102 
THE PEAEL OYSTER 
the animal before commencing a new growth of its shell " 
(Tryon's Structural and Systematic Conchology). Seeing 
that they do not follow the common law in the secretion of 
shells — the law that what is no longer required is absorbed 
for utilisation in the new formation — but are retained in 
great numbers and well ordered regularity, one is led to 
conclude that it is for a purpose that they are retained, and 
to conjecture what that purpose may be. Obviously arma- 
ture cannot be the object, for that could be better served 
by making the spines round and sharp and needle-liie, as 
in Ct/therea (Dione) lupanaria, Desh. Looking again more 
closely at a shell fresh from the sea, we find that there are 
not only projecting spatulated spines in about twelve marked 
radiations, but that between them are shorter spines, and 
after them they fine off to lesser radiations in the direction 
of the two ears, and along the base of all runs a thin folia- 
tion projecting slightly from the shell, so that at every new 
formation of shell, and these are at very close intervals, there 
is a thin knife-like ledge running all round the former 
contour of the shell and projecting slightly with further 
projections from it in short and long spines. 
Here we may recall what Dr. Kelaart writes : — 
" I have thi'ough the microscope ascertained the kind of food 
" pearl oysters live on. This consists of minute algoe or weeds, 
" animalcules and shells, called foraminifera. Diatoms also, 
"those minute vegetable forms which can scarcely be detected 
" with the naked eye, are found growing on the external surface 
" of the shell, where a host of infusorial and microscopical 
" objects likewise find a pasturage, so that the oyster maj- be 
" said to carry on its back the food on which it fives." 
For my own satisfaction, I have confirmed these remarks 
by personal observation ; and it struck me that this was 
the primary purpose which the formation of the outside of 
the shell of the young oyster was intended to serve. The 
numerous projecting spines presented a siu'face admu-ably 
