2. =Free Balloons= are more or less surrounded by heated or stagnant air, because they drift with the wind, and on account of the sluggishness of the thermometers, the temperatures observed at a given height in a balloon are generally higher during the ascent, _i.e._ when passing from warm to cold air, than during the descent, when the conditions are reversed. Again, it is not possible to study the progressive changes in the atmospheric conditions at one place, because observations in a drifting balloon are not comparable with simultaneous ones made at a station on the ground below. With kites, however, the possibility of making frequent and nearly vertical ascents and descents permits observations to be obtained almost simultaneously in superincumbent strata of air. The height of the kite can usually be determined with an accuracy not attainable by the barometer in a balloon. 3. =Captive Balloons=, although constructed so as not to be driven down by wind, cannot rise nearly so high as kites on account of the weight and resistance of the cable necessary to control them, and even the German kite-balloon, on account of its large surface, would hardly withstand the strong winds in which kites can fly. 4. =The Cost= of installing and operating either mountain stations or balloons is much greater than for kites. The exploration of the lower two miles of air with kites flown from Blue Hill is no doubt the most complete ever made at one place. Nearly two hundred records have been obtained in all kinds of weather conditions, and the progressive attainment of greater and greater heights is shown in the table in the preceding chapter. The records from the flights have been discussed by Mr. Clayton; those until February 1897, with the Blue Hill Observations, in Vol. xlii., Part I., of the _Annals of the Astronomical Observatory of Harvard College_, and later records in two _Bulletins_ of the Blue Hill Observatory, in which the changes of temperature and humidity with height, and their relation to the positions of cyclones and anti-cyclones, are investigated. The use of kites for weather predicting, as was said, has been tried by the United States Weather Bureau, but it is certain that further studies, such as have been made on Blue Hill, are necessary before the sequence of the conditions at the earth's surface to the phenomena observed in the upper air is definitely known, so that the latter can be utilized in forecasting.