BY MOLECULAR COALESCENCE. 89 
more numerous; but, as might have been expected, they 
have not the feathery appendage. Notwithstanding, their 
ends are free and prominent. In the tubes represented in 
the plate alluded to no projecting ends are shown, these 
having been most probably ground off in making the pre- 
paration, but, in other respects, they perfectly agree with 
those I am describing, but in no respects with the flexuous 
Imes before described. I have thin sections of crab’s 
claw which had not been boiled, showing these tubes, but 
they are troublesome to make. They can be most easily 
shown in the very young crab, but the best way to see 
them is in the decalcified shell. Now, I can easily ima- 
gine that tubes thus constructed, with one of their ex- 
tremities expanded, and thus rendered favorable for the 
operation of extensive endosmose or exosmose, and with 
the other extremity close to the part of the shell where 
the globular carbonate of lime is chiefly found, would 
serve admirably to convey the water in contact with the 
surface of the animal, containing salts of lime, to the 
membrane lining the shell, where a quantity of sub- 
carbonate of soda is always present—(this circumstance 
will be treated of more fully hereafter)—and thus to bring 
into operation all the conditions required to form the 
globular carbonate which occurs in this situation. But, 
on the contrary, in the ivory of the tooth, circumstanced 
altogether different to the claw of the crab, such tubes 
could be of no use, especially as they would be covered 
with a layer of enamel, and in some animals with a layer 
also of cementum. Now, as it is extremely probable that, 
with organs so decidedly tubular, there would not exist 
another set of tubes to perform the same function. An 
additional reason is here afforded in favour of the opinion 
of the structure of these alternating dark and light lines, 
