The position, which is, with few exceptions, in the angle between the nucleus and the nebenkern, induces me to consider the differently staining bodies as the same organ of the cell. For some time after the nebenkern has disappeared the acrosome keeps its position; then it approaches the nuclear wall, flattens against it, and later wanders to the apical end, where it forms the tip of the spermatozoon. (See figs. 32-36 and 39-42.) Fig. 16 shows the acrosome back of the nebenkern. It is not an isolated example, still, I think, it is an abnormality. Fig. 37 shows an apparent division of the acrosome. I did not see enough instances to consider it a regular occurrence.
As the cell is elongating, the axial filament is seen apparently growing out of the nucleus. I do not mean to say that it grows out of the nuclear substance, but in almost every cell observed, where the axial filament was incomplete, there was rather a large mass of chromatin gathered at the place where the axial filament was attached to the nucleus. It may be that the smaller one of the extra-nuclear bodies in figs. 4, 7 and 9 is a centrosome which is passing to the equatorial region of the nucleus, where it later develops the axial filament. As already noted, the axial filament does not pass through the nebenkern, but only over its surface. Fig. 29, _a_, _b_, and _c_, shows different cross-sections through the elongating cell. In _b_ the axial filament was cut at an angle.

I have not followed the centrosome through its migrations. In fig. 2 one of the two bodies is the centrosome, but I did not trace its changes farther.

The cells, up to the stage of figs. 13 or 14, have been scattered promiscuously through the cyst; but as the cell begins to elongate, the one end becomes the anterior-nuclear end, and it shifts to the periphery of the cyst. The central part of the cyst now shows the so-called central lumen. As the lengthening goes on the heads of the forming spermatozoa are all turned toward the distal end of the cyst. The cyst becomes very narrow and long, apparently preparing to contain the long, slender spermatozoa. Near the rachis the spermatozoa seem much twisted, and I surmise that they turn so as to have the head foremost when set free into the _vas deferens_. Fig. 42 may indicate some such behavior.






