288 DR J. STEPHENSON ON 
markedly towards their posterior limit ; anteriorly, before fading away completely, they 
become mere tubercles. They are in constant movement, in some degree reminding the 
observer of cilia; the resemblance is frequently increased by the regular propagation of 
the beat along the line of the gills, in a postero-anterior direction, like the propagation 
of a ciliary wave; but there is this difference, that the movement of the individual gills is 
a swinging movement from side to side, 7.e. in a direction at right angles to the direction 
of propagation of the wave, not, as in the case of ciliary movement, in the same direction 
as the wave. The structure of the gills will be described with that of the body-wall. 
In addition to the oscillations of the gills themselves, the whole posterior part of the 
body of the worm is continually performing undulatory movements. 
The body-wall (figs. 2-5) consists of the usual layers. The surface epithelium is 
high and columnar over the body generally, low over the gills (fig. 6). The subjacent 
circular muscular layer is thin, and extends unbroken throughout the body. The longi- — 
tudinal muscular layer, on the other hand, is thick, but is characterised by inequalities 
of distribution ; in the anterior part of the body (fig. 2) it is of the same thickness at 
all parts of the body-circumference, but in the gill region inequalities appear, while 
posteriorly (fig. 3) it is absent from parts of the body-wall. The position of the bands of 
fibres that persist varies: they may be lateral, one on each side; or one dorsal and one 
ventral; or there may be one, a ventro-lateral, band only. The position of the absent 
fibres is taken by an indefinite network with a few cells, the appearance being some- 
what that of a scaffolding or meshwork from which the fibres have dropped out. 
An inner circular layer of muscular fibres is well marked in the anterior part of the 
body (fig. 2). The muscular fibres of the septa are continuous with the fibres of this 
layer ; and special bundles of fibres, passing vertically between the sacs of the dorsal 
and ventral sete on the same side, may be considered as belonging to it (cf also” 
Inmnodrilus socialis, post). 
In the gill region the circular muscular layer of the body-wall is continued over the 
gills without splitting, while the longitudinal layer closes the base of the space within 
the gill, the fibres maintaining their straight course without being deflected. In the 
interior of the gill, within its muscular layer, are a number of branched cells whose 
processes anastomose (a, fig. 6); these may entirely fill the distal part of the gill-cavity. 
Proximally, near the base of the gill, there is always a space; cells and cell-processes 
may stretch across it, and other cells may form a definite, almost epithelial layer on the 
inner surface of the muscular coat (fig. 6). The two vessels of the gill run on opposite 
sides, within the muscular coat. 
Brpvarp, who describes a septum shutting off the cavity of the gill from the general 
body-cavity, nevertheless speaks of the gill-cavity as “evidently belonging to the 
celom.” J do not think this follows either from his description or mine; the gill- 
cavity appears to be a space between the muscular layers of the body-wall, whic 
undergo a separation consequent on the circular layer alone being continued over the 
& 
gill. Brpparp also states that the movements of the branchize are caused by muscular 
