In difficult cases, especially those treated in cool weather, I sometimes add, at the third week, one half-ounce of cod-liver oil, given half an hour after each meal. If it lessen the appetite, or cause nausea, I employ it thrice a day as a rectal injection; and in cases where the large doses of iron used cause intense constipation, I find the use of cod-oil enemata doubly valuable, by acting as a nutriment and by disposing the bowels to act daily. This may be given as an emulsion with pancreatic extract. This will suit some people well, and result in a single passage daily, but in others may be annoying, and be either badly retained or not retained at all, and may give rise to tenesmus. The question of stimulus is a grave one. In too many cases which come to me, I have to give so much care to break off the use of all forms of alcoholic drinks that I am loath to resort to them in any case, although I am satisfied that a small amount is a help towards speedy increase of fat. Its use is, therefore, a matter for careful judgment, and in persons who have never taken it in excess, or as a habit, I prefer to give, with the other treatment, a small daily ration of stimulus: an ounce a day of whiskey in milk, or a glass of dry champagne or red wine, seems to me useful as an adjuvant, and as increasing the capacity to take food at meals. Nevertheless, alcohol is not essential, and for the most part I give none, except the small amount--some four per cent.--present in fluid malt extracts. Even this is found to excite certain persons, and it is in such cases easy to substitute the thicker extracts of malt, or the Japanese extract, made from barley and rice. So soon as my patient begins to take other food than milk, and sometimes even before this, I like to give iron in large doses. In hospital practice the old subcarbonate answers very well, being cheap, and not unpalatable when shaken up in water or given in an effervescent draught of carbonated waters. In private practice large doses of salts of iron, as four to six grains of lactate at meal-time, are satisfactory; but the form of iron is of less moment than the amount.