121 
new brood of young would drift over from the original parent 
colony and settle among the young, already partially grown, attaching 
themselves to their cirri or their pinnules, wherever they could 
find an available place. It is conceivable that the young of the 
first breeding period would not be so large that the little penta- 
crinoids of the second breeding period could not, by their long 
stems, reach up above the plane of their arms, and thus obtain 
nourishment in spite of them, and complete their metamorphosis: 
on breaking off from their stems, however, they could not settle 
down between the larger young, but would have to swim about 
until they could find a vacant spot for themselves, or perish. By 
this process, all the vacant places available would be taken very 
quickly, and the new crinoid mass after the space of a very few 
breeding periods would become as dense and compact as the old 
parent mass, and would then be incapable of receiving further 
additions to its numbers. Hence, the mass would be entirely com- 
posed of individuals the difference of whose ages would not be 
more than two or three breeding periods at the most. 
It is possibly in this connection that we get an explanation 
of the ten-armed stage of the multibrachiate comatulid. These 
ten-armed young are, in all cases where they have been observed, 
very active, and are continually swimming about; thus they are 
enabled to pick out for themselves suitable localities before they 
transform to the adult multibrachiate condition, after which swimming 
is impossible, and the animal is fixed for life. 
All the extremely small specimens of species of the genera 
Capillaster and Nemaster which I have been able to examine are 
multibrachiate, even when apparently freshly broken from the larval 
stem. This would, theoretically, greatly hinder their swimming, 
if not stop it entirely, through the resultant irregular number of 
the arms on different rays; and it may be that this is sufficient 
to explain the fact that the individuals of species of these genera 
åre always scattered, and therefore never so common in any given 
