INTRODUCTORY.THE experiments to be described were devised to determine whether raw cow's milk can replace orange-juice quantitatively as an antiscorbutic substance. If this proved to be the case it was proposed to utilise the results to estimate the antiscorbutic value of human milk.It has been shown by Chick, Hume and Skelton [1918, 1] that raw cow's milk has a low antiscorbutic value. In the case of guinea-pigs, when less than 50 cc. were consumed\ daily, the animals showed no protection from scurvy. If 50-85 cc. were taken weight was better maintained. When 100-150 cc. formed the daily ration the animals grew normally and showed no symptoms of scurvy.When such large quantities of milk have to be administered daily, the conditions of the experiment become difficult, as unless the animal co-operates heartily, it is impossible to obtain a daily consumption of more thaAi about 80 cc.It follows that if human milk has a comparable antiscorbutic potency, its direct quantitative estimation would not be easy to carry out.Orange-juice, on the other hand, is very rich in antiscorbutic substance. Davey [1921] has carried out careful experiments to determine within narrow limits what is the minimum ration which will prevent scurvy. It was found that the animals receiving only 0 5 cc. of orange-juice developed scurvy. Of eleven animals receiving 1-5 cc., six were protected, three died of intercurrent disease, while only two succumbed to scurvy. The minimum ration which will allow of normal life is therefore 1-5 cc., though a certain proportion of animals employed will not show complete protection when receiving this amount.If raw cow's milk and orange-juice could replace one another in direct proportion to their respective antiscorbutic values, by using both these substances at one time as the source of the antiscorbutic factor, much smaller quantities of milk would have to be administered. The conditions of the experiment would be facilitated, and by carrying out parallel experiments with human and cow's milk, it would be possible to obtain comparable values for cow's milk and human milk in terms of orange-juice.