae, Thalamoporellidae, Adeonidae, Catenariidae, Celleporidae, Hippo- 

 thoidae, Mgriozoidae and Tubucellariidae. 



Finally, we may briefly refer to a new Callopora species, from the Fseroes, 

 which apparently shows the largest amount of variations in the number, struc- 

 ture and mutual relation of the spines, which have yet been found in any Bryozoa. 

 While some zocecia only have 4 short spines, others have a very varying number 

 of longer ones, which sometimes have the same breadth in their whole length, 

 sometimes the tip expanded or bifurcated. These longer spines in more or fewer 

 zocecia may be united with one another to form a cover, pierced by transverse 

 furrows, just as we find in the species of the genus Membraniporella. 



The primary aperture. In a number of species we meet two different forms 



