GRANULAR ERYTHROCYTES 51 



the cell to divide until that division was arrested 

 "by total death on the staining of the centrosome. 

 But the red blood-corpuscles could not be induced 

 to divide on such jellies. 



In the paper before quoted (3) it was noted 

 that healthy erythrocytes, when examined on 

 coefficient jelly which contains salts and stain, 

 show in about one in every hundred corpuscles 

 a slight granulation. These granules are scattered 

 irregularly inside the cell, and they stain at the 

 same time and under the same conditions as the 

 granules of leucocytes ; they give the chromatin 

 reaction. These granular red blood-corpuscles 

 correspond to the punctate or granular basophilia 

 described by many pathologists. By the older 

 fixed -film methods of blood -examination the 

 granules would sometimes give the chromatin 

 reaction, or more commonly would stain (say, by 

 Leishman's method) a pale blue, contrasting with 

 the eosin-coloured normal erythrocyte. 



In the book already referred to (1) it was stated 

 that in cases of cancer these granular cells were 

 considerably increased in numbers and that the 

 numbers of their granules in individual cells were 

 increased too until in certain erythrocytes there 

 appeared conglomerate masses of granules packed 

 together in the centre of the cell. By the jelly 

 method these granules could be easily distinguished 

 as such, but by the fixed-film method the cells 

 appeared as nucleated red blood-corpuscles. By 

 the jelly mode of examination it was at once seen 

 that the so-called nucleus is composed entirely of 

 granules packed together as described, and that 

 there is no nuclear wall or limiting membrane. 



