able. They are probably not far out of the way. Mr. Rock wood has endeavored to construct one series of rations at a very low cost, eleven or twelve cents a day or thereabouts, and another series at higher rates, up to forty-five cents a day. Of course, it is not proposed that any person or family should attempt to follow these exactly. For that matter the chemist will evidently have to consult the cook if he proposes to construct dietaries to accord with ordinary tastes. The rations will, nevertheless, help to show how foods may be economized, and what proportions would suffice for the nourishment of ordinary people, and what are the constituents and costs of the different materials. They are estimated to supply very nearly the amounts of nutrients in Voit's standard ration for a laboring man at moderate work.

DAILY RATIONS COMPUTED TO FURNISH NUTRIENTS EQUIVALENT TO THOSE OF STANDARD FOR LABORING MAN AT MODERATE WORK.

Standard ration-Protein (118 grams), 0.26 Ib.; fats (56 grams), 0.12 b.; carbohydrates (500 grams), 1.10 lbs. A. DAILY RATIONS COSTING TWELVE CENTS OR LESS.

D.-DAILY RATIONS COSTING FROM TWENTY TO FORTY-FIVE CENTS.

[Since the delivery of the above address, Hon. Carroll D. Wright, the President of the Convention, has asked me in what form the statistics of dietaries of the people would be most valuable as data for determining the actual amounts of nutritive material consumed. The following brief statement suggests some points in answer to this question.

If the pecuniary cost of the dietary is the only or the principal matter to be considered, the statement of the total amounts of meats, vegetables, and other food-materials purchased may suffice. If, however, it be desired to determine the actual nutritive value of the dietary, and with this to answer the question whether the amounts of the several classes of nutrients, protein, fats and carbohydrates, are insufficient, excessive

or properly adapted to the needs of the consumer, the statements should be of such sort as to permit a reasonably accurate estimate of the actual amounts of nutrients in the food. It is likewise important to know whether all the food is actually eaten, or whether a part of it is wasted.

It is accordingly essential that the amounts of each of the food-materials should be stated in pounds. The legal bushel for certain articles may have one weight in one State, and a different weight in another State. When a chemist is told that the dietary includes so many bunches of onions or cans of corn, though he may know the average composition of corn and onions as they ordinarily grow, he is at a loss to tell how much of the nutritive materials the dietary contained, because of this lack of definiteness in the statement of the amounts of materials. If the statement is made that a family consumed so many pounds of beef in a year, he is again unable to estimate the amounts of nutrients it contained, because there are such wide differences in the composition of beef of various kinds, and espectally in the portions or "cuts," as the butchers call them, from the different parts of the same animal. A glance at the tables of composition of food-materials given in the address above will illustrate this point. I have been much perplexed in some attempts toward calculating the nutritive materials in dietaries from the fact that the amounts and kinds of materials are not stated with such definiteness as to permit an at all satisfactory estimate of their composition. In brief, if it is desired to ascertain the actual dietetic value of the food, or its fitness for the needs of the consumer, or if we seek an answer to the question as to whether it contains nutrients in proper amounts and proportions, and how it might be altered so as to make it better for his nourishment or for his purse, the weight of each food material, and as complete a description as possible of each article, should be given. That an accurate estimate of any given dietary is possible even with data in the form suggested is, of course, not to be assumed. The only way to determine the exact amount of nutrients is to analyze the food actually cousumed. But if the statistician's figures and descriptions are reasonably accurate and detailed, an at least approximate calculation of the composition is possible.]