THE POSITION OF SPONGES, ETC. 453 



parts. Noldeke, like Goette bases his theory of the loss of 

 of a whole germ layer in sponges entirely on observations 

 upon the development of Spongilla. How cautious we 

 should be, however, in drawing any far-reaching con- 

 clusions from this apparently highly modified type is 

 obvious from the fact alone that its development has 

 been investigated by Ganin, Goette, Maas, Delage and 

 and Noldeke, and that the observations of all these authors 

 contradict each other upon even most essential points. In- 

 deed, there is hardly any statement, even of detail, which is 

 borne out by the testimony of a majority of the witnesses. 

 I refrain on that account from a detailed criticism of 

 Noldeke's descriptions and figures, though a close exam- 

 ination might perhaps yield, as in the case of Wilson, results 

 not quite in harmony with the author's conclusions. I will 

 only say that while on the one hand the author's observations 

 as to the immigration of the ciliated cells at the metamorphosis 

 are most valuable and highly to be welcomed, on the other 

 hand the theory of the complete absorption of these cells 

 requires far more convincing proof, both for Spongilla and 

 in the case of other types, before it can be entertained. 



Goette, whom Noldeke follows, attempted to give his 

 theory of the loss of the "ectoderm" in sponge larvae 

 a wider basis of fact by interpreting in this sense the 

 observations of Metschnikoff upon Calcareous Sponges — 

 Sycandj-a and Ascetta {Clathrind). As regards Sycandra, few 

 will, I think, deny that Schulze's well-known investigations 

 render such an interpretation impossible. This conclusion is 

 further borne out by my own observations on the amphiblas- 

 tula of Lezccosolenia, where I was able to trace the origin of 

 the collar cells from the ciliated cells of the larva. In the case 

 of other Ascons also I must crive a most decided denial to the 

 supposition that the ciliated cell layer is lost in development. 



It may be said, therefore, that the embryological inves- 

 tigations upon sponges published since my review in the 

 pages of Science Progress, to which reference has already 

 been made, tend on the whole to strengthen greatly views 

 there advocated ; namely, that a complete reversion of the 

 layers takes place in the metamorphosis, whereby the ciliated 



