630 TRANSACTIONS LIVERPOOL BIOLOGICAL SOCIETY. 
of the appearance of the neuropodial crotchets of the 
middle region, both neuropodial and notopodial crotchets 
appear in the posterior part. 
The posterior region is terminated by a conical caudal 
appendage, situated dorsally, which seems to have been 
overlooked by Fewkes but is described by Claparéde (loc. 
cit. p.. 76;, Pl. X.,, fig:, 14). The last “ seamen: sama 
Fewkes’ figs. 10, 11, 18, probably represents this 
appendage, being devoid of parapodia. None of the Port 
Kirin specimens show the broad terminal “segment” of 
Fewkes’ figs. 12, 12a, perhaps because they are not old 
enough. The anus opens ventrally on the last segment. 
In Fewkes’ larvae three .crotchets “arise from the 
dorsal region of the parapodium. In addition to these 
appendages the posterior body segments also bear on a 
ventral elevation smooth spines similar to those on the 
anterior and middle regions *’ (loc. cit. p. 171). In balsam- 
preparations of the Port Erin larvae I have been unable 
to find any capillary setae on either the middle or 
posterior body-regions. Fewkes also found, in the case of 
the provisional setae, that those of the anterior region 
(except the first segment) were completely lost at a com- 
paratively early stage, and in this his observations are in 
agreement with those of Claparede. Specimens from 
Port Erin, however, show these setae even after the loss 
of the larval tentacles. This is probably another example 
of the great variability in the time at which such 
phenomena occur in Spioniform larvae (see above, pp. 
609 and 610), a fact noted by McIntosh in this particular 
connection (1894: p. 69). 
A ‘‘Magelona-like” larva has been described from 
the Cape Verde Islands by Hacker (1898: pp. 20-23; 
Pl. II., fig. 19). The larval tentacles of this form each 
bear a prominent swelling about one-third of the way 
