270 DR. J. B. GATBNBY ON THE GERM-CELLS AND 



9. The Question of the Origin of the Germ-cells. 



That ova and spermatozoa ultimately arose from collared cells was the 

 conclusion of Haeckel, but Schultze and others claimed that they arose from 

 wandering amoeboid cells of the mesoglea, which they supposed were 

 unrelated to the collar-cells. Polejaeff and Glorich also consider that germ- 

 cells arise from wandering mesogleal cells, which are not connected with 

 the gastral epithelium. Jorgensen likewise derives his germ-cells from 

 the so-called mesoderm (mesogleal) cells, either resting stellate connective- 

 tissue cells, or amcebocytes. Dendy, as before mentioned, in 1915 published 

 a paper in which he disinterred Haeckel's " heresy," and courageously stated 

 that in his opinion collar-cells did metamorphose into germ-cells. In the 

 present paper I am finally obliged to follow Dendy's unpopular view, simply 

 because I find myself unable to overlook directly confirmatory evidence, 

 derived from examination of my own and other workers' sections. It is my 

 opinion that not only do collar-cells migrate into the mesoglea and subse- 

 quently become germ-cells, but also that collar-cells in situ may metamorphose 

 into oogonia. Schultze, Polejaeff, Jorgensen, and others may be correct in 

 considering the germ-cells as originating from mesogleal amoeboid cells, but 

 I claim to have shown here that collar-cells are the main source of the 

 above-mentioned amoeboid cells, so that in the long run Haeckel and Dendy 

 are correct. 



10. How Collar-cells become Amceboid Cells of the Mesoylea. 



In PI. 22. fig. 24 I have drawn by means of the camera lucida a part of 

 a sponge showing the migration of collar-cells into the mesoglea. The cells 

 marked a 1, a 2, and a 3 are three stages in the process; in PI. 19. figs. 7 

 and 8 are two later stages showing the matamorpbosis of an amoeboid cell 

 into an oogonium. The cells drawn in PI. 22. fig. 24 can only be doing one 

 of two things — either they are migrating into the mesoglea or they are 

 migrating from the mesoglea to the gastral cavity. That they are not doing 

 the latter seems indicated by the fact that just near the figure a 2 in PI. 22- 

 fig. 24 are gaps in the epithelium showing that cells have left their place, and 

 the cell marked by the figure «3 is opposite one of these gaps. When the 

 collar-cells begin to migrate inwards they lose their collar and flaoellum 

 and become amoeboid. As Dendy and Garter showed, collar-cells teased 

 out alive often begin to thrust out pseudopodia. Many amoeboid cells of 

 the mesoglea contain similar cytoplasmic inclusions to those of the ordinary 

 collar-cells, but later their mitochondria often become more marked, as in 

 PL 19. fig. 7. Most of the pieces of sponge I have sectioned exhibit appear- 

 ances such as drawn in fig. 24 of Plate 22, and I am led to believe that this 

 behaviour of collar-cells is of constant occurrence during the life of the 

 sponge individual. 



