A REVIEW OF THE TRILOP.ITES. 283 



staele to their assumption, because new forms are being constantly 

 brought to light ; and, by widening out the genera as here done, am- 

 ple space is left for the reception of these new comer3, and both gen- 

 eric and specific distinctions rendered far more rigorous and minute. 

 In a linear system of arrangement, like that necessarily employed, it 

 is extremely difficult, if not indeed impossible, to convey a just idea 

 of the relations of these families to one another. An attempt to 

 effect this has been made, however, in the following distribution, in 

 which, with certain unavoidable exceptions, each family will be seen 

 to form a natural transition from that which precedes to that which 

 follows it. The weakest point in the connexion, occurs perhaps be- 

 tween the second family and the third. 



lLL<ENIDiE. 



Beontid^e. 



LlCHASIDiE. 



acidaspidje. 

 Ceeatteid^e. 



PhaCOPSIDjE. 



Teinuclid.e. 

 a s aphid .e. 



Proetidjb, 



caltmenidj2. 



Harpesidje. 



Olenid^:. 



paeadoxid2e. 

 Here the stream of affinity flows from the asapMdce in an upward 

 and downward direction. Thus, through the genus stygina, the asa- 

 pMdce connect with the trinuclidce. These latter have certain affin- 

 ities with some of the pltacopsidce, and the asapMdce and phacopsida. 

 are still nearer related. The cerauridce and phacopsidce, again, have 

 the same number of body-rings, the same expanding character of the 

 glabella, the same facial sutures. "With acidaspidce and lichasidce, 

 the cerauridce have also much in common. Brontidce are but slight- 

 ly related to the family below them in the list, but in both the tail- 

 rachis is very small, and the pygidium itself of a peculiar character. 

 The illcenidie and the brontidce are closely related by the large 

 buckler and pygidium, the slightly-developed tail-rachis, and the 

 normal number of body-rings, with other characters to be pointed 

 out in the sequel. 



On the other hand, the proetidce form a transition group between 

 the asapMdce and the calt/menidce, whilst these families are also 

 more or less immediately related by the genus Iwmalonotus. Between 



