STUDIES ON INSECT SPERMATOGENESIS. 345 



the telophase and as in the ring derivatives of Centrums are sub- 

 sequently divided. In these two cases there appears for the first 

 time a relationship between the spindle and the chondriosomes 

 which is essential to an equal, mass division. In Blaps (Dues- 

 berg, '10) the mitochondria! rods are very numerous and much 

 smaller in diameter, and, what is of most interest, are oriented 

 toward one pole of the cell during the prophases of the firtst 

 maturation division. What relation this bears to the centrioles 

 is not stated, but the arrangement is clearly a forerunner of the 

 " palisade " of rods which encloses the spindle during division, 

 and is divided across as in the two preceding cases. A further 

 step in the definite orientation of the mitochondria with respect 

 to the spindle is found in the case of the honey-bee described by 

 Meves ('07). In the bee the early history of the mitochondria is 

 very similar to that which I have described in Euschistus, the late 

 growth period being characterized by a tangled mass of threads 

 enveloping the nucleus on all sides. During the prophases, how- 

 ever, the threads are all aggregated into a single tangled mass 

 situated opposite that centriole which ultimately comes to lie in 

 the functional sperm, and therefore presumably in the vicinity 

 of the other centriole. As the spindle forms, free ends of threads 

 begin to appear around the edge of the mass of mitochondria; 

 and these gradually move toward the opposite centriole until the 

 whole mass is untangled and the threads stretch from centriole 

 to centriole, parallel to the spindle and enclosing it more or less 

 completely in a mantle of threads. Wilke ('13) has also de- 

 scribed the orientation of mitochondria toward the centrioles in 

 the spermatocyte divisions of Hydrometra, but his figures give a 

 rather inadequate idea of the exact chain of events. Finally, the 

 case of Euschistus, described in an earlier section, demonstrates 

 conclusively that the arrangement of the mitochondria prepara- 

 tory to division depends in these forms upon some influence that 

 is as it were focused in the centrioles. There is thus a complete 

 series ranging from a mass of scattered granules dependent for 

 equal division apparently on the exigencies of chance, up to a 

 mass of threads definitely oriented towards the centrioles to the 

 end that an equal division of their substance may occur. How 

 can these apparent contradictions be harmonized ? A possible 



