342 MR. G. C. ROBSON OX THE RESPIRATORY 



on the other hand, says that it fits into a groove in tlie latter. 

 It is possible that, in the spirit specimens 1 ha,ve examined, the 

 ridge is conti'acted and drawn away from the roof of the pallial 

 cavit}". 



In the first place it is unfortunate that we can have no direct 

 evidence on this subject. In a living animal we are obviously 

 unable to see whether the pallial roof is in contact with the entire 

 length of the ridge. We have, therefore, to fall back upon in- 

 direct evidence. I'he following facts appear to indicate that the 

 separation is not complete. The ridge is sometimes quite high 

 enough anteriorly to touch the roof of the cavity ; but it almost 

 invariably decreases in height posterioily so that it is difficult to 

 see how complete insulation of the two chambers can be thereby 

 eflfected. Secondly, I have never found a groove deep enough to 

 hold the top of the ridge stead}?- against the j)i'essure of water. 

 Thirdly, I have frequently found the albumen gland so much 

 distended that it jn^esses down a portion of the ridge and keeps 

 the pallial roof a.way from it, thus opening a more complete 

 communication between the two cavities. Fii:ially, there can be 

 no actual need to prevent water from entering the so-called pul- 

 monary pallial cavity from the side of the gill, as it may be seen 

 frequently entering it from the left-hand side. It is, perhaps,, 

 from this latter fact that we ma}- derive an explanation of the 

 function of the ridge. Water apparently enters the pallial cavity 

 from the left-hand side, passes backwards, and then is passed 

 forwards over the gill to be finally discharged from the right-hand 

 side. It seems likely that the ridge functions in guiding the- 

 current of water, and its structure ( usually high anteriorly and 

 diminishing in size posteriorly) is in accordance with such a 

 function. The criticism will at once suggest itself that similar 

 arrangements for directing the cui'rent of water in branchial 

 respiration are not found in the large assemblage of other Proso- 

 branchs which employ branchial respiration. Difficult as this, 

 .argument may be, it appears to offer less obstacles to a solution 

 thaii the fact that a, similar ridge is found in Yivipara in which 

 no lung is found ! 



The function of the ridge in that genus has been discussed by 

 Sewell (9), who regards it as ' respiratory,' though he does not 

 make its precise r61e clear. The possibility that the ridge may 

 form part of a gutter down which the renal excretions pass to 

 J^he exterior should not be overlooked. Bouvier (2) suggests that 

 it acts in this capacity in AonpuUaria, as a result of the displace- 

 ment of the gill into the area of renal excretion. It should not 

 be overlooked, however, that the ridge is foiuid in Vlvijxtra in 

 which there is a, fully-formed ureter as well ! 



B. The next point of interest is whether during siphonal 

 respiration there is a direct coimection between the base of the 

 siphon on its inner side and the orifice of the lung. It must be 

 admitted in advance that the evidence to be adduced is mainly 



