PHYSIOLOGY OF SPONGES. 321 



food- vacuole's contents by a pellet of: gelatinous matter containing refractive 

 or opacpie granules, this is likely to be also the process of digestion in a 

 mesoglceal phagocyte*. The evidence is that the ''masses of granules" 

 are the last products of digestion, and within a phagocyte, or group of 

 phagocytes, may be taken as showing where the combat with an intruding 

 organism has been victorious and complete. 



This speculation has partly originated in the attempt to explain the cell 

 drawn in fig. D, alluded to at the meeting of the Linnean Society on Dec. 11th. 

 With a superficial resemblance to a pylocyte, it differs in its thick body and 

 large vesicular nucleus ; in each of these particulars it resembles Dendj^s 

 fig. 81. Also the plane of its crescent is radial instead of tangential to 

 the chamber into which it projects, instead of framing an aperture in the 

 wall. 



The appearance of the cell, either in fig. 1 D or in Dendy's fig. 81, is 

 unquestionably that of a gonocyte. In describing my feeding experiments, 

 and confirming my previous observations that thegonocytes feed on the basal 

 spherules of the collar-cells, which are stores of digested food (1892 b, p. 474 ; 

 quoted 1893, p. 220 ; rediscovered 1914, pp. 328, 344), I wrote :— " A large 

 number of gonocytes are in contact with collar- cells which contain plentiful 

 carmine ; in only two of them I found carmine-grains, and it is tempting to 

 deduce that vacuoles and undigested food do not pass into the gonocyte " 

 (1895, p. 30). Cotte (1903, p. 448) uses similar language: — " Les coupes 

 de Sycandra montrent quelques grains de carmin ou de charbon englobes par 

 des amibocytes ; c'est la, on peut le dire, une veritable exception." But our 

 experiments show only that the amoebocytes of Calcaronea do not ingest 

 particles which reach the choanocytes from the water of the chamber. This 

 negatives in no way their power to engulf foreign bodies which they may 

 find in the mesogloea. And beside the classical experiments of Metschnikoff 

 (1879), Cotte has shown that they have this power in Reniera (p. 448) and 

 SpongeUa (p. 455); while Dendy (1914) has given a most interesting 

 description of phagocytosis by oocytes and other amoebocytes in Grantia 

 compressa. 



I iuterpet fig. 1 D as an amcebocyte which has digested some foreign 

 substance encountered in the mesoglcea, and is dragging the excreta to be 

 cast out into the efferent water of the flagellate chamber t. And Dendy's 

 fig. 81 indicates that the substance partially digested there, wholly digested 

 in my fig. 1 D, is a lightly-staining sphere of ti fi diameter (with nine spots, 



* [Since this paper was read I have seen van Trigt's interesting work (1919, p. 1G2), in 

 which he records similar faces formed by the amcehoid cells of Spongilla and by the walls of 

 the efferent canals. 



Van Trigt considers that this sponge feeds upon and digests the cells of the given alga 

 which give its colour.] 



t [Cf. van Trigt, /. c] 



