NO. 10 CRUSTACEAN METAMORPHOSES — SNODGRASS 69 



tion of the polychaete parapodia (A). The nephridia (B, Nph) and 

 the primitive genital ducts open mesad of the leg bases suggestive of 

 their openings on the coxae in many of the arthropods. Though 

 modern Onychophora are terrestrial animals, there can be little doubt 

 that they had aquatic ancestral relatives represented by the Cambrian 

 Aysheaia of the Burgess Shale, and perhaps by the Pre-Cambrian 

 Xenusion described by Heymons (1928). 



The arthropod limbs are developed on the embryo from latero- 

 ventral budlike rudiments that lengthen and become segmented. We 

 may therefore suppose that from the ancestral onychophorans (fig. 

 I A) a form was evolved with longer legs (B), which later, with 

 sclerotization of the integument, became the jointed appendages of the 

 ancestral arthropods (C). It then required a long period of Pre- 

 Cambrian evolution to produce a trilobite on the one hand, and some 

 ancestral form of crustacean on the other. The differentiation between 

 the two groups, however, was first in the form of the body, not in 

 that of the appendages, as seen in the legs of a trilobite (fig. 25 A) 

 and those of Marella and Btirgessia (B, C). Though there is no valid 

 reason for regarding the primitive arthropod appendage as being a 

 biramous limb, the crustacean appendages later acquired their charac- 

 teristic biramous structure, which is usually lost in the ambulatory 

 limbs (F). 



Many carcinologists hold the view that the phyllopodial type of 

 limb is primitive, at least for the Crustacea, and this concept has been 

 well elucidated by Borradaile (1926a, 1926b). It is supposed that the 

 primitive crustacean appendage was a flat, unsclerotized lobe with a 

 fringe of hairs on the mesal border. Then the inner margin was 

 broken up by the development of a series of endites. Next, the limb 

 became more rigid by a sclerotization of the integument, but this 

 necessitated lines of flexibility that led to a system of jointing, and 

 naturally the joints were formed between the endites. Thus the 

 endites are explained as the precursors of the later developed limb 

 segments. Finally, with the departure from the phyllopodial form and 

 the suppression of the endites, some of the limbs became slender, seg- 

 mented, leglike appendages. In favor of this theory it may be noted 

 that in many of the branchiopod api^endages there are six endites on 

 the mesal margin and a free lobe at the apex (fig. 27 A, B). If all 

 the parts of such a limb became segments there would be seven seg- 

 ments in all, the terminal lobe becoming the dactylopodite, which gives 

 the usual number of limb segments in the Crustacea generally, though 

 Borradaile holds that the maximum number is nine, which would 

 include the doubtful "precoxa" of the trilobite. 



