NATURE AND AFFINITIES OF THE SPONGES. 183 



an intermediate group between the Protozoic and Metazoic sections. 

 Proceeding yet further, it is maintained by the present author that the 

 structure and relationship of the Spongida is altogether Protozoic, and that 

 the phenomena exhibited by the life and developmental history of the 

 ciliated reproductive bodies now under discussion are, equally with the 

 structural composition of the adult sponge, reducible to and capable of 

 correct interpretation only in association with a Protozoic standard. 



Before proceeding to an examination of the evidence that can be 

 adduced in support of this declaration, a passing note may be made of one 

 or two of the points involved in Mr. Balfour's argument. In the first 

 instance, the so-called larval sponge form distinguished by Hacckel by the 

 title of an " amphiblastula," and consisting of one hemisphere of amoeboid, 

 and the other of flagellate cells, cannot, as Mr. Balfour suggests, be 

 accepted as the typical and ancestral form of these bodies, it occurrino- 

 only, and then not persistently, in the group of the Calcispongiae ; a still 

 simpler type having all the cellular constituents alike in form and cha- 

 racter, and thus presenting a closer structural conformance to a simple 

 vesicular morula, is mostly dominant. The important issue at stake, 

 recognized by Mr. Balfour, relating to the digestive functions of the 

 amoeboid and flagellate elements of both the adult and embryo sponge, can 

 fortunately be completely disposed of, and in such a manner as, on Mr. 

 Balfour's own admission, demonstrates that neither the so-called ectoderm 

 nor endoderm of Spongida will correspond with the similarly named layers 

 in the Coelenterata and other Metazoa. Nutrition and digestion are, in 

 fact, accomplished by both the collared flagellate and the amoeboid cells, a 

 circumstance which would require for their strict correlation with the 

 equivalent Metazoic elements, the possession of nutritive functions by both 

 the ectodermal and endodermal layers. 



Passing now to a consideration of the interpretation of these special 

 reproductive sponge structures maintained in this volume, it may be 

 affirmed to be substantially identical with that submitted by the author in 

 the 'Annals and Magazine of Natural History' for August 1878, previously 

 embodied in the communication made to the Linnean Society in June 

 1877, and as extensively confirmed by subsequent investigation. The fact 

 of the case is, that in almost every one of the accounts and illustrations 

 given by contemporaneous authorities, that which may unhesitatingly be 

 pronounced to represent the most important and significant structural 

 element in these reproductive bodies, has been persistently ignored. The 

 omission referred to is the fundamental composition of the free-swim- 

 ming sponge-embryos of collared flagellate zooids, in all ways identical 

 with those that line the interstitial cavities, and constitute the essential 

 factors of the adult sponge. The so-called ciliated sponge-larva is, it is 

 here maintained, in its typical phase of development, not an individual 

 germ or larva, but a motile swarm-gemmule, consisting of a more or less 

 ovate colonial aggregation of typical collared zooids, as shown at PI. IX. 



