The skin is cleansed with soap and water and dried. A number of little spots are denuded of their superficial epithelia by twirling a small brad-awl, which should not scrape deeply enough to draw blood. Most workers speak of scratching the skin but the brad-awl scrapes to the proper depth more quickly and easily. The spots are marked with the names of the foods to be tested, as milk, beef, potato, oats, etc., and a drop of a five per cent solution in water of the respective proteins is rubbed into the spots. One spot is left as a control, into which normal saline or 3% solution of milk sugar is rubbed, as the proteins of commerce are made up with milk sugar. Within five or ten minutes, there appears a redness and swelling, as with the pollens. As with the pollens, a patient who at any time has been poisoned or, as we now say, _sensitized_ by any of these foods, still has circulating in his blood or fixed in his skin the reactive bodies to that food. These reaction bodies react to that food on the skin by redness and swelling. Food proteins that cause no redness and swelling are thought harmless for that patient. Foods that cause the reaction are thought to be those to which the patient has been sensitized and to which he has not developed or maintained an efficient defence. There is a contradiction here; for the reaction merely shows the presence of defense bodies in the blood and does not tell us whether that defence is or is not efficient. However, even if the argument limps, the results reported are encouraging. Some striking cures have been reported by simply excluding these foods from the diet. The test is simple and harmless if the scratch is not too deep and _if the protein is not injected beneath the skin_. If injected beneath the skin or rubbed into a deep scratch, the food proteins, like the pollen proteins, may be dangerous. If they are absorbed rapidly into the circulation of a patient who happens to have been sensitized to any of them, there is serious danger of anaphylactic shock. If these observations prove reliable, here is a method of selecting a diet for the individual patient that surpasses in accuracy anything that we have ever known. If hay fever is ever a food anaphylaxis, this method of testing the food sensitiveness of the individual patient promises much; but these observations are still too new and unconfirmed and the skin reaction too uncertain to rely on it implicitly yet. There was a time, back in 1908 to 1910, when the skin reaction for tuberculosis too was highly valued. Enthusiasts proposed to test all the school children and all the soldiers and all the factory workers and segregate the tubercular by the skin test. The diagnosis of tuberculosis was to be put in words of one syllable.