times the employment of chemical preservatives has acquired a new aspect through the increasing tendency of manufacturers to add to food products antiseptic chemicals in wide variety and of dubious physiological effect.

It is not so easy and simple as it might appear to declare that no substance that is poisonous shall be added to food. The scientific conception of a poison is one involving the amount as well as the kind of substance. Common salt itself is poisonous in large doses, but, as everyone knows, small amounts are not only not injurious, but absolutely necessary to health. Well-known and very powerful protoplasmic poisons such as strychnine and quinine are frequently administered in minute doses for medicinal purposes, without causing serious results.

How complicated the question of using food preservatives really is appears in the case of smoked meats and fish, which owe their keeping qualities to the creosote and other substances with which they are impregnated by the smoke. Although these substances are much more highly poisonous than chemical preservatives like benzoic acid, over which much concern has been expressed, but little if any objection has been made to the use of smoked foods.

The use of benzoic acid (benzoate of soda) as a food preservative illustrates several phases of the controversy. Observations by Wiley in 1908 upon so-called "poison squads" were thought by him to indicate that benzoate of soda administered with food led to "a very serious disturbance of the metabolic functions, attended with injury to digestion and health." On the other hand, the experiments of the Referee Board of Scientific Experts (1909), conducted with at least equal care and thoroughness, were considered to warrant the conclusions that:

(1) Sodium benzoate in small doses (under five-tenths of a gram per day) mixed with the food is without deleterious or poisonous action and is not injurious to health. (2) Sodium benzoate in large doses (up to four grams per day) mixed with the food has not been found to exert any deleterious effect on the general health, nor to act as a poison in the general acceptance of the term. In some directions there were slight modifications in certain physiological processes, the exact significance of which modification is not known. (3) The admixture of sodium benzoate with food in small or large doses has not been found to injuriously affect or impair the quality or nutritive value of such food.

Still later experiments under the auspices of the German government (1913) showed that in the case of dogs and rabbits relatively large doses of benzoic acid (corresponding to sixty to one hundred grams per day for a man weighing one hundred and fifty pounds) were necessary in order to produce demonstrable effects of any kind. This finding may be considered to confirm in a general way the finding of the Referee Board that four grams per day is harmless.

Probably the evidence respecting the effect of benzoic acids and the benzoates when used as food preservatives constitutes as favorable a case as can be made out at the present time for the employment of any chemical substance. Benzoic acid is present in noteworthy amounts in many fruits and berries, especially cranberries, and its presence in these natural foods has never been connected with any injurious action. In point of fact, substances present in many ordinary foodstuffs are converted within the human body first into benzoic acid and then into hippuric acid. Folin's masterly summing up is worth quoting:

We know that the human organism is prepared to take care of and render harmless those small quantities of benzoic acid and benzoic acid compounds which occur in food products or which are formed within the body; we know how this is accomplished and are reasonably sure as to the particular organ which does it. We also know that the mechanism by means of which the poisonous benzoic acid is converted into the harmless hippuric acid is an extremely efficient one, and that it is capable of taking care of relatively enormous quantities of benzoic acid. In this case, as in a great many others, the normal animal organism is abundantly capable of performing the function which it must regularly perform in order to survive. From this point of view it can be argued, and it has been argued with considerable force, that the human organism is abundantly capable of rendering harmless reasonable amounts of benzoic acid or benzoate which are added for purposes of preservation to certain articles of our food. In my opinion this point of view is going to prevail, and the strife will resolve itself into a controversy over how much benzoic acid shall be permitted to go into our daily food.

But we ought to be exceedingly cautious about accepting any definite figure, certainly any large figure, as representing the permissible amount of added benzoic acid in our food. The very fact that we are in possession of an efficient process for converting poisonous benzoic acid into harmless hippuric acid indicates that there is a necessity for doing so. It suggests that even the small quantities of benzoic acid which we get with unadulterated food, or produce within ourselves, might be deleterious to health except for the saving hippuric acid forming process. And because that "factor of safety" is a large one with respect to the normal benzoic acid content of our food it does not follow that we can encroach on it with perfect impunity. What the effect of a general, regular encroachment on it would be cannot be determined by a few relatively short feeding experiments. It is known that while certain chemicals may be taken in substantial quantities for a month or a year without producing demonstrably injurious effects, nevertheless the continued use of the same substances, even in smaller quantities, will eventually undermine the health. Perhaps the final solution of the benzoic acid problem could be best obtained directly from the people at large. If they were to consume benzoic acid as knowingly as they consume, for example, sodic carbonate in soda biscuits, or caffeine and theobromine in coffee and tea, it would not require more than a decade or two before we should have a well-defined and well-founded public opinion on the subject, at least in the medical profession.[46]

With respect to other familiar and more or less poisonous substances used to preserve foods, defense of their harmlessness is far more difficult. Formaldehyde, salicylic acid, sulphurous acid, and sulphite are compounds definitely poisonous in relatively small amounts, their injurious action in minute successive doses in animal experiments is quite marked, and their use in human food products practically without justification. Boric acid and borax are perhaps on a slightly different footing, but are never present in natural foods, and there is no good evidence that their long-continued ingestion in small doses is without injurious effect. It must not be forgotten that all such substances owe their preservative or antiseptic power to the poisonous effect they have upon bacterial protoplasm. It is fair to assume that, in general, bacterial protoplasm is no more easily injured than human protoplasm, and this raises at once the propriety of bringing into repeated contact with human tissues substances likely to produce injury even if such injury is slight and recovery from it is ordinarily easy. In every case the burden of proof should be properly placed on those who advocate the addition of bacterial-restraining substances to food intended for human consumption. It is for them to show that substances