14 INTRODUCTION". 



5. THE VALUE OF GENERIC DIVISIONS IN THE 

 CYCLOSTOMATA. 



1. Simplicity of Structure. Examination of this list of the 

 structures of tubicolar Bryozoa, shows that the skeleton is never 

 a very complex one ; and as many of the above structures are 

 confined to the Trepostomata, or to exceptional genera of Cyclo- 

 stomata, the majority of the members of the latter order have 

 characters, which are both simple and variable. This renders the 

 accurate diagnosis both of genera and species difficult, if not 

 impossible. Among the Cheilostomata there is, at least, an ap- 

 proximation to an agreement as to the taxonomic value of the 

 different skeletal elements. Thus, the suborders are founded on 

 the thyrion or front wall; the families upon the nature of the 

 openings to the exterior, and upon the presence of ooecia or 

 gonoecia ; and the genera on the characters of the zocecia, and 

 the arrangement of the avicularia and appendages. 



But when we turn from those Bryozoa in which the zorecia are 

 box-shaped, and there are spines, avicularia, and vibracularia, to 

 those in which we have to rely solely upon the modifications and 

 grouping of simple tubes, agreement as to the range of genera 

 ceases. 



2. The Stomatopora-Diastopora Series. As an example we may 

 take the series of forms grouped round Berenicea. Haime, in 

 his monograph on the Jurassic Bryozoa, accepted four genera, 

 Stomatopora, Proboscina, Berenicea, and Diastopora. Hincks, how- 

 ever, with a similar series of variations in recent specimens, 

 has merged these into the two genera Stomatopora and Diasto- 

 pora. Ulrich, again, has made a genus, Hitoclema, for a Silurian 

 Bryozoon, and, in reply to criticisms, maintains that it is quite 

 different in structure from Entalophora ; nevertheless, Waters and 

 Vine claimed it as a normal species of that genus, and the former 

 even placed it in the species Entalopliora verticillata, Goldf. Haime 

 included in the genus Heteropora specimens which D'Orbigny had 

 distributed among the following fourteen genera : Cava, Ceriocava, 

 Ceriopora, Crescis, Heteropora, Multicrescis, Multinodicrescis, Nodi- 

 cava, Nodicrescis, Polytrema, Reptomulticava, Reptomulticrescis, 

 Reptonodicava, and Reptonodicrescis. Haime even included within 



