664 H. GRAHAM CANNON 



disappears. The stages in which the two halves are far apart 

 (Text-fig. F and j) very probably illustrate this process taking 

 place. 



In a recent paper on the sperm of Hemiptera Bowen (1) 

 enumerates fifteen cases by other authors besides his own 

 description in which the acroljlast of the spermatid in giving 

 rise to the acrosome, which forms the tip of the spermatid, 

 gives off a body termed by him ' the Golgi remnant ', wdiich 

 is lost in the protoplasm of the tail region of the spermatid. 

 Probably the case of the horse-louse must be added to this list. 



In the dog-louse and in H. consobrinus the acroblast 

 is single as in Pediculus. 



Whether or not this bod}'^ which we tentatively called the 

 acroblast is really flic liomologue of the Golgi apparatus in 

 other cells cannot be said with certainty until a more precise 

 definition of the Golgi bodies is found. In the dog-louse in 

 Mann-Kopsch preparations there are indications of the acroblast 

 arising from two or three scattered granules which may be the 

 true Golgi bodies. However, these are not impregnated by 

 prolonged fixation with osmic acid. A character that is as 

 specific of Golgi bodies as is their staining reactions is that they 

 always show a definite relation to the centrosome during 

 mitosis. It is significant that in all the lice examined the 

 acroblast is peculiar in that it exists as such during the sperma- 

 tocyte division, and also it does not show any definite spacial 

 relations to the centrosomes of the dividing cell. 



With regard to the mitochondria, preparations of dog-louse 

 material fixed in Mann-Kopsch completely confirmed the account 

 that we gave of the development of the mitosome in Pediculus. 

 There was a slight difference in that the earliest spermatogonia 

 in the dog-louse showed the c^'toplasm completely filled with 

 vacuolated mitochondria, whereas in Pediculus Professor 

 Doncaster was of the opinion that some of the earliest sperma- 

 togonia showed granular mitochondria. However, the gradual 

 fusion of these mitochondrial vacuoles to form a mitosome 

 consisting of a central chromophilic mass surrounded by two 

 large vacuoles took place exactly as in Pediculus and ■\\'ill not 





