The traveller should, at the end of each day, give his opinion of the nature of the weather, saying whether he felt it hot or cold, relaxing or bracing, close or fresh. Such observations have no necessary relation to degree of temperature or humidity as recorded by instruments; but the human body is the most important of all instruments, and everything which affects it should be studied. By paying attention to the foregoing instructions, an observant traveller will bring home a far better meteorological log without instruments than a more careless person could produce by the diligent reading of many scales. Yet, in enforcing the importance of non-instrumental observations, we must not leave the impression that the readings of instruments are of little value. It is, in ordinary circumstances, only by the readings of instruments that the climate of one place can be compared with that of another, and only the best results of instrumental work are precise enough to form a basis for climatological maps. INSTRUMENTAL OBSERVATIONS. The minimum requirement of instrumental observations by a traveller is the reading twice daily of the barometer and of the dry and wet bulb thermometers, to ascertain the temperature and humidity of the air, also the reading once daily, in the morning, of the minimum thermometer which has been exposed all night, and on days in camp of the maximum thermometer also. It is very desirable to expose a rain gauge whenever it is practicable to do so. Unless special meteorological researches are to be carried out, nothing farther in the way of observations need be attempted. A very useful supplement to the necessary observations is the use of a self-recording barograph or thermograph; but these are delicate instruments, liable to get out of order unless very carefully handled, and it will not always be possible to make use of them. The observer must understand what his instruments are intended to measure, how they act, and how they should be exposed, read, and the reading recorded. He must know enough about all these things to be able to dispense with unnecessary precautions only possible at fixed observatories, and, at the same time, to neglect nothing that is necessary to secure accuracy in the results.