SPERMATOZOA OF GTJANCHA AND SYCON. 303 



spots within the spherules of the gonocytes are probably artefact. In some 

 cases the spherules are indistinguishable except in number from those in the 

 smaller unripe ova underlying the collar-cells, the spherules which I have 

 shown (1895, p. 30; 1892 h, p. 477, fig. 4) to be the basal spherules of the 

 collar-cells, withdrawn through the pseudopodia of the ovum for its susten- 

 ance (cf. Dendy, 1893, pp. 221, 222, fig. 50 ; 1914, pp. 328, 344, fig. 48). 

 The subcloacal gonocytes (gland-cells of Dendy, 1891, fig. 26) are ova in 

 the process of being fertilised through a micropylar pseudopodium per- 

 forating the pinacocytal epithelium (cf. Dendy, 1891, fio-s. 36, 37; 1914, 

 p. 325). I have added figs. 5 and 6 (PI, 24) to show the structures more 

 clearly, but must postpone more detailed description and discussion to another 

 paper. The difference is very startling from the fertilisation under the 

 collar-cells of Grantia eompressa through the intervention of a nurse-cell, as 

 described by Mr. Gratenby and accepted by Professor Dendy at the Linnean 

 Society (Dec. 11, 1919), and implies an unexpected genealogical separation. 



The organism of figs. 1 &2 (PI. 24) was measured 6'7/i greatest x 4 - 6 /j, 

 smallest diameter, button "2'2fix 1'6/t, tail 6/x. It would therefore be about 

 one-third the volume of the dolly-cell, which is 8 /u. to 10 /t in diameter, with 

 spherical button-nucleus 2'ifi to 3'6ju.. The difference is large, but not 

 beyond the possibility of differing nutrition. In each case the cell contains 

 one or more large clear bodies, which Mr. Gatenby points out to me would 

 correspond to mitochondria or middle-bodies, if these are spermatozoa — for 

 the present I will give them the non-committal name of " grains." These 

 grains and the button-nucleus appear to me to identify the dolly-cells with 

 the organisms of figs. 1 & 2. Fig. 3 I have added merely as a record of 

 fact, for those who have reason to believe in an altogether smaller sperma- 

 tozoon ; though the tail was 8"0 /x long the body was only 1'3/i diameter, 

 or -^q the bulk of a dolly-cell. [Identity of fig. 1 and dolly proved, Oct. 25.] 



The stained object in the ovum (under the nurse-cell) of Dendy's fig. 52 

 (1914), which, as I understood Professor Dendy to state on Dec. 11, he 

 now considers to be the male pronucleus, measures 24/ix l'l fi, dimensions 

 fairly reconcilable with those of the button-nucleus alike of figs. 1 & 2, or 

 of the dolly-cells. (The latter show often a second nucleus, with spindle.) 



EXPLANATION OF PLATE 24. 



e=cloaca. <2= dolly-cell. e = cloacal epithelium (pinacocytes). /=flagellate cells 

 (choanocytes). f.c. = cavity of flagellate chamber. <7 = gonocyte. m = rnesoglcea. 



The observations were on Naples sponges ; the time of year is given at which the sponge 

 was taken. 



Fig. 1. ( x about 1800.) March 24. Sketches in different positions of an organism found 

 in a living section of Syan raphanus t. aquariensis in bismarck brown. No part 

 of this organism was stained by the brown. The stiff tail was about ift thick, 

 being noted a> twin- tin thickness of the llagella of the collar-cells. 



23* 



