NEBALIOPSIS 211 



row setae. It extends backwards and attaches to the body wall at the base of the boss. 

 I call it the flexor muscle. Its contraction must tend to lift the tip of the limb towards 

 the body wall and, at the same time, it must swing the tip round towards the middle line. 



The recovery of the limb back to the position shown in Figs, i and 2, must be mainly 

 due to the blood-pressure, but it is probably aided by the other muscle in the limb. This, 

 the extensor, is markedly smaller than the flexor. It is attached to the ventral surface 

 of the limb just below the anterior attachment of the flexor, and runs ventro-posteriorly 

 to a point of attachment near the tip of the boss. From Fig. 5 C it will be seen that it 

 also runs postero-laterally. Its contraction must therefore tend to swing the tip of the 

 limb outwards and downwards. 



Thus the typical trunk limb oscillates in metachronial rhythm, moving up and down 

 towards the body wall and against the limb in front, the tips of each pair of limbs 

 approaching each other as they move upwards. 



The result of this movement is, I believe, a general drift of water away from the limbs. 

 As each limb moves upwards it throws out the water between it and the limb in front. 

 The tip of the limb will be moving fastest, the base not at all. It will touch the limb in 

 front and so occlude completely the inter-limb space in this region. Thus the water 

 towards the tip of the limb will be forced outwards. It will pass out mostly between 

 the ends of the two limbs and the epipodite which stretches between them laterally. 

 The forward drift will obviously be more pronounced anteriorly where the carapace 

 covers the limbs laterally. 



A simpler way of considering the limb movement is to look on each limb as an oar 

 used as a scull over the stern of a boat. The oar produces, or tends to produce, a move- 

 ment of water away from the boat along the length of the oar. The limb, I believe, acts 

 in a similar manner. 



To replace the water thrown forwards, water will drift towards the body from all 

 directions in the hinder part of the trunk. Anteriorly the trunk is covered by the carapace 

 so that water cannot pass in laterally and, in addition, the momentum of the moving 

 mass of water passing forwards over the limbs will suck in the surrounding water, but 

 this will be drawn into the stream and not into the inter-limb spaces. 



Now this is totally difli'erent from the conditions in Nebalia. Here, all the water 

 which passes into the inter-limb spaces, is sucked from the mid-ventral chamber 

 through the filter wall. In Nebaliopsis, while there are filter walls to this median space, 

 there are no valves on the outer sides of the limbs to cause such a suction. On the con- 

 trary, there is a large open space dorsal to the attachment of the epipodite which, as far 

 as I can see, cannot be closed however the limb oscillates, and this must be at least one 

 of the passages for water entering the inter-limb space. There is also a gap distal to the 

 epipodite and there is nothing to hinder water entering here as well. 



I have summarized what I believe to be the main water currents around a feeding 

 Nebaliopsis in Fig. 5 B. Now the passage forwards of a definite stream over the surface 

 of the limbs must mean that there will be a drift in the same direction between the 

 limbs in the mid- ventral space. I do not mean that there will be a marked anteriorly 



