BRYOZOA. 363 



external resemblance, but little can be brought showing true 

 relationship. On the contrary, in the form of the zooecia there 

 is close agreement between them and the STICTOPORHLE. It also 

 appears highly probable that the ARTHROSTYLID.E constituted a 

 distinctively Silurian type, since no representatives are as yet 

 known from either the Devonian or Carboniferous systems. 



The jointed condition of the zoarium is the most conspicuous 

 and perhaps also the most important feature of the family. It 

 is well shown in all the genera excepting Nematopora, in which 

 the zoarium is a dichotomously branching stem above the basal 

 articulating extremity. (PI. XXIX, fig. lla). In Helopora 

 Hall, Sceptropora Ulr., and Arthrostylus Ulr., the segments are 

 simple and terminally joined together. In Arthroclema Billings, 

 the zoarium forms a plumose expansion, and is divisible into 

 numerous primary, secondary and tertiary segments, those of 

 the first and second order being connected terminally and ranged 

 in straight lines. A deep socket occurs on two opposite sides 

 of each of the joints of the first and second series, in which the 

 first of the series of the second and the third order is inserted. 



The zooecia are arranged in a radial manner around a central 

 axis, and, excepting Arthrostylus, open on all sides of the sub- 

 cylindrical segments. In the excepted genus one of the sides is 

 marked only with longitudinal striae. The primitive cell is 

 wedge-shaped, but inclined to be tubular in the larger species. 

 As nearly all of the zooecia commence at the center of the zoar- 

 ium the reason is obvious; their form must adapt itself to the 

 distance between the axis and the surfaces, and as it is greater 

 in the large species than in the smaller, the primitive cell ap- 

 pears, consequently, somewhat tubular (compare figs. 15 and 

 18) . Ordinarily they are oblong quadrate or hexagonal in out- 

 line, and aside from their wedge-like shape (due to their arrange- 

 ment) agree closely with those of the STICTOPORID.E. "Hemi- 

 septa" I have not observed. 



(9) RHABDOMESONTID.E: This interesting family must be re- 

 garded as forerunners of the VINCULARID^. Indeed the external 

 resemblance to species of that later type is often so very striking 

 that a real affinity between them is forcibly suggested. Minute 

 comparisons show, however, that the differences which distin- 

 guish the groups of Palaeozoic Bryozoa from their more recent 



