266 Proceedings of Royal Society of Edinhurgli. [sess. 
significance of this arrangement in Notarclius is not far to seek, as 
the animal is often observed, especially when irritated, to inflate 
this large parapodial bag with water, which is then forcibly ejected, 
the impetus sending the animal backwards in the opposite direction. 
It will be remembered that in another group of the Mollusca (the 
Cephalopods) exactly the same physiological arrangement is brought 
about by the development, however, of the mantle, not of the para- 
podia, as here. Probably a similar function is subserved by the 
arrangement in Dolabella, though no direct observations on this 
point have been recorded. 
Mantle : In Notarclius the parapodia so completely envelop the 
visceral mass as to render special protection for the gills superfluous, 
and hence the mantle has all but disappeared. In this form, how- 
ever, the parapodial opening is immediately over the gills, extending 
over their whole length, and we And the mantle well developed 
and completely covering the gills (fig. 2, m). It might be supposed 
that such an arrangement would be highly unsuitable for the access 
of water. This is secured in Notarclius by the gills being almost 
entirely uncovered by a mantle, and floating freely in the parapodial 
cavity. By two special arrangements in Dolabella, however, a 
definite current of water is directed over the gills. Pirst, the front 
of the mantle is closely attached to the left parapodium, so as to 
divert the current of water entirely to the right side, on which the 
gills lie (fig. 3, a). Secondly, the posterior rim of the mantle, after 
being folded up to form an exhalent sipho (fig. 3, s), is continued 
backwards round the anus, and downwards, and then forward over 
the floor of the mantle cavity to the genital opening (fig. 3, / and 
fig. 2, /), thus forming a ridge, the effect of which would be to 
guide the current of water over the gills and up to the exhalent 
sipho. This arrangement is of such importance, suggesting as it 
does a phylogenetic connection between the mantle and the genital 
furrow, that it requires an accurate examination. Fig. 3, m, is 
intended to show the mantle and its continuation into this fold 
and ultimately into the genital furrow (gf.). It will be noted that 
the mantle is not uninterruptedly continuous with this fold, but 
that there is a small gap of about one mm. in length. This gap 
(g.) may, however, be accounted for by the fact that the button or 
cushion-like arrangement previously mentioned occurs at this point. 
