STUDIES ON INSECT SPEEMATOGENESIS. 411 



them (see Bowen '22a), the chromatin seems to contract toward the 

 mid-hne of the nuclear cavity, and there condenses to form a single, 

 thread-like axis. This is subject to various peculiarities in the early 

 stages, and it is in respect to these that the large sperm heads seem to 

 differ most conspicuously from the smaller ones. In many (all?) 

 cases, this thread is first formed as an irregular, usually incomplete 

 helical thread, or at least such seems to be its shape (Fig. 18, which 

 does not show the tip of the head). Sometimes this seems to be merely 

 irregular chromatin masses, and occasionally cysts occur in wliich the 

 chromatin is more or less broken up into a series of distinct, well sep- 

 arated, beadlike masses. I do not know what role these may play, or, 

 indeed, whether they are to be considered part of the normal procedure. 

 The first mentioned condition is not unlikely an early stage in the 

 formation of the axial thread of chromatin; while the latter condi- 

 tion is perhaps connected with the end stages in the same process. 



In any event, this irregular thread soon becomes straightened out, 

 probably in part by its further condensation, and the head of the 

 future sperm is thus produced, corresponding remarkably with the 

 structure of the sperm head of normal size. Through the center of 

 the former nuclear cavity passes a smooth core of chromatin, which is 

 at first enclosed in a layer of protoplasm corresponding to the limits 

 of the old nucleus (Fig. 19).^ I have never been able to follow this 

 layer over an entire sperm, for obvious reasons, but originally it 

 probably forms a complete mantle as in the small sperm. A charac- 

 teristic vacuole (perhaps two) is also formed along the head during the 

 final steps, as in the normal sperm, and its origin and fate is doubtless 

 the same, though uncertain, in both cases. During the final conden- 

 sation of the chromatin thread the anterior part of the head often has a 

 cork-screw shape, as though an originally coiled thread were just in 

 process of straightening out. 



Stage p. The spervi completed.— Finally, the outer protoplasmic 

 envelope disappears as in the normal sperm, and the completed sperm 

 (Fig. 20) is now ready to be released into the efferent duct of the testis, 

 a protoplasmic mass having been already sloughed off the tail after the 

 manner customarily followed in normal sperm formation. In stages 

 o and p, the tips of the heads are so very delicate and so closely bunched 

 (in sections) that it is quite impossible to separate a single one from its 

 neighbors. I have accordingly added the anterior part of the sperm 



9 Compare with Figures 168 and 174 (Bowen '22a); the Figures 51, 53, 54 

 and 55 of Paludina given by Meves ('03) are also interesting in this connection. 



