190 PKOTISTS AND DISEASE 



be passing from the pycno- to the chasmatoplasson state, 

 resulting (in 16 and 17) in the production of stellate amoe- 

 bulae. In 21 is an early anaphase in an intranuclear parasite 

 with chromidial particles distributed in the plasm : 20, a 

 similar parasite at metaphase. 



The two views of mitosis 18 and 19 show chromosmes 

 quite unlike those of any human cell and quite like those of 

 Plasmodiophoraceae and Mycetozoa. 



Expansion into and formation of stellate subdivisions 

 and mitotic subdivision are here alternative processes. 

 Fig. 55, 21 shows a mixture of the two. A few instances of 

 subdivision of intranuclear parasites I observed also in the 

 mammary sarcoma described above. 



The sections of the choriocarcinoma which Professor 

 Primrose kindly gave me are perfectly stained with iron- 

 haematoxylin, a process which shows certain nuclear features 

 better than does acid haematoxylin. 



Conclusions and Comparisons. If the sarcoma of the 

 breast (Figs. 53 and 54) of which a fuller account is given in 

 Part II, were the only tumour known, and the cytology of 

 granulation-tissue were also known, we should be justified 

 in concluding that the tumour was caused by a parasite which 

 first invades the nuclei of connective-tissue cells, escapes 

 thence into the cytoplasm of the host-cell and thence again 

 into the intercellular spaces. The parasite subdivides either 

 with or without the formation of chromatin into amoebulae 

 which tend to assume a stellate form. When chromatin 



