SYSTEMATIC PART. 161 
But while the armlets of the Silurian Botryocrinus bear no pinnules, those of 
the Subearboniferous Steganocrinus give off pinnules from the armlets. 
Bather’s classification and theories respecting the evolution of the arms 
are open to many criticisms, and are certainly not confirmed by a study of 
the Carboniferous Fistulata. He cannot put together certain Poteriocrinide 
with the Dendrocrinidx, and others with the Decadocrinidx, nor place the 
non-pinnulate Homocrinus and Parisocrinus in a family with pinnule-bearing 
forms, His classification is based principally upon two things: the presence 
or absence of infrabasals, which he makes a subordinal character, and the 
relations of the plates at the anal side among themselves and toward adjoin- 
ing plates, upon which he separates the families. He was perhaps not aware 
that in Baryerinus the radianal may be present or absent in the same species, 
and that Parisocrinus would be substantially identical with Cyathocrinus but 
for the presence of the radianal. As family characters he also relies upon 
the mode of branching in the arms, and the structure of the tegmen, which 
latter he has been able to observe in only a few genera. The remarkable 
development of the posterior area into a ventral sac, which in 1890 was 
regarded by him as an excellent ordinal character, is omitted altogether in 
his present classification. 
While therefore we cannot agree with Mr. Bather upon his classification, 
we fully acknowledge the excellence of his specific and generic descriptions 
in his late work on the Swedish Inadunata, and the many fine observations 
which he has brought out. His discoveries upon the orientation of the radials 
in the Pisocrinide and Calceocrinide are of the utmost value, and have 
thrown new light upon these difficult groups. 
We regard as the best family distinctions among the Fistulata the 
presence or absence of infrabasals, the presence or absence of pinnules, 
and the relative size of the ventral sac. The structure of the tegmen, if we 
knew more about it, might perhaps also afford good distinctions; but the 
modifications that occurred at the anal side of the cup, and the mode of 
branching of the arms, can be utilized only for distinguishing genera. The 
biserial arm structure did not obtain a foothold among the Inadunata until 
near the close of the Subecarboniferous, and at no time became a constant 
character. More important for generic separation is the mode of union 
between the radials and brachials, and the form of the facet — whether 
horse-shoe shaped or forming a straight line; also the form of the ventral 
sac, and whether one or more of the radials are compound. 
21 
