SPIRAL VALVE I\ THE GENUS RATA. 53 



the axis, and become a straight line, thus giving a sort of columella to the spiral, and 

 effectually preventing the passage of food in any but a spiral direction. 



It is evident that, retaining the same general character as that described, a much 

 simpler and less efficacious form of valve would be produced if the free edge, still re- 

 maining parallel to the attached edge, formed a more open spiral. This condition of 

 things is well shown in the last four turns of the valve in fig. 1, PI. X. (the anterior 

 turns have undergone a modification presently to be described), the width of which is 

 not greater than half the semidiameter of the intestine ; and, in consequence, a wide 

 passage is left down the axis of the latter, along which its contents could pass directly, 

 or without taking a spiral direction at all. It is a valve of this sort which Perrault 

 figures in his " Eenard marin " {Alopecias vulpes), and which he aptly compares to " un 

 escalier tournant sans noyau." 



But in this valve (fig. 1) — the simplest I have met with, except one unreliable dried 

 specimen — the anterior turns do not retain the simple character described. They have, 

 in fact, undergone a notable increase in width, the second turn becoming as wide as 

 the intestine itself, and, being now prevented from taking a transverse direction, have 

 become more or less folded and deflected, the direction of the deflection being, on the 

 whole, a backward one. Thus the inner edge of any given turn of the valve, instead 

 of being in the same plane as its attached edge, comes to be in the same plane as the 

 attached edge of the turn next behind, or even as that of the next but one. In fig. 1, 

 for instance, the free edge of the fourth turn is on the same level as the attached edge 

 of the sixth. 



This state of things is carried to an extreme in the cases shown in fig. 8, PI. X., and 

 fig. 1, PI. XI. In these the width of the second turn is so great that its free edges come 

 to be on about the same level as the attached edge of the seventh ; and the width of the 

 valve gradually diminishing from the third to the seventh turns, the free edges of these 

 also are brought approximately into the same plane. This is well shown in the cross 

 section, fig. 9, PI. X. The successive turns therefore form, as it were, a " nest " of 

 imperfect truncated cones, placed, with their apices directed backwards, one inside the 

 other, and gradually diminishing in height from the innermost to the outermost. 

 Moreover the successive " cones " adhere so closely to one another that there is prac- 

 tically produced a central conical chamber, with a peripheral spiral cavity wound round 

 it. This condition is, in fact, actually produced if such an intestine be dried ; for then 

 the successive turns become completely fused together wherever they are in contact. 

 In the diagram of this type of valve the free edge is seen to form a very curious and 

 complicated figure (fig. 2, PI. XI.). 



In fig. 1, PI. X., it will be seen that the first two turns, although on the whole 

 deflected backwards, turn at first forwards or towards the pylorus. If this anterior 

 deflection were persevered in, and took place in every turn, we should have such a 



