Besides the general work of the International Committee, special investigations have been undertaken by the French, who formed an Aerostatic Commission in Paris. The services of the eminent physicists, MM. Cailletet and Violle, have been enlisted, while a generous patron has been found in Prince Roland Bonaparte. The apparatus of M. Cailletet to bring down samples of air from the high regions may now be described. When the balloon has reached its greatest height a cock of special construction, turned by clockwork, opens and allows the air to enter a reservoir in which a vacuum exists, and then the cock is automatically and hermetically closed. As it is known that the balloon reaches its extreme height in about an hour and a quarter, the time of opening the cock is so regulated, the closing taking place a little later by its further rotation. In order to protect the moving parts from the extreme cold, a receptacle filled with fused acetate of soda is placed in the box containing the motor, so that, notwithstanding the intense cold of the high regions, this salt in assuming a crystalline state gives out enough heat during several hours at least. During the ascent of an _AÈrophile_ air was collected at 50,000 feet, which when analyzed by M. M¸ntz showed what was supposed, viz. that at this altitude the composition of the air does not differ much from that of the lower air. The slight excess of carbonic acid found in the upper air might be due to the oxidation of the grease used on the cock, and the smaller quantity of oxygen, as compared with normal air, might be caused by the absorption of this gas by the grease or even the absorption by the tinned sides of the copper reservoir. By eliminating all possible sources of error in future ascensions, M. M¸ntz thinks that it can be proved whether there are real differences in the air at different heights, for the methods of analysis are to-day accurate enough to show such differences if they exist. But since it is probable that in the regions which can be explored by the _ballons-sondes_, the air undergoes the same mixing that renders the lower air nearly uniform, only the smallest variations in its composition would naturally be found, requiring minute precautions against errors. This is no doubt why previous measures agreed in showing the invariable composition of the air at lower altitudes.