284 



A REVIEW OV THE TRILOJilXES. 



the last family and the liarpesidce and olenidce there might appear, 

 at first sight, to be few connecting points ; but we have here the 

 same general tapering form of body, the gradually diminishing py- 

 gidium, the increasing segments, and the contracted glabella. The 

 genus cyphaspis (usually placed with the proeiidce) is undoubtedly 

 related to each of these three types. Finally, the olenidce and pa- 

 radoxidce have so much in common, that in general they are united 

 into a single family. The opposite character, however, of the gla- 

 bella (and of the buckler generally) should keep them distinct. If, 

 indeed, it could have been so contrived without breaking through 

 other relations, the paradoccidce would have been placed higher in 

 the series; for I think it will be found that paradoxides is related 

 to the genu3 PJmcops, much in the same way as Jiarprs or olenus 

 is related to calymene. A certain transition, at least, i3 presented 

 through the genus remopleurides, with its largely-developed buck- 

 ler and glabella, its eleven thoracic segments, and its dwarfed and 

 modified pygidium. 



Although, when viewed in the manner just pointed out, the above 

 arrangement indicates to a certain extent the relations existing be- 

 tween the families adjacent to one another in the series, it is yet in 

 other respects confessedly of an imperfect character. It is obvious, 

 however, that such must necessarily be the case, where it is at- 

 tempted to shew these natural transitions in a purely linear sys- 

 tem of arrangement. Thus, in the above method, the asaphidce, 

 required as a transition group, are placed in a ceutral position, un- 

 avoidably remote from their allied forms, the illanidce. But where 

 complicated relations exist, it is impossible for all to be met in a 

 satisfactory manner in any linear distribution of the kind. 



Four type-forms appear to hold a prominent place amongst the 

 trilobites, and indeed, when considered in all their modifications, 

 to constitute centres of classification, as it were, around which the 

 other types may be at least conveniently if not naturally grouped. 

 Admitting this, we obtain the distribution exhibited in the following 

 table : — 



; 



Illceniafs. 



Illoenid;e. 

 Brontida\ 



ASAPHIAXS. : CALTMENIAyS. 



Asaphidse. 



Trinuclida 

 Proetida?. 



Oalymenidse. 

 ELarpesidse. 



Olenida-. 



Phacopsiasts. 



Phacopsidte. 



Ceraurida . 

 Lichasida-. 

 Acidaspid®- 

 Paradoxida\ 



