356 ME. H. M. BERNARD ON THE SYSTEMATIC [Aug. 1 895, 



case, while persisting as swimming-plates, are anteriorly adapted for 

 clinging to weeds, and in the latter case were early changed into 

 filamentous crawling-legs. 



In my former paper I suggested two possible genealogies, one in 

 which the trilobites appeared as offshoots specialized for a creeping 

 manner of life, from the main crustacean stem, and the other in 

 which Apus and the higher Crustacea are deduced from the trilo- 

 bites by a secondary loss of pleurae along the trunk-segments. The 

 persistence in Apus of cylindrical trunk-segments but loosely bound 

 together inclined me to pronounce strongly in favour of the former. 

 This choice is now amply justified. Not only in the form of the seg- 

 ments, but in their number and in the character of the appendages, 

 Apus stands far below the trilobites. Apus, in spite of the speciali- 

 zation of its mouth-parts, in the possession of traces of over sixty 

 cylindrical body-segments, and of its but slightly modified leaf-like 

 limbs, is clearly in the direct line of descent between the original 

 annelidan ancestor of the whole group and the higher Crustacea. 

 The trilobites, on the other hand, while retaining the homogeneity 

 of the head and trunk-limbs, and other primitive annelidan cha- 

 racters which have been lost in Apus, are yet specialized by the 

 great reduction in the number of body-segments, by the development 

 on every segment of large pleurae, and by the early modification of 

 the phyllopodan appendages. 



The trilobites, therefore (as exemplified by Triarthrus), in spite 

 of their extremely primitive mouth-formula, do not stand in the 

 direct line of descent of the Crustacea, but are lateral offshoots, 

 specialized for a creeping manner of life. 



Specialisation of the Limbs. — In the modification of the original 

 phyllopodan appendages it is the endopodites in both Apus and the 

 trilobites which are most modified. The exopodite, which I think 

 was the original notopodial cirrus, retains in both cases its phyllo- 

 podan specialization as a swimming-organ. 



Triarthrus possessed the gnathobases on the trunk-limbs, some- 

 what as we find in Apus. This, as I endeavoured to show in my 

 former paper, was rendered highly probable by an examination of 

 Walcott's results, 1 and by a comparison of his sections of Calymene 

 with sections of Apus. This point is now set at rest. The phyllo- 

 podan limbs of the larval segments, in developing into walking-legs, 

 retained throughout their whole length the gnathobases which, 

 round the mouth, become the jaws. Dr. Beecher has accordingly 

 given an emended diagrammatic section — emended, that is, from 

 Walcott's well-known figure in his classical paper just quoted. If 

 I might be allowed to suggest a further emendation, it would be in 

 the direction of making the line of insertion of the limb much 

 longer transversely. "Walcott's sagittal sections, compared with 

 sagittal sections of Apus, appear to me to leave no doubt that the 



1 ' The Trilobite : New and Old Evidence relating to its Organization,' BuLL 

 Mus. Comp. Zool. Harvard, vol. viii. 1880-81. 



