5i8 MICRO-ORGANISMS AND DISEASE [chap. 



same bodies in stratified epithelium of a variety of tissues in 

 normal, and particularly in pathological, conditions which 

 have nothing to do with cancer. The same is shown by 

 D'Arcy Power. Hence I am justified in denying their 

 parasitic nature, not only because they occur in normal 

 stratified epithelium, but also, and chiefly, from the fact that 

 in every respect they resemble epithelial cells wherein the 

 main body of the protoplasm has shrunk around the nucleus, 

 leaving a peripheral portion surrounding it like a capsule. 

 Whether this apparent encysting of nucleated epithelial cells 

 is connected with and dependent on the structural differ- 

 ences of the marginal and central portions of the cell proto- 

 plasm, as is suggested by Heidenhain's researches, I am 

 unable to say ; but as evidence at least tending in this 

 direction, I may state that, however partial or extensive 

 coagulation-necrosis occurs in stratified epithelium, such 

 apparent encysting of the main coagulated mass of the 

 epithelial substance does occur. To this class of appear- 

 ances belong epithelial cells in which the nucleus is shifted 

 to one side and compressed, owing to the presence within 

 the central part of the cell of an almost homogeneous 

 spherical body ; and in this remnants of granular matter 

 may be often recognised. These clear intracellular globules 

 are represented by some observers as the cancer parasites ; 

 but, apart from their remarkable dissimilarity to anything 

 resembling a sporozoon, the possibility of their being indica- 

 tive of and due to a hydropic or colloid change of the cell 

 protoplasm must not be lost sight of. Moreover, the pre- 

 sence of fluid colloid fatty matter or other material within 

 the cell protoplasm is not at all of rare occurrence in 

 epithelial and other cells in various pathological states. 



2. The presence of well-outlined, more or less clear, 

 intracellular bodies showing a more or less distinct peri- 



