NATURE AND AFFINITIES OF THE SPONGES. 187 



wards fecundated by spermatozoa independently generated within the 

 same sponge-body. 



Had such a process of production and development been actually 

 substantiated, the Metazoic affinities of the Spongida would undoubtedly be 

 capable of considerable development, but, as a matter of fact, no substan- 

 tial proof of the existence of any such process has been as yet adduced. 

 The evidence available for the gauging of this important question favours, 

 on the other hand, an entirely opposite conclusion. It is here maintained, 

 indeed, as first suggested in the author's communication to the 'Annals of 

 Natural History' for July 1878, that the amoeboid oviform bodies are not 

 independent products of the adult sponge-stock, but simply retromorphosed 

 collar-bearing zooids that have retreated within the cytoblastema, and 

 assumed, as is common to them after passing their matured collar-bearing 

 stage, an amceboid condition. It is these amoeboid units that through 

 coalescence with their fellows attain by degrees the comparatively colossal 

 proportions they present in their most advanced phase of development, and 

 then by the process of segmentation produce the characteristic moruloid or 

 amphiblastuloid ciliated gemmule. There is in this process of evolution no 

 concourse of male and female or true ova and spermatozoa as occurs among 

 the Metazoa, but the phenomena exhibited are in all ways identical with 

 those that obtain among the most simple Flagellate Infusoria, such as 

 Monas and Heteromita, in which the typical flagellate zooids, passing the 

 zenith of their adult condition, enter into an amceboid state, and coalescing 

 in pairs, or even socially, give rise by segmentation to a new generation of 

 flagellate units. On a yet larger scale, and in a manner more closely 

 corresponding with what obtains among the Spongida, identical phenomena, 

 as described at page 42, are encountered in the group of the Myxomycetes 

 or Mycetozoa. The only fundamental point that distinguishes the segmen- 

 tation process in the two groups of the Spongida and ordinary Flagellata 

 is that, whereas in the more simple Flagellata the products of such 

 segmentation are scattered apart throughout the inhabited fluid medium, 

 and maintain an independent existence, in that of the sponge-gemmule 

 these flagellate units are intimately bound to one another at the time of 

 their exodus from the parent colony-stock, and remain associated within 

 their subsequently developed common gelatinous matrix or cytoblastema for 

 the whole term of their existence. Substantial corroboration of the opinion 

 here maintained that the reproductive sponge-gemmules, or so-called 

 ciliated larvae, are the product of the coalescence or fusing with one 

 another of a large number of metamorp~hosed collared zooids, and that 

 they are not independently generated after the manner of true ova, is 

 afforded by the circumstances under which they are naturally met with 

 in the tissues of the parent sponge-stock. In this connection, both in 

 accordance with the author's experiences, and as distinctly shown in the 

 figures given by all the more notable investigators of this organic group, it 

 is invariably found that these bodies are, as shown at PI. VII. Fig. I, 



