1900.] POLYNOID FBOM NEW ZEALAND. 977 



narrow transverse grooves, which lead inwards to a median dorsal 

 groove or channel lying between the ridges of the right and left 

 sides. This channel is incompletely divided into two, longitudinally, 

 by a series of firm, truncated tubercles, which posteriorly form a 

 single median row, but anteriorly a double row, enclosing a spindle- 

 shaped raised area, the channel passing forwards outside the 

 tubercles (Plate LX. fig. 3). The result is that the channel bi- 

 furcates anteriorly ; but on the second segment there are again a 

 couple of median tubercles (Plate LX. fig. 3 & Plate LXI. fig. 4), 

 and the channel is thus carried forwards right to the base of 

 the prostomium (Plate LXI. fig. 4). Posteriorly, the tubercles 

 cease on segment xix, while the channel continues backwards 

 to segment xxii. The tubercles serve to support the mesial 

 moieties of the elytra, thus leaving a clear subelytral channel, 

 which, from its function, may be termed the dorsal " respira- 

 tory channel." Towards the posterior end the channel becomes 

 deeper, and is closed by the transverse union of the parapodial 

 ridges of the last elytriferous segment. The channel thus ends 

 just under the aperture mentioned above, formed between the last 

 pair of elytra. 



Looking more closely at the parapodial ridges, a number of small 

 processes are observable, at their outer ends and especially towards 

 the margins of the transverse canals. These processes may be as 

 many as twenty in number on a cirriferous segment, rather fewer 

 on an elytriferous segment (Plate LXI. figs. 7 & 8). They are little 

 finger-shaped evaginations of the body -wall, rarely branched, and 

 evidently serve as gills \ Unfortunately there were no living 

 specimens obtainable, so that we could not verify the following 

 inferences, but the appearances seem to justify them : — currents of 

 water are brought by ciliary action 2 into the transverse canals ; 

 the water is filtered by the bunches of notopodial hairs, and passes 

 thence over the branchise, oxygenating the blood ; the water then 

 passes into the dorsal median canal and finds its way out by the 

 posterior aperture. 



Haswell, in his Monograph of Australian Aphroditea [2], 

 mentions that both "Williams and Quatrefages record a respiratory 

 current in Aphrodita and Hermione, caused by the rhythmical 

 movements of the elytra under the felty coating of the back. But 

 he adds : — " In species in which the felt-like dorsal covering does 

 not exist, this function would appear to be in abeyance, and in 

 Polynoe and allied genera, so far as I have observed, the elytra 

 remain perfectly motionless while the animal as a whole is at rest." 



It is suggested that, in L. giyanteus, while probably no actual 

 movement of the elytra is necessary to produce the stream, yet, by 

 means of the "respiratory channel " and the dorsal tubercles 



1 Compare Buchanan's figure 4 in illustration of Eupolyodontus comishii, in 

 Q. J. Micr. Sc. vol. xxx\ . 



2 Cilia have been observed by several naturalists on the sides of the para- 

 podia in Polynoids ; but Mr. Thomson did not examine the histological 

 structure of the body-wall in the present annelid. — W. B. B. 



