IO8 \V. M. SMALLWOOD. 



spermatogonium divides, the division does not completely sepa- 

 rate the two cells, but leaves them in connection with the blasto- 

 phore. Each cell surrounding the blastophore has a direct pro- 

 toplasmic connection with all of the other cells. Voigt finds as 

 miny as 128 cells surrounding the blastophore. One hundred 

 cells were located with the camera lucida on the upper half of 

 the blastophore of BrancJiiobdella instabilia, which would give 

 about two hundred cells in all. The whole structure is so small 

 that an accurate account is practically impossible. But from this 

 estimate we can get some idea of the number of generations of 

 spermatogonia and spermatocytes. 



As the spermatid elongates, it remains attached to the blasto- 

 phore by the tip end of the head and continues so connected 

 until the spermatid has become a full-sized spermatozoon. The 

 blastophore becomes reduced in size during this growth, vacuoles 

 appear in it and the plasma stains give a more marked reaction. 

 The reduction in size is so great in some instances that the blasto- 

 phore is no larger than the leucocytes, but can be distinguished 

 from them because of its non-nucleated condition. I believe that 

 the blastophore is nutritive in character because of the origin and 

 attachment of the developing spermatids and the accompanying 

 degeneration in it as the spermatozoa approach full size (Calkins, 

 p. 276), concludes his discussion of the blastophore with the fol- 

 lowing statement : "It lives as long as the developing germ cells 

 are connected with it and dies when deserted by the spermatozoa. 

 Nor is there any reason to suppose that it provides nutriment for 

 the spermatid cell." 



The phenomena of reduction are not readily determined in 

 BrancJiiobdella instabilia. Continued search has been made to 

 ascertain the number of chromosomes and at what stage the 

 reduction occurred but thus far I have been unable to satisfy my- 

 self or to secure constant results. The size of the cells shown 

 in Figs. 5 and 6 is the main reason for my failure. A study of 

 the karyokinetic division of the cells surrounding the blastophore 

 shows that the spindle fibers arise from the nucleus and termi- 

 nate in a definite centrosome at each pole. No astral rays are 

 present. The centrosomes lie close to the nuclear wall which 

 gradually breaks down (Figs. 5 and 7). 



