TYPE ANNELIDA. 243 



much less closely related than they are here supposed to be. The Eehiurese 

 are still held to have Annelidan affinities, whUe the Sipunculacea are as- 

 signed to the next type to be described. This tendency has its origin in 

 the attachment of too great importance to the metamerism which is indi- 

 cated in the Echiurid trochophore but lacking in the Sipunculid larva. 

 There seems no good ground for supposing Lhat its absence in the latter 

 group may not be suJflciently explained by the assumption that it repre- 

 sents the final stage of the reduction of metamerism of which the transient 

 segmentation of the Echiurid is a stage. In their anatomical character- 

 istics the adult forms of the two groups are too much alike to be assigned 

 to different types and the similarities of detail too numerous to warrant the 

 belief that they have been independently acquired. It seems much more 

 probable that both orders have descended from segmented ancestors— the 

 degeneration, if degeneration it can be called, having been carried to a 

 greater extent in the Sipunculacea than in the EchiuresB, and having in 

 consequence been thrown back upon the larval stages and so obscuring the 

 developmental evidences of the phylogeny. 



A connecting link between the Echiureee and the Polychaeta has been 

 traced by some authors in the genus Sternaspis, at one time associated with 

 the Gephyrea but now universally assigned to the Polychaeta. In this 

 genus the metamerization, though to a certain extent reduced, is still pro- 

 nounced, 8. arouata consisting of from twenty to twenty-two metameres, 

 of which the anterior seven, together with the head-lobes, may be invagi- 

 nated — the introvert of the Sipunculacea being thus recalled. On the ven- 

 tral surface near the posterior extremity of the body are two shield-like 

 plates armed with setse, and at the posterior extremity, as in Priapulus, 

 are a number of filamentous appendages which are regarded as branchiae. 

 Setae are present on all the metameres except the fifth, sixth, and seventh ; 

 those of the eighth to the sixteenth metameres being, however, concealed 

 beneath the hypodermis. The digestive tract is somewhat convoluted, but 

 opens terminally ; the ventral nerve-cord shows traces of ganglionic swell- 

 ings, and at the posterior end of the body possesses a marked enlargement ; 

 and only two nephridia are present. The musculature and the vascular 

 system resemble those of the Polychaeta rather than those of the Gephyrea, 

 while the reproductive organs are peculiar in possessing special ducts, 

 which, it has been held, show no indications of being modified nephridia. 



In many respects, accordingly, Sternaspis does hold a position interme- 

 diate between the Echiureae and the Polychaeta, and it seems not improb- 

 able that it may represent an offshoot from near the base of the line along 

 which the Gephyrea have been differentiated. Whether this be the case or 

 not, it is exceedingly probable that the Gephyrea have been derived from 

 the Polychaeta, the Echiureae preserving more numerous traces of their an- 

 cestry than do the Sipunculaceae. 



