Mr. W. S. Kent on the Embryology of Sjjoiiyes. 147 



of segmentation from a primary reptant amoeboid and sub- 

 sequently spheroidal unit in a manner identical with that 

 already detailed of the free-swimming ciliated gemmules. 

 Plate VII. figs. 1 to 7 serve to illustrate the successive 

 phases of this development of a ciliated chamber as observed 

 by me first in a species of Halisarca apparently identical 

 with H. lohularisj and since confirmed by the investigation of 

 innumerable other forms. Figs. 1 to 4 exhibit no deviation 

 whatever from the normal process of segmentation producing 

 the moruloid phase of the so-called ciliated embryo ; and it is 

 only when the separated units or blastomeres assume their 

 next more characteristic and uniflagellate condition that the 

 distinction becomes apparent. Here, as shown in section at 

 fig. 5, the flagella are developed on the interior instead of the 

 exterior border, and project into a central cavity instead of 

 into the surrounding water. The matured development of 

 the same chamber, in which the individual units or zooids 

 have attained their typical form and characteristic collars, is 

 similarly shown at fig. 7. As will be immediately recognized, 

 it needs merely the eversion of this inward-turning spheri- 

 cal aggregation of collar-bearing monads to produce the typical 

 free-swimming gemmule or so-called ciliated embryo repre- 

 sented by fig. 12 of the preceding Plate. At fig. 8, Plate VII., 

 half a dozen monads from the same mature ampullaceous sac, 

 but more considerably magnified, are delineated ; and close to 

 them (fig. 9) is placed, for the purpose of comparison, an 

 example of an independent collar-bearing form, described by 

 me in my monograph of the group under the title of Desma- 

 rella moniliforinvis. This type, which occurs somewhat rarely 

 in salt water, forms small chain-like, free-floating colonies of 

 from two to six or eight individuals only. Apart from the 

 explanation here given, it would be scarcely possible to distin- 

 guish it from the separated spongozoa of the ampullaceous sac ; 

 and it affords another illustration of the close relationship that 

 exists between the sponges and these more simple independent 

 collar-bearing types. Throughout these latter, indeed, when 

 extensively known, types are constantly recurring that mani- 

 fest in their narrower cycle of existence a correspondence with 

 some isolated developmental phase of the separated zooids of 

 the former. 



Although the symmetrically ovate shape, with the collars 

 and flagella of the separate units forming an even and unin- 

 terrupted elegant frill-like border throughout the peripheral 

 surface, as delineated at Plate VI. fig. 12, represents what 

 may be accepted as the most typical and characteristic expres- 

 sion of the so-called ciliated sponge-embryo, it will be found 



10* 



