At the risk of repetition let it be clear that too much weight must not be attached to any of the so-called dietary standards, i.e. to any attempt to state the requisites of an adequate diet in terms of quantities of certain nutrients. As Atwater sought strongly to emphasize, a dietary standard at best is "only an indication, not a rule." Some of those who have been most active in recent investigation are most emphatic in warning against the expectation that dietary standards can be made to embrace all the qualities which a diet must have in order to be permanently adequate. Thus Hart, McCollum, Steenbock, and Humphrey in a very recent article * say:

"With this recognition of all the normal factors for adequate nutrition there must not simultaneously arise a desire for a mathematical expression of these factors in feeding standards. It is doubtful if this can ever be done, at least for certain of them. For example, the role of the mineral nutrients is so varied, including such widely separate functions as construction and control through antagonism, as to make it seem futile to attempt an expression of absolute requirements when natural foods, with their diversity of mineral content, are involved. Even the recognition of differences in the quality of proteins and their relation to nutrition will make it more difficult to continue expressing protein requirements in exact quantities than before the development of such knowledge; and what can be said of the quantitative requirements of fat soluble A and water soluble B and their supply in feeding materials? We need more effort placed on the accumulation of information on the physiological behavior of feeding stuffs than on the attempts to bring out new mathematical expressions of feeding standards."

* Proceedings of the National Academy of Sciences, Vol. 3, page 374 (May, 1917).