CELLULAEINE AND OTHER POLYZOA. 321 
be regarded as the more primitive. I am not sure that this is always the case,, 
and T think it is at least possible that a multiserial branch may in some cases 
have been derived secondarily from a preceding biserial condition. I do not 
consider it necessary to discuss this question in the present paper, in which 
I desire to point out, however, that the extent to which the splitting oi: the 
original lamina has taken place differs in various genera and species ; and 
that the study of the actual facts of the bifurcation of the branches may be 
of sreat assistance in systematic work. So far as I am aware, there has been 
no previous comparative account of this subject, although the importance of 
the mode of branching has not been overlooked by other writers. I may 
refer especially to the account given by Davenport (1891, Bull. Mus. Comp. 
Zool. Harvard, xxii. p. 41) of the branching of Bugula, and to two papers by 
Waters (1897, J. Linn. Soc. xxvi. p. 2 ; 1913, Proc. Zool. Soc. p. 473), 
dealing with the mode of bifurcation in Scrupocellaria and Menipea. These 
papers record important facts, but they do not exhaust the subject. 
Davenport is concerned with the general laws of branching, and he does 
not attempt to show how his results can be applied to classification. In his 
1913 paper Waters divides Menipea into several groups, based on differences 
in the mode of branching ; but I think he has not been altogether successful 
in his conclusions. Waters (1913) defines a group " C," for instance, which 
he supposes to include species having a common type of bifurcation ; but he 
places in it Menipea patagonica and Bugulopsis peacliii, two species which, 
according to my own results, differ essentially in their mode of bifurcation. 
In studying the bifurcation, particularly in the Family Scrupocelhiriida;, 
it is necessary to devote special attention to the formation of the chitinous 
joints with which many species are provided. In the majority of cases, each 
of the branches is jointed at its base, in such a way that the actual bifur- 
cating point is immediately succeeded by a joint at the origin of each branch. 
In some cases, only one of the branches is thus jointed ; an arrangement 
which may give rise to a sympodial form of stem, unilateral when the joints 
all occur on the same side, or bilateral when they are formed alternately on 
the two sides. The joint is developed in a manner which is remarkably con- 
stant in the most diverse members of the calcareous Polyzoa ; occurring in 
fundamentally the same way in (!yclostomata (Crisia) and in a variety of 
Cheilostomata, which seem to have no close connection with one another. 
The process has been described by various authors, among whom I may 
mention Waters (1881, Q. J. Geol. Soc. xxxvii. p. 320 ; 1887, Ann. i\Iag. 
Nat. Hist. (5) xx. pp. 89, 92 ; 1897, J. Linn. Soc. xxvi. p. 2 ; 1913, Proc. 
Zool. Soc. p. 472) ; Pergens (1890, Bull. Soc. Beige Geo], iii. p. 313) ; and 
Lomas (1889, Proc. Liverpool Biol. Soc. iii. p. 219). The joint is formed 
across one or more zooecia, the calcareous walls of which are at first complete 
and continuous. A chitinous lining is developed on the inner side o£ the 
calcareous wall and in close contact with it (c/. Pergens, text-fig. 8, on 
