DEVELOPMENT OF HALISARCA LOBULARIS. 615 



that of the Metazoa by the fact that in it the hypoblast con- 

 sists of collared flagellate cells. 



Perhaps I may be allowed to add some remarks on a possible 

 mode of origin of the Spongise. 



It is generally admitted that one must begin one's supposi- 

 tions with a colony of Protozoa. Let us then assume the 

 existence at some early period of a closely aggregated colony of 

 choano-flagellate Infusoria, which need not at first have been 

 free-swimming. Encystment in some form or other is frequent 

 in this group of Infusoria, and one may suppose that at some 

 point in its life cycle the members of our choano-flagellate 

 colony passed through a resting stage in which they existed as 

 amoeboid cells, enveloped in a gelatinous cytoblast. 



If now some of the flagellate individuals became ''enjellied" 

 — one can hardly say encysted — at a difiereut time to the rest, 

 the colony would become diff"erentiated into a set of cytoblastic 

 and a set of flagellate individuals. This unstable diff'eren- 

 tiation might very probably prove of advantage to the colony, 

 saving it from becoming wholly inactive at any given time, for 

 while some individuals rested others would remain on duty 

 securing food for the coenobium. 



Some such stage as this appears in point of fact to exist, if 

 we may trust Mr. Saville Kent's description (which certainly 

 stands in need of confirmation) of that compound flagellate 

 Infusorian which he has named Protospongia Haeckeli. 



We may next suppose that the individuals of the colony 

 which remained active derived some special advantage from the 

 existence of the cytoblastic portion of the colony ; possibly it 

 led to a slight economisation of energy by retarding perhaps 

 loss of heat, perhaps osmotic action. By this economising 

 of energy the active individuals may have been relieved from 

 the necessity of encystment, and, growing more rapidly than 

 the cytoblastic portion of the colony, may have formed a con- 

 tinuous layer over it. A double advantage would follow from 

 this, for the increase of mass of the colony not involving in- 

 crease of surface would necessarily lead to some slight saving 

 of heat, and the cytoblastic portion would both serve as a layer 



