444 SCIENCE PROGRESS. 



epithelium. A stage showing the internal cavity of the 

 young sponge lined by columnar cells is compared with a 

 later stage in which flagellated chambers open into a gastral 

 cavity lined by flat epithelium. From the comparison it is 

 inferred that the flat epithelium of the gastral cavity owes 

 its origin to a flattening of the columnar " endoderm " cells. 

 Two other inferences, however, are just as possible, either 

 that the lining of flat epithelium originates by an exposure 

 towards the gastral cavity of the outer germinal layer when 

 the " endoderm " is aggregated into the chambers, or that the 

 columnar cells of the earlier stage are not all "endoderm ".^ 

 In any case Plakina is, like Halisarca, one of the forms 

 which requires renewed investigation, the facts of its de- 

 velopment as at present recorded standing in irreconcilable 

 contrast with the development of other forms. 



It must therefore be granted that, with all the respect 

 deservedly due to so great an authority, Schulze has not 

 brought forward absolutely convincing proof of the origin 

 of the internal flattened epithelium by modification of the 

 collared gastral layer. On the other hand such direct 

 embryological evidence as exists with reference to the 

 origin of the flat epithelium of the interior shows it to be 

 derived from the dermal layer, the inner cell mass of the 

 larva. Both Maas {1892 and 1893) and Delage (1892) are 

 united upon this point, and agreed that the ciliated layer 

 of the larva furnishes the chambers, and nothing else. To 

 establish what may be termed the continuity of the ciliated 

 cells is, as we shall see, the tendency of recent investigations 

 upon sponge embryology ; that is to say, the ciliated cells 

 of the larva become the ciliated cells of the adult, in spite 

 of the very great changes in position this implies, and 

 nothing else is formed from them. In other words, the 

 gastral layer when once differentiated remains uniform in 

 nature throughout the life cycle. 



Nothing is more difficult than to establish a negative, 



and the above conclusion as to the uniformity of the gastral 



layer would be at once upset if it were shown clearly and in 



a trustworthy manner that in any single case collar cells are 



^Compare Maas (1893), pp. 423, 424. 



