208 



evaporated, the membrane of the entire network of the 

 branchial laminse broken by lesions and contractions ; and 

 their fig. 3. in the plate has much the aspect of such ruptures. 

 " I cast this idea on the waters," as Southey did " his little 

 book," and it may have as much value as it deserves. I had 

 scarcely written these lines when I found that my conjecture 

 might be right. Having opened in a gill-plate an inter- 

 branchial tube that retained the injected mercury, I cleared 

 it of the mineral, and being dry it was placed in water to 

 recover pliability, for fixing on a tablet, on which it was care- 

 fully spread without stretching ; I found that in the central 

 portion of the membrane of the plate almost every parallelo- 

 gram was ruptured, which under the microscope showed no 

 previous solutions of continuity, and each fissure proved a 

 fac-simile of those delineated in the ' Annals/ vol. viii. N. S. 

 pi. 15. fig. 3. 



The area of the portion of the gill-plate examined contained 

 about 2000 parallelograms in rows, and by its size caused the 

 sphere of contractibility to centre in the middle, whilst towards 

 the margins, a less resistance and greater elasticity prevailing, 

 many of the rows of network preserved their integrity. I 

 then prepared another portion of ten transverse and as many 

 longitudinal rows ; in this diminished area not a mesh was 

 ruptured, and the membrane of the blood-vessels remained 

 perfect. It appears then, that the moistening of the gill- 

 plate with fresh water and of course with alcohol a much 

 greater effect is produced may have caused all the fissures in 

 Messrs. Alder and Hancock's specimens, thus fully accounting 

 for the singularly different results of our respective injections 

 of the anal siphon. 



If I am right in these points, the question of in- and ex- 

 currents by cilia and separate siphons is disposed of. The 

 data of these gentlemen to show a communication between 

 the anal and branchial vaults through the membrane of the 

 network of the gill-laminse not being tenable, of course their 

 theory falls to the ground, on the principle of <f sublata causa, 

 tollitur effectus;" consequently mine, as published in the 

 'Annals/ 1850, has not yet been proved incorrect. 



