DEVELOPMENT OF SPONGES. 45 
pears which may be seen through the body-walls. At this 
time the germ consists of two layers, the inner layer of cili- 
ated cells (endoderm) forming a closed sac, enveloped in the 
spiculiferous layer. Such are the observations of Metschni- 
koff on the development of Sycon. According to the ob- 
servations of Barrois, the larva or gastrula fixes itself by what 
are destined to be the ectodermal cells, and which are the 
round non-ciliated cells forming the posterior end (Fig. 31, 
C) of the free-swimming blastula. About this time the 
mesoderm separates from the endoderm, either before or 
just after the gastrula becomes stationary, according to the 
group to which it belongs. 
When the young sponge becomes stationary it does not 
differ from the gastrula, except that it becomes more or less 
Fig. 81.—Development of a sponge (Sycon ciliatum).—After Metschnikoff . 
irregular in form. Then appear the food or digestive cavi- 
ties in the endoderm, in Sycandra becoming radiating tubes 
lined with ciliated, collared, monad-like cells ; or in Leucon. 
and Halichondria, and their allies, forming scattered pock- 
‘ets, called ‘‘ ampullaceous sacs.’’ In most sponges (except 
some calcareous species) there is no general body-cavity in 
‘the gastrula, nor in the young after the larva becomes sta- 
tionary, according to Barrois. After the formation of the 
ampullaceous sacs the pores open through the mesoderm 
and connect the sacs and ciliated channels, as the case may 
