162 ME. A. W. WATERS ON 



stomata arises, in the first place, from the small number of 

 readily available characters, but some of the confusion results 

 from the fact that so many attempts at classification have been 

 based upon fossils ; and here almost valueless characters have 

 been used, in consequence of which very exaggerated ideas as to 

 the number of Cretaceous and other fossil Cyclostomata have 

 been received. 



Formerly, in the Chilostomata generic divisions were based 

 upon the zoarial form, but now the zooecial characters are known 

 to be of chief importance, and the zoarial shape is hardly ever of 

 more than specific value, though there may be exceptions, such 

 as the Catenicellidse. On the other hand, in the Cyclostomata a 

 large number of species and genera have been, and are still, 

 based entirely upon the shape of the zoarium ; and until other 

 characters have been made available these groupings must con- 

 tinue to be used, to enable us to record what has been met with, 

 even though these may be catalogue genera and not natural 

 genera. 



The number of characters available can be increased; and 

 since Smitt published his papers on the Northern Bryozoa, the 

 importance of the ovicell in the classification of the Cyclo- 

 stomata has been more fully recognized. Smitt was, moreover, 

 the first who gave a number of figures of ovicells of Cyclo- 

 stomatous Bryozoa. In my papers I have always referred to 

 the ovicells where there was suitable material, and have dealt 

 with some in two communications *, and since then Dr. S. F. 

 Harmer has published a most valuable series of papers on the 

 embryos and ovicells of Crisia t, Lichenopora J, and Tuhulipora § ; 

 and it is now fully established that useful classificatory characters 

 are furnished by the position and shape of the ovicell, as well as 

 by the nature of the oceciopore. We must see how far the 

 number of tentacles is of value ; then the position and form of 

 the closures || are other ustful characters ; and there are ovarian 

 and sexual characters which must be examined, such as the size 



* "On some Ovicells of Oyclostomatous Bryozoa," Journ. Linn. Soc, Zool. 

 vol. sx. p. 275 ; and " On the Ovicells of some Lichenoporffi," op. cit. p. 280. 



t Quart. Journ. Micr. Sci., n. s. vol. xxxii. pp. 127-181, pi. 12. 



X Ibid. vol. xxxix. pp. 71-144, pis. 7-10. 



§ Ibid. vol. xli. pp. 73-157, pis. 8-10. 



II " Closure of the Oyclostomatous Bryozoa," Journ. Linn. Soc, Zool. 

 vol. xvii. (.1884) pp. 400-404, pi. 17. 



