424 CYTODIAGNOSIS 



By making smears as for blood beautiful preparations may be obtained. I prefer 

 Giemsa for differentiating cells and Gram's staining for bacteria. 



The wet Giemsa method described for blood gives good results with puncture 

 fluid sediments. 



At the time of securing fluid for cytodiagnosis, cultures should be made on blood- 

 serum for various pyogenic bacteria and, if tuberculosis is suspected, inoculation 

 of a guinea-pig is indicated. 



The interpretation of cellular sediments is more difficult than many 

 books would indicate, there being many factors which tend to compli- 

 cate the findings. 



The polymorphonuclears in purulent fluids often show fatty degeneration, swollen 

 and faintly staining nucleus or a breaking up of the nucleus into small deeply 

 staining masses (nuclear fragmentation). Such fragments in the smear may be con- 

 fusing. The endothelial cells often show fatty degeneration in the cytoplasm and 

 we often note bacteria and other cells which have been phagocytized by them. 

 Where proliferation of endothelial cells is going on actively the cells show a rather 

 deeply staining cytoplasm as compared with the light staining cytoplasm of the cells 

 in transudates. Some authorities attach importance to the Foulis' cells in connec- 

 tion with malignant processes in the peritoneum; often those associated with malig- 

 nant types of ovarian cysts. Such cells are large, often multinucleated and may 

 show appearance as if budding. 



The following are the leading differentiations: 



1. A smear showing almost entirely lymphocytes with a few red 

 cells and very rarely a polymorphonuclear indicates a tuberculous 

 process. 



2. Where a pyogenic process is engrafted on a tuberculous one, we 

 have still the red cells, some degenerated lymphocytes, and in particular 

 polymorphonuclears showing fragmentation of their nuclei. 



3. When a hydrothorax results from chronic hear tor kidney disease, 

 the characteristic cell is the endothelial cell, which greatly resembles a 

 large mononuclear. These cells often are arranged in plaques. 



4. Some authorities consider that the cancer cell can be recognized 

 by its occurring in masses and having a markedly vacuolated cytoplasm. 

 It has been claimed that they contain glycogen by which means we can 

 distinguish them from endothelial cells which they so much resemble. 

 If such cells should show mitosis the finding would be suggestive. 

 For mitotic figures wet fixation with some bichloride fixative, with H. E. 

 staining, is best. 



Jousset introduced inoscopy as a means of diagnosing tuberculosis. The fluid 

 was allowed to coagulate and was then digested with an artificial gastric juice. The 

 digested material was then centrifuged and the sediment examined for tubercle 





