It was not without some misgivings that I thus undertook to make a man of nearly sixty change his lifelong habits of eating. That is something which I consider no physician has a right to do unless there is some very imperative reason for it. Here was, however, a desperate case. It was in the late afternoon particularly that this patient craved drink so much that he could not deny himself. As he ate but very little breakfast and had a hasty scanty lunch, he was at the very bottom of his physical resources at that time, and at the end of a rather demanding day's work. We had {159} to break up his other habits in the hope of getting at the craving. He had taken coffee and a roll for breakfast. I dictated a cereal, two eggs and several rashers of bacon and several rolls. I insisted on fifteen minutes in the open before lunch and then a hearty lunch with some substantial dessert at the end of it. This man proceeded to gain at the rate of a little more than three pounds a week. By the end of two months, he weighed about one hundred and eighty pounds and had not touched a drop of liquor in that time and felt that he had no craving for it. That is some ten years ago, and there has been no trouble with his alcoholic cravings since. He has maintained his weight; he says that he never felt so well and that above all he now has no more of that intense tiredness that used to come to him at the end of the day. Every now and then he says to me in musing mood,--"And to think that I had never learned to eat enough!" For these very tired feelings so often complained of by nervous patients, once it has been decided that there is no organic trouble--for of course kidney or heart or blood pressure affections may readily cause them--there are just two things to be considered: These are {160} flat-foot or yielding arch, and undereating. When there is a combination of these two, then tiredness may well seem excessive and yet be readily amenable to treatment. Persons with occupations which require standing are especially liable to suffer in this way.