132 DR. L. A. BOllRADAILE ON THE 
along tbem. They widen from above downward, and it would at first seem 
as if the current must flow in that direction. But (1) as each channel widens 
it grows sliallower, so that its capacity is no greater, but rather less ; (2) its 
lower end is probablj"^ closed by the edge of the branchiostegite, and in any 
case leads to a part of the chamber that does not directlj' communicate with 
the exhalent passage; and (3) the direction of the leaflets is such as at the 
lower end to cast the water into the channel in an upward direction. It is 
therefore probable that the current Adas upwards. On the other hand, near 
the upper end the channel narrows so much that its capacity is greatly 
diminished, and here the water must be overflowing f lom it into the space 
over the gills, which is near the top of the collecting»-space. The grooves 
between the" convex surfaces of adjoining gills, which may be known as 
" interbranchial channels," form a similar system. 
All the ivater enters the exhalent passage from above and within — that is, 
from the collecting-space. This is due partly to the fact that, as has already 
been shown, access to the passage from behind is barred, and partly to the 
fact that the stroke of the scaphognathite is made from below upwards, against' 
the roof ; and, as has also been shown, it has the important effect of causing 
the water to flow through instead of over the gills. 
The loorldng of the scaphognathite has already been described. The 
effective stroke is the upward one, but the downstroke must act upon the 
water which has entered the lower part of the passage during the upstroke. 
Probably the bulk of this is driven into the collecting-space and a small 
portion sent forward to the exhalent orifice. 
The stream that issues from each branchial chamber is directed, as has been 
said, obliquely across the e[)istome lo the opposite side of the front, and often 
it does in fact take that course. But normally it meels in the middle line the 
current of the other chamber, and the two deflect one another so that they flow 
■ forwards under ths antennules. A further modification of the direction of 
the current is brought about by the activity of the flagella of the maxillipeds. 
Flicking to and fro extremely rapidly, these exert their force more on 
the outward stroke when they are drawn by their extensors in the direction 
of their concavity as a cilium moves than when they are returning by their 
own elasticity in the direction of their convex sides. They not only reinforce 
the current very notably, but also turn it outward, and in particular, I think, 
b)' means of the hairs with which they are fringed they draw away particles 
which might otherwise lodge upon the organs of special sense. 
'I he regulation of the gill-stream is brought about in two ways — by 
alterations in the size of the openings, and by changes in the beat of the 
scaphognathite and the exopodites of the maxillipeds. I have already shown 
how the flow through the anterior iuhalent opening is regulated by the third 
maxilliped, and how the legs can close the openings which lie behind them. 
The size and form of the exhalent opening must also have an important 
