ON TEILOBTTES. 115 



(•").) As the Trilobiles have no known Post-Pakuozoic representatives, and are prac- 

 lioally confined to strata representing the earlier and middle periods of the Pakeozoic Age, 

 broad dislimtions, surh as those which separate the Pakeozoic Criuoids and Eehinida from 

 Mesozoic and higher forms oi' these orders, cannot obviously be looked for in any 

 classification. Added to this disadvantage, there are other drawbacks which beset at 

 present, and render dilFieult, the systematic grouping of the Trilohites. lirielly stated, 

 these comprise, first of all, the very lax manner in which fragmentary examples have been 

 referred to known types, or placed, as new genera, in families with which they have but 

 few points of resemblance. Definitions, applicable enough in the first instancfe, become 

 thus, after a time, of almost impossible application — the original definition being so 

 extended as to iiicludt> forms of very dissimilar structure. Mathematically rigid defini- 

 tions in the grouping of organic forms cannot, of course, be strictly enforced without, at 

 least, greatly multiplying our classification groups ; but better an undue multiplication 

 of orders and families, than this license of indefiniteness that now so habitually prevails. 

 Another source of difficulty lies in the remarkable dissimilarity of aspect which exists in 

 tho .•■)se of many species commonly referred to one and the same genus, as seen in 

 Asaphus, Proetus, Cheirurus and other forms. This has led to the creation of numerous 

 sub-genera, by which, how^ever, the diihculty is in no way lessened — family definitions 

 still remaining vague and diffuse. 



(G.) In the classification shown, as regards its leading features, in the annexed 

 synopsis, the Trilobites, considered as an order of Crustacea, are arranged under four sub- 

 orders, or primary groups, founded on general structure and configuration, with subordi- 

 nate sections based, as far as possible, on some striking or typical character. This latter, 

 however, is to be taken in connection with the general characters of the sub-order to 

 which the section may refer. The names attached to these sections refer necessarily to 

 single characters, and they are thus to be regarded as names only, not as defmit.'ons. It 

 is thought, however, that they may serve to bring rapidly before the mind's eye the more 

 salient or special features of the types to which they refer. In the collocation of the 

 families, an attempt is made to place them in as connected a sequence as possible, so that 

 each family shall present relations to the family which tbllov>'s it, whether placed in 

 the same section or under contiguous sections ; but in a linear system of arrang(>ment, 

 and especially where the objects to be classified are not very numerous, it is not possible 

 to avoid occasional breaks in the continuity of the series. 



In offering this new classification, I am not sanguine enough to expect that it will 

 meet with general acceptance. In its subdivision of genera, and the recognition of these 

 subdivisions in one or two cases as the types of new families, and in its disregard of 

 stratlgraphical considerations, it will necessarily rullle existing prejudices. Hut it may 

 tend to bring under more immediate notice the indefinite and hence unsatisfactory 

 character of our commonly received subdivisions, and so lead eventually to a more 

 successful groxiping of these ancient forms. 



