242 * KIYOSHI MASUI. 



those of the opossum can not be seen within or ^\-ithout the nucleus, but 

 in the leptotsue fis well as the syuaptic stages a small numlier of the 

 mitochondrial granules usually appear in tha cytoplasm. As Dcesberg ('11) 

 stated it is true that in the synaptic stage most of the granules always 

 appear near the pole of the nucleus where the chromatin spiremes converge, 

 and that they slightly increase in number. But it can not te admitted that 

 this increase in number of the granules is du 3 to the extrusion of chromidial 

 pariicles from the nucleus, for the nuclear wall appeal's distinctly, a;;d neither 

 the granules nor the particles can be found within or without the nucloits at 

 the converging point of the chromatin thro:ul3. But from actual obssrvation it 

 is more probaole that the mitochondrial granules do not originate from the 

 nucleus in the synaptic stage, they usually exist in the cytoplasm throughout 

 every stage and increase by division of the original ones. A number of facts 

 favourable for this view are tD be seen in my materials, of which the following 

 can be cited : 1. The mitochondrial granules are usually found in the sperma- 

 togonial cells. The same facts have been roported by Ddesberg ('11) in the 

 guinea-pig and Jordan ('12) in opossum. 2. The growth of every granule can 

 clearly be seen in the prophase of the first division. 3. Tai th same cell the 

 size of the granules varies considerably. 4. Lineal arrangement of the granules 

 can sometimes be seen. 5. The numl)C:r of the granules differs considerably 

 in diffcrout stages of the cells. 



The only difficulty which confronts us in the assumption of the cytoplasmic 

 origin f)f the mitochondrial griinules, is perhaps the fact that in the synaptic 

 stage they are always gathered ne-u' the pole where the idiosome situated. But 

 this difficulty may be interpreted as follows : By the enormous expansion of 

 the nuclear wall at this stage, the cytoplasm becomes pushed towtwrds the pole 

 where the idiozome is situated, and with this the mitochondrial granules alsJ 

 move toward the same pole. This change of the position of the gi-auulas can, 

 however, be ascrited to the attraction of the centrosomes as BacHSER ('13) 

 indicates. 



VI. The CHRO-^rATOiu Body. 



Tlio cliromatoid body was long ago described in uüimmals b}- many 

 investigators (Meves, Hermann, Leuhossek, Niessig and Duesberg). ßecently 



