592 TRANSACTIONS OP SECTION A. 



' I indicated tlie.se peculiarities for the first time in October 1901, in a 

 communication to the Luftschifi'art-Verein at Berlin, then in a communication to 

 the Meteorological Society of France in March 1902, and I have developed these 

 conclusions in a note to the Academic des Sciences in April 1902. 



' A short time after, in the early part of May 1902, Prof. Assmann showed from 

 ascents of six rubber balloons that not only was there a cessation of the decrease 

 of temperature but also an inversion. This inversion had also been very marked 

 in the first ascents by Hermite and Besan^on, but Prof. Assmann sought to 

 explain it as being due to the efiect of solar radiation on the thermometer, while 

 the ventilation produced by the rapid ascent of the balloon showed that it could 

 not be referred to such an error in the thermometer record. 



' Having once demonstrated the existence of this isothermal layer for places 

 in the neighbourhood of Paris, we sought to find the evidence of it in other 

 regions, in order to show that it was a general phenomenon. Ascents made by 

 us and our assistants in the winter of 1900-1901, by M. de Quervain in Piussia, by- 

 Mr. Eggenberger at Bath in England in 1'902, have made it evident that the 

 phenomenon was a general one. On referring to the results of the international 

 ascents made in different countries, it is seen that the cessation of the tempera- 

 ture decrease is found ^in the case of all the balloons sent up, and that it is 

 impossible to refer it to insufficient ventilation, since the phenomenon was well 

 marked in ascents made during the night. Since this time ascents made on 

 board the •' I'rincess Alice " by Prof. Hergesell in 1905 have furnished evidence of 

 the existence of the layer near the Azores ; ascents made in the United States by 

 Mr. A. L. Potch have furnished evidence of its existence there, with the pecu- 

 liarities I have indicated, i.e., high up over high-pressure areas and low down over 

 low-pressure areas. 



' The expeditions of the " Otaria," organised in conjunction with my friend Mr. 

 llotch, have proved the existence of the zone in the tropics, and have shown that it 

 is further from the earth near the equatorial regions, where the Trade Winds meet. 



' Finally, the ascents made at the end of the winters of 1907 and 1908 by the 

 French-Swedish expedition organised by the Observatory of Trappes, with the 

 support of Prof. Ilildebrandson, have shown that near the Arctic Circle at Kiruna 

 the layer exists and possesses general characteristics analogous with those found 

 in these regions. 



' The results of series of daily ascents for eight, ten, or more days in eucces- 

 siott in February 1901, March 190;!, and May 1904 have proved that the change of 

 altitude of the point where the temperature ceases to fall is accompanied by 

 changes of temperature of 10°, 15°, 20° C, in an interval of a day or two at 

 lieights between nine and thirteen kilometres, variations great enough to be felt 

 near the surface during the same time. 



' Thus the equalisation of temperature in the course of the year which had been 

 supposed to be nearly complete at eight or nine kilometres' altitude does not exist, 

 but, on the contrary, sudden changes of temperature occur with the passage of 

 cyclones and anticyclones, which would furnish to an observer in those regions 

 the chief evidence of the changes occurring at the surface. 



' Causes of the Isothermal Layer. — The summai-y of the observed phenomena 

 has led me to this conclusion : that the cessation of the temperature diminutioTl 

 is due to the fact that there is at these heights no considerable vertical con- 

 vection. 



' The fact that one meets with layers of air thousands of metres thick where 

 the temperature increases and decreases rapidlj', and others where it is stationarv, 

 is incompatible with the existence of motion of the air accompanied by pressure 

 variations, which always tend to produce a vertical temj)erature gradient more or 

 less near that for the adiabatic state. It does not follow that the movement in 

 the isothermal layer must be horizontal, but that it takes place along the isobars 

 without crossing these surfaces nearly in the manner in which a body rolls on an 

 inclined plane. 



' These ideas have been developed in several communications, in particular at 

 the Conference d' Aerostat ion scientifique at St, Petersburg in September 1901 



