Papers from the Marine Biological Laboratory at Tortugas. 223 



is bounded by the mitochondria. A cross-section through the posterior 

 region of the spermatosome shows that the mitochondria of the secondary 

 ring are closely applied to the bundle of axial fibers (fig. 42). The completed 

 separation of the mitochondrial apparatus into an anterior and posterior 

 portion is shown in figure 43. The peculiar structure of the secondary ring 

 is maintained for a long time and the mitochondria connected with it persist 

 in the adult spermatozoon. The mitochondria of the anterior portion, on 

 the other hand, disappear very suddenly, at least no traces of them can be 

 found after the stage represented by figure 43. It will be noticed that here 

 already there has been a diminution of the denser cytoplasm around the 

 anterior end of the bundle of axial fibers; by the time the centrosomal plate 

 has reached the cell membrane practically all of it has disappeared (figs. 44 

 and 45). While it is possible that the anterior mitochondria may be dis- 

 solved in situ along with the other substance, two other explanations may 

 be offered for their disappearance: (i) they may move in among the axial 

 fibers and there give rise to occasional mitochondrial granules that are later 

 found scattered throughout the length of the bundle or (2) they may be 

 drawn back to the base of the cell and augment the secondary ring that has 

 been formed there. 



By comparing figure 42 with figure 37, it will be seen that after the forma- 

 tion of the secondary mitochondrial ring the posterior portion of the bundle 

 of axial fibers has become more nearly circular, that its diameter is decreased, 

 and that the fibers themselves have drawn closer together. The reduction 

 of the diameter of the bundle is also shown in figures 41 and 42. The process 

 seems to continue proportionately with the gradual growth of the fibers 

 (figs. 44 and 46). 



At first the growth of the axial fibers is very much slower than that of 

 the flagella. It will be remembered that the flagella begin to grow out 

 before the centrioles have divided. By the time the distal centrioles have 

 traversed one-third the distance across the cell, the flagella have reached 

 their maximum growth and are more than twice as long as the axial fibers 

 (figs. 33, 34, 35, 36, and 39). From this point on they gradually become 

 shorter and more ragged in appearance, eventually fusing at a very late 

 stage to form the posterior tip of the spermatozoon (figs. 40, 41, 43-49, 

 63, and 64). This shortening of the flagella is probably due to their partial 

 retraction into the cell, although their ragged appearance would suggest 

 that they have been worn down by attrition with other free cells. 



V. Brunn ('84) has described a similar action on the part of the flagella 

 for the oligopyrene spermatozoa of Paludina. His observations were con- 

 firmed by Meves ('03), who explains the phenomenon on the ground that' 

 the flagella are the free ends of the axial fibers, each one passing through a 

 ring formed by one of the proximal centrioles upon the cell-membrane. 

 The shortening of the flagella is caused by the downward growth of the 

 cell substance around and between them, so that, eventually, a portion of 

 each flagellum comes to lie within the body of the cell. The facts that have 



