m'clung: spermatogenesis of xiphidium fasciatum. 257 



first spermatocytes. 



During the prophase of the first spermatocytes there occur 

 changes in the chromatin elements that make it possible to de- 

 termine something of their structure and final disposition. 

 With regard to the accessory chromosome I have not much to 

 add to my former descriptions ('99, '02). A spireme stage, 

 such as I figured for Orchesticus ('02, fig. 5), appears, but not 

 so clearly as in other genera of the family. This would seem 

 to be a common condition, for it has been recognized in 

 Locusta by Otte ('06, p. 531), who apparently was unaware of 

 my earlier account. 



Because of the mistaken interpretation by this author of the 

 changes experienced by the accessory chromosome it would 

 seem advisable to present the history of this element somewhat 

 more in detail than I have before done. The long, slender, 

 much coiled thread of the spireme gradually shortens, thick- 

 ens, and uncoils, the result being to produce, generally, a thread 

 bent upon itself at or near the middle (fig. 5). This double 

 thread may lie straight, but more commonly is convoluted with 

 the free ends in approximation. There is, however, no fixed 

 type form, and almost any shape that a flexible thread may take 

 is represented. As the thread shortens and thickens still fur- 

 ther the irregularities of form decrease in number. There is 

 still manifest a tendency for the thread to be bent, usually near 

 the middle, but not infrequently it is straight. Finally, as it 

 approaches the metaphase, it takes on the boomerang form de- 

 scribed in my earlier paper ('99, p. 190). 



The longitudinal cleavage of the thread is not at any time 

 very apparent, but indications of it are not lacking. The point 

 of interest at this stage is that before the metaphase is reached 

 the rod has almost invariably become almost straight, and 

 never shows the U-shape of the early prophase. It was the 

 failure of Otte to notice this point that led him into error with 

 regard to the division of the accessory chromosome. 



The ordinary chromosomes of the prophase show the typical 

 rings, rods and crosses that have now so often been described 

 in insect spermatocytes. There is every reason to believe that 

 they conform to the type of tetrad that is found in other mem- 

 bers of the order. No unusual phenomena were observed. 



During the metaphase of the first spermatocyte a clearer 



