BRYOZOA. 349 



side by two close rows of exceedingly minute dots (see fig. 2ft). 

 Untabulated mesopores occur frequently, but in no case known, 

 vesicular tissue. The following genera are placed here: Ptilo- 

 dictya Lonsdale, Clathropora Hall, Tseniodictya, Phzenopora 

 Hall, Ptilotrypa, Graptodictya, Artbropora, Stictoporella, In- 

 trapora Hall, and Stictotrypa. Of these, the first, second and 

 third are distinguished mainly upon constant zoarial peculiari- 

 ties; the fourth has two mesopores occupying the transverse 

 space between the ends of the zooecia apertures; the fifth, very 

 oblique apertures and a small accessory cell in front; the sixth 

 and seventh have a filiform peristome and vermicular striae on 

 the interspaces; the eighth has more or less numerous meso- 

 pores, very thick-walled tubular zooecia, with dilating apertures, 

 and no hemisepta; the ninth also has numerous mesopores, but 

 thinner walled zooecia, with quite an abrupt aperture and an 

 inferior hemiseptum; the tenth has a decided peristome. The sys- 

 tematic position of the last (Stictotrypa) is somewhat doubtful. 

 (4) FEXESTELLID-E: This very characteristic Palaeozoic group 

 of Bryozoa presents us with an almost endless array of beauti- 

 ful forms, whose delicate net-work generally affords us proof of 

 the exceeding constancy with which, what may appear to the 

 uninitiated as very trivial zoarial features, recur in species after 

 species. The exceeding abundance of specific modification causes 

 a continual effacenient of sharp divisional lines, and I know of 

 no group better calculated to show that such do not occur in 

 nature. On the contrary, the more our investigations are ex- 

 tended, the more evident it becomes that our classifications are 

 necessarily largely based upon arbitrary divisions. Obviously 

 then, that classification is the best which offers the most con- 

 venient and at the same time the most natural (genealogical) 

 arrangement possible, a desire not by any means easily realized. 

 In my subdivision of the family I have been guided by the test, 

 that the value of a character is determined by its degree of 

 prevalence. Accordingly, when I have found a number of species 

 that agreed in one or more features without disagreement in 

 others known to have greater importance, I have regarded them 

 as entitled to recognition as a separate genus. Subgeneric di- 

 visions, being both inconvenient and useless, I am opposed to, 

 and have not adopted them excepting in such cases where they 

 seem to merit generic distinction. 



