It is difficult to treat any of these cases without a resort at some time more or less to the use of milk. In most dyspeptic cases--and few neurasthenic women fail to be obstinately dyspeptic--milk given at the outset, and given alone by Karell's method for a fortnight or less, enormously simplifies our treatment. Even after that, milk is the best and most easily managed addition to a general diet. As to its use with rest and massage as an exclusive diet in obesity alone or in extreme fatness with anæmia, I spoke in a former edition with a confidence which has been increased by the added experience of physicians on both sides of the Atlantic. Finally, there are exceptional cases of intestinal pain of obscure parentage or seemingly neuralgic, of dyspepsia incorrigible by other treatments, which, having resulted in grave general defects of nutrition, are best treated by several weeks of milk diet, combined with rest, massage, and electricity. Milk, therefore, must be so much used in these cases in connection with the general treatment I am describing that it is perhaps as well to say more clearly how it is to be employed when given alone or with other food. I am the more willing to do this because I have learned certain facts as to the effects of milk diet which have, I believe, hitherto escaped observation. In fact, the study of the therapeutic influence and full results of exclusive diets is yet to be made; nor can I but believe that accurate dietetics will come to be a far more useful part of our means of managing certain cases than as yet seems possible. We are indebted chiefly to Dr. Karell, of St. Petersburg, for our knowledge of the value of milk as an exclusive diet, and to Dr. Donkin for the extension of Karell's treatment to diabetes. I shall formulate as curtly as possible the rules to be followed in using milk as an exclusive diet in dyspeptic states, and in anæmia with obesity, and in the latter state uncomplicated by defective hæmic conditions. For fuller statements as to the reasons for the various rules to be observed in using milk, I must refer the reader to Karell's paper and to Donkin's book.