iS6 STRUCTURE AND LIFE-HISTORY OF A SPONGE. 
appear to have travelled along very much the same road as the 
sponges, but now they part company, the ccelenterata originated 
in the growth of the ciliated cells over the cilialess cells, so that 
the latter formed the digestive lining inside the resulting gastrula^ 
as we have agreed to call the sac formed by the invagination of 
the preceding form or blastula \ in the sponges, on the contrary, 
as we have seen, the ciliated cells withdraw into the cilialess layer, 
which thus becomes a protecting, instead of a digestive, layer. 
But now it is worth while recollecting that though this is the 
normal process with the sponge, yet that the opposite one is 
frequently passed through as a transitory stage, preliminary to it, 
and thus we may conjecture that the larva which becomes a sponge 
now, by invagination of the ciliated layer, is a descendent of a 
form which used to become a coral by the invagination of the 
other layer, that is, that a form on the way to become a coelenterate, 
took the wrong turn for once, and so ended in a cul-de-sac, and 
became a sponge. Thus the abnormal kind of invagination in 
Sycandra may be an instance of what is termed ^ reversion to the 
ancestral type*] on the other hand it may simply indicate the 
balancing play of forces on the young organism, so that it 
looks as if it could not make up its mind, and was undecided as 
to whether to turn the flagellated layer inside and become a 
sponge, or outside, and become a coelenterate. Between these 
alternative possibilities we cannot decide, the day has not yet come 
when the development of an animal may be represented by a 
mathematical formula, and that wonderful man, the mathematical 
physicist, has not yet taken the matter in hand, 
To resume our history. The young gastrula, which we believe 
to have led an independent life, a free swimmer in the sea, whose 
parent lived and died a gastrula, took another step in development, 
by becoming sessile, and here again its contrary disposition was 
displayed, for instead of settling base downwards, as the ccelen- 
terate gastrula did, when it similarly resigned a free existence^ it, 
