MALIGNANCY 35 



that cancer cells mould or determine connective tissue " to 

 their requirements." This may mean much or nothing, 

 for I confess to having seen few such loose statements. 

 Waldeyer's opinion was more complicated than Thiersch's, 

 and fuller of assumptions. He held that the epithelium 

 was weakened and, being pressed on by the connective 

 tissue, was in parts isolated, and thereby in some inexplic- 

 able way liable to transformation into cancer cells. Since 

 this transformation is the problem, we should be no further 

 advanced if insistance on the material of change were not 

 distinctly useful. Durante and Cohnheim, also, seem to 

 have been in favour of the theory that the epithelium and 

 connective tissue directly influenced each other; but 

 Cohnheim was led away by the sequestration or " cell- 

 rest " theory which is due to him. Modern research seems 

 to support his opinion that tumours are frequently to be 

 attributed to such causes ; but the malignity of some and 

 the benignity of others is still to be explained. It is absurd 

 to suppose that embryonic " rests " always occupy the 

 sites of tumours started by irritation, and Cohnheim himself 

 excepted certain cases where that seems the immediate 

 cause of malignancy. Ribbert held that such " rests " can 

 be created post-natally, and that epithelium, when cut off 

 from its ordinary physiological control, can proliferate 

 malignantly. Implantation tumours by themselves are suffi- 

 cient disproof of this view. Adami attributes cancer to an 

 acquired " habit of growth." The cells devote themselves 

 to mitosis. After what I have said I need not add that 

 this is merely re-description. It deals with " how ? " 

 not with " why ? " Green attributes cancer largely to the 

 influence of the combustion products of coal or peat, with 

 a high percentage of sulphur, as well as to low-lying valleys. 

 While such may be contributory factors to a loss of sym- 



