94 REPORTS ON THE STATE OF SCIENCE. 
ascents give for the mean value 5:1, the later 4:8, while the kite ascents © 
give 4:7. It is therefore to be expected that the mean temperature of the 
air in contact with a mountain 3,000 m. high will be 2° to 3° C. below that 
at the same height in the free atmosphere. The elevated parts of the 
earth’s surface exercise a cooling influence on the upper air, 7.e., the moun- 
tains are not cool because the upper air is cooled by adiabatic convection, 
but they are cool because of radiation to space. It follows from this 
that convection does actually raise the temperature of the atmosphere 
up to 3 km. altitude above what it otherwise would be, a fact poimted 
out from theoretical considerations by Gold.' 
The results of direct comparison of simultaneous observations are in 
agreement. Berson? found from a comparison of the temperature observed 
in balloons with that observed on the Brocken (1,140 m.) that the mountain 
was 0°'9 C. colder than the free atmosphere. 
Shaw and Dines* found from twenty-eight kite ascents made in 
July, August, 1902, that the temperature on Ben Nevis (1,343 m.) was in 
all cases lower than that in the free atmosphere at the same height over 
the sea to the west of the mountain, the mean difference being 2°°6 C. 
Additional evidence in support of their result was furnished by the fact 
that the height at which the kite reached the clouds was invariably 
greater than the height at which the clouds were observed over the 
neighbouring hills. They suggested that the difference might be due to 
the westerly stream of air rising to cross the mountains and producing an 
approximately adiabatic gradient of temperature. 
Schmauss ‘ has recently considered the simultaneous values observed 
on Zugspitze (2,965 m.) and recorded at the same height in balloon ascents 
from Munich, 90 km. distant. He found a mean difference of 1°6 C. 
between the synchronous temperatures, and 1°1 C. between the tem- 
perature recorded in the free atmosphere and the mean temperature of the 
day at Zugspitze. In both cases the free atmosphere had the higher 
temperature. Schmauss deduced also from a comparison of the tem- 
peratures on Zugspitze and Sonnblick that the latter was 0°-6 C. colder 
than the former at the same height, and consequently a mountain in the 
middle of a mountainous district is colder than one on the edge of such 
a district. This may be taken as further evidence that the atmosphere 
is cooled by the mountain. 
In dealing with the registering-balloon results, the mean temperatures 
at each kilometre for each month of the year have been formed for ten 
stations: Berlin, England (Pyrton Hill, Ditcham Park, and Manchester), 
Koutchino by Moscow, Munich, Paris, Pavlovsk (near St. Petersburg), 
Strassburg, Uccle, Vienna, Zurich. From these means the mean yearly 
temperature at each height has been calculated for individual stations, and 
the mean monthly temperature at each height for the stations taken 
collectively. 
The following table gives the mean gradient of temperature determined 
from the general mean values :— 
Height . ~ 7021 1-2 2-3 3-4 4-5 5-6 6-7 7-8 
Gradient . =9a'0 4:3 52 58 63 68 72 74 
Height . . 8-9 9-10 10-11 11-12 12-13 13-14 14-15 
Gradient . se 50 3:3 O7 -08 00 -01 
1 Proc. Roy. Soc., vol. 1xxxii., 1909, pp. 47, 67. 
2 Wissenschaftliche Luftfahrten. 3 Phil. Trans. A., vol. ccii. 
* Registrierballonfahrten, Miinchen, 1908. 
