The loftiest atmospheric phenomenon that we can observe directly is the aurora, which has been photographed up to heights of more than 300 miles. The altitude of the aurora is determined by simultaneous observations made at two or more points, and the same is true of shooting stars and their trails, which seem to be especially numerous between the levels of sixty and ninety miles. The so-called “noctilucent clouds,” which shone by reflected sunlight throughout the night for some years after the great eruption of Krakatoa and were supposed to consist of fine dust from that volcano, were probably about fifty miles above the earth. From the duration of twilight we infer that above about forty-five miles the air is so tenuous that it cannot reflect sunlight to the earth. Clouds furnish information concerning the movements of the air at various levels up to ten miles or more. Observations on mountains contribute further to our knowledge of the atmosphere above the ordinary levels of habitation. Of all methods of exploring the atmosphere in a vertical direction, the most fruitful is the use of kites and balloons. In recent years investigations of this character have become so extensive and so highly specialized that they are regarded as forming a separate department of meteorology, known as Aerology. It is by virtue of developments in this field that meteorology has become “a science of three dimensions.” Formerly meteorologists could do but little more than study the _bottom_ of the weather, so to speak; but now they observe it and chart it at all levels. The weather forecaster has daily reports of conditions aloft to aid his predictions both for dwellers on _terra firma_ and for the aeronaut; while the accumulated data of upper-air observations are throwing new light on many difficult atmospheric problems. Scientific balloon ascents are no novelty. Some were made in the eighteenth century, and many famous ones in the nineteenth, including those of Biot, Gay-Lussac, Glaisher, Tissandier, and other daring _savants_. The “record” height for such personal ascents was attained in 1901, when Berson and Süring rose to 35,400 feet above Berlin. Kites were sent up for meteorological purposes even before Benjamin Franklin’s immortal experiment in 1752. Modern aerological methods have, however, little in common with these pioneer undertakings. Existing types of box kites, pilot balloons, sounding balloons, and self-registering meteorological apparatus for upper-air research were developed in the latter part of the nineteenth century, but their use did not begin to bulk large in meteorology until about the beginning of the present century. The epoch-making event in these undertakings was the discovery of the _isothermal layer_.