There are yet other considerations concerning the relation of diet to cancer which are worthy of attention. We have mentioned some principal agents, which seem unquestionably to have an influence in the production and continuance of cancer, namely, proteids, coffee and tea, and alcoholics. But millions of human beings partake of these with apparent impunity, while in the relatively few they appear to have cancer-genetic powers. This need not surprise or puzzle us any more than do the many other problems in medicine which we are seeking to solve: for we know how often it happens that the system reaches a point where certain things, once well borne, are no longer tolerated. We know, for instance, that Port and Madeira wine certainly can cause gout, but with many individuals they may be indulged in freely for some time before this result follows: likewise that tobacco may even be abused for a long time, without apparent ill effects, when suddenly there is a revulsion of the system and the slightest use of tobacco will be intolerable: also that many edible substances which have long been well borne, will at a certain time act unfavorably and excite eruptions, urticaria, acne, eczema, etc. Psoriasis also furnishes an illustration which may be of service in understanding the relation of diet to cancer; for psoriasis is characterized by a disordered epithelial growth, which both shows on the surface and manifests itself by epithelial prolongations into the corium, which are quite comparable to the ingrowing cellular masses of early cancer: moreover cancer is not very rare in psoriasis patients. In this eruption it has been very clearly demonstrated, clinically and experimentally, that error in nitrogenous metabolism is commonly at the bottom of the eruption, which has been seen to promptly disappear entirely simply under an absolute vegetarian diet, correctly regulated, excluding also coffee and alcohol, without the use of any medical treatment whatever, internal or external; but, of course, this result cannot always be obtained, and sometimes the eruption will relapse during what is thought to be a strict vegetarian diet. There must, therefore, be some systemic disturbance which causes nutritive material, at some particular time, thus to derange cell action in the eruptions mentioned, and the same is true in regard to the production of cancer. Some years ago Braithwaite called attention to the occurrence of cancer among certain peoples who were vegetarians, and attributed it to the great amount of salt which they consumed. While the suggestion of salt being a cause of cancer has been ridiculed, it is quite possible that a great excess of sodium chloride may disturb the salt equilibrium in the blood, by replacing the potassium which is so necessary for proper cellular nutrition, and also by hindering the excretion of uric acid, as Haig has pointed out.