32 INTRODUCTION". 



The subdivisions of these groups accepted by D'Orbigny have 

 been recognized by nearly every subsequent worker to be artificial 

 and useless, and the amount of truth the classification contains 

 has therefore not been recognized. 



Busk followed in 1859 with a classification, of which the only 

 part that is now accepted is the separation of the Crisiidae from the 

 rest as the Articulata. Srnitt in 1865 accepted this arrangement,, 

 and two of D'Orbigny's divisions, for which he used the same names 

 but in a Latin form, viz. Tubulinea and Fasciculinea ; as he dealt 

 only with recent species, he had not to consider the forms included 

 in the division Foramines. Hincks in 1880 and MacGillivray in 

 1887, also each dealt only with living species, which are so few in 

 number that family divisions are sufficient in their classification ; 

 Hincks therefore referred the British species to four families, and 

 MacGillivray those of Victoria to five. Pergens and Meunier, 

 who in the same year described the Cretaceous Bryozoa of Faroe, 

 also used only family divisions, ten in number. Marsson in 1887 

 divided the Cyclostomata into two groups, the Solenoporina and the 

 Metopoporina, including in the latter only Melicertites and its 

 allies. Pergens in 1890 accepted the same division, though he 

 abandoned the name Metopoporina; and he made many great 

 changes in the families. 



3. The Classification proposed. Without attempting a formal 

 revision of the classification of the whole order of the Cyclostomata, 

 which to be of value must be based upon, and prove applicable to, 

 the Cretaceous fauna, it will be useful to show the grouping into 

 which the Jurassic representatives of the order may be arranged. 

 The Jurassic species are few in comparison with the Cretaceous, 

 but they offer the great advantage of showing the commencement 

 of the lines of development, which by the succeeding period had 

 given rise to extraordinarily varied forms. The lines of evolution 

 can therefore be recognized unobscured by the great secondary 

 variations of the Cretaceous fauna. 



The Articulata section of Cyclostomata is not known to be repre- 

 sented in the Jurassic, and thus this group may be dismissed. 



The Jurassic Cyclostomata fall fairly readily into two groups. 

 In the first, all the zocecia are functional, open, and tubular; in 

 the second, normal zocecia occur surrounded by aborted, closed 

 zocecia (or dactylethrse), which generally form a large proportion 

 of the zoarium. In the Cretaceous period there is a third type, in 



