94 SECTIONAL ADDRESSES. 



long spouts in front and behind. The evolutional changes in the gill- 

 cavity are somewhat complicated morphologically, but physiologically 

 can be summed up in a single word, sanitation. The various readjust- 

 ments amount to a series of experiments in the more efficient separation 

 between the respiratory and excretory arrangements, and finally result 

 in the substitution of a simple system which cannot go wrong for one so 

 intricately balanced that it will only work if its owner keeps perfectly 

 still. It is no accident that Zygobranchism is associated with a sedentary 

 rock-life, and that Azygobranchism is distinctive of the snails with 

 versatile and wandering habits. 



The true Limpets (Docoglossa) of course gained the same end by 

 different means, sacrificing first one, then both gills in the mantle-chamber, 

 and substituting for them an entirely new system of marginal folds outside 

 the primitive gill-chamber altogether. 



From this survey of the facts it seems to be a legitimate inference that 

 the reversal of the mantle-chamber did in fact introduce some serious 

 difficulties into the adult life of the first Gastropods. Retention of the 

 complete ancestral organisation was rendered impossible except by an 

 immediate modification of the mantle margin, and even this permitted 

 no deviation from a very restricted mode of life. ' Radiation ' into other 

 environments requiring greater activity was inhibited by the delicacy of 

 the respiratory adjustments, consequent on the partial blocking of the 

 branchio-cloacal aperture. Had the torsion taken place by instalments in 

 successive generations, some of the modifications which were subsequently 

 introduced (with the Azygobranchia) would almost certainly have been 

 accomplished en route, and would not have been deferred until the rotation 

 was complete. The nature of the earliest post-torsional modifications 

 thus corroborates the more direct evidence that torsion was, so to say, 

 imposed upon the adult stage, and not primarily developed in its 

 interest. 



But the marginal slit has bearings on the general jjroblem which are 

 direct as well as corroborative, since it provides us with a test case of 

 the origin of a typical adult character. We know from Boutan's account 

 of the development of Fissurella that there is not a sign of the slit before 

 the sedentary stage is entered upon, and his figures show that an area of 

 shell is produced equal to that of the whole embryonic coil before the 

 marginal slit begins. This area is a mere trifle compared with the ultimate 

 size of the adult shell, but it is enough to show that the slit is a purely 

 adult character and arises at the outset of the adult life. 



Now the history of the slit is engraved upon the face of every Zygo- 

 branchiate shell, and the date of its commencement in the adult life is 

 to be got by following the ' slit-band ' to its source. In every Zygo- 

 branchiate living to-day the ' slit-band ' begins, like the hole of Fissurella, 

 near the apex of the shell in front of the larval coils. Moreover the 

 inscription on the shell is so distinct, and the shell so durable, that it can 

 be read on the shells of the earliest Cambrian and Silurian fossils. Here 

 also in every case, even in the primaeval Bellerophon, which retains perfect 

 bilateral symmetry in its nautiloid coil, the slit-seam runs up from the 

 margin of the shell nearly to the apex of the coil. There has accordingly 

 been no change from first to last in the period at which the slit develops. 



