Great interest attaches to the subject of Evaporation, on account of its connection with rainfall and water supply. It is to be regretted, therefore, that the results hitherto obtained in the endeavour to measure its rate and quantity do not merit much confidence as regards their applicability to the evaporation occurring in nature, owing to the exceptional manner in which the observations have been made. There is this uncertainty about evaporation, that all the experiments relate to that taking place from an exposed water surface of a, comparatively speaking, infinitesimally small area, and can therefore have but a very partial applicability to the conditions occurring in nature. There are two main reasons for this statement. Firstly, the proportion of the surface of the land on the earth which is covered with lakes and rivers is very limited, and the experiments above indicated throw no light on the evaporation from the soil. Secondly, the evaporation from the surface of a small atmometer erected on the ground, with comparatively dry air all around it, is certainly very different from that which would take place from an equal area in the centre of a large water surface, such as a lake. It is of course easy to make experiments on the evaporation from the soil by means of a balance atmometer, but in order that these should possess a practical value, the investigation must be extended so as to include a wide variety of soils, &c., &c. As regards the second point which has been raised, it is recommended by the Vienna Congress to erect atmometers in the centre of water surfaces; but it is not a very easy matter to conduct such experiments with accuracy, owing to the risk of in-splashing from waves. [Illustration: 20. Atmidometer. Scale about 1/5.] BABINGTON’S ATMIDOMETER measures evaporation from water, ice, or snow, and in form resembles a hydrometer, with the difference that the stem bears a scale graduated to grains and half grains, and is surmounted by a light, shallow copper pan. When in use, the hydrometer-like instrument is immersed in a glass vessel having a hole in the cover, through which the stem protrudes. The copper pan is then placed on the top, and sufficient water, ice, or snow placed therein to sink the stem to the zero of the scale. As the evaporation proceeds, the stem rises; and, if the _time_ of commencing the experiment is noted, the rate as well as the amount of evaporation is indicated in grains. III.—RAREFACTION. The diffusion of aqueous vapour through the air and the rarefying influence of heat jointly effect an alteration in the weight of the atmosphere. This alteration of _weight_ is determined by the Barometer, an instrument invented by TORRICELLI, in 1643, and in so perfect a form that in its essential features it has not been superseded. [Illustration: 21. and 22. Construction of Barometer. Scale about 1/18.]