Thus also is explained the marvellous food-action of these substances. Tea and coffee enable a man to live on less beefsteak. The Peruvian mountaineer, chewing his coca-leaf, accomplishes incredibly long tramps without stopping to eat. And every hardy soldier, in spite of Mr. Parton, has that within him which tells him that he can better endure severe marches and wearisome picket-service if he now and then lights his pipe. The personal experience of any one man is, we are aware, not always conclusive; but our own, so far as it goes, bears out the general conclusion. It was when we were engaged in severe daily mental labour, that we first conceived the idea of employing tobacco as a means of husbanding our resources. Narcosis being steadily avoided, the experiment was completely, even unexpectedly, successful. Not only was the daily fatigue sensibly diminished, but the recurrent periods of headache, gloom, and nervous depression were absolutely and finally done away with. That this result was due to improved nutrition was shown by the fact that, during the first three months after the habit of smoking was adopted, the average weight of the body was increased by twenty-four pounds--an increase which has been permanent. No other dietetic or hygienic change was made at the time, by which the direct effects of the tobacco might have been complicated and obscured. The statement that smoking increases the average weight of the body[23] is not, however, universally true. We have here an excellent illustration of the impracticability of laying down sweeping rules in physiology. Many persons find their weight notably diminished by the use of tobacco; and we frequently hear it said that smoking will not do for thin people, although for those who are fleshy it may not be injurious. In this there is a very natural but very gross confusion of ideas, which a little reflection upon the subject will readily clear up. It is true that moderate smoking sometimes increases and sometimes diminishes the weight; and it is no less true that in each case the result is the index of heightened nutrition! This seems, of course, paradoxical. But physiology, quite as much as astronomy, is a science which is constantly obliging us to reconsider and rectify our crude off-hand conceptions. [23] "Tobacco, when the food is sufficient to preserve the weight of the body, increases that weight, and when the food is not sufficient, and the body in consequence loses weight, tobacco restrains that loss." Hammond, _Physiological Effects of Alcohol and Tobacco_, Am. Journal of Medical Sciences, tom. XXXII. N.S., p. 319.