96 REPORTS ON THE STATE OF SCIENCE. 
than the other stations. At Pavlovsk the temperature is continually 
lower than at the other places up to 10 km., after which it is higher than 
for the others. If the values for Pavlovsk are taken to be representative 
of the conditions for lat. 60° and those for Strassburg for lat. 50°, the mean 
difference of pressure between the two parallels at a height of 10km. will 
be nearly 7 mm. if any difference in humidity is neglected. If allowance 
is made for the diminished density of the air at this height, it follows that 
such a difference of pressure would correspond to a steady W. wind of 
about 24 m.p.s. (metres persecond).' Above 10 km. the temperature over 
Pavlovsk is higher than in lower latitudes, so that the difference of pre- 
sure would diminish with further increase in height. It would indeed 
diminish more rapidly than the density, so that the wind also would 
diminish in intensity above a height of 10 km. and the mean wind velocity 
would have a maximum value at about this height. It may be noted 
that the effect of the diminished proportion of water vapour present in 
higher latitudes would be to accentuate the difference of pressure in the 
upper air. The increase must however be small, and could not exceed 
2 mm. even if the air over Pavlovsk were perfectly dry ; the actual value 
is probably only a fraction of a millimetre. 
The higher values found for Paris and in England indicate that there 
is a slight horizontal gradient of temperature from W. to E., and this will 
produce in the upper air a corresponding gradient of pressure also from 
W. to E. or from ocean to continent. 
IV. (6) Temperatures under Cyclonic and Anticyclonic Conditions. 
One of the most important questions which arise refers to the possible 
difference in the vertical gradient of temperature over cyclones and anti- 
cyclones. Hann ? deduced from mountain observations that the gradient 
was less for anticyclones than for cyclones, and the difference was so con- 
siderable that the mean temperature of the atmosphere up to 3°5km. was 
5° C. higher over anticyclonic regions than it was over cyclonic. Gre- 
nander * used the observations made in the free atmosphere at Hald and 
Berlin and found similar results both for winter and summer. The fol- 
lowing table gives the mean fall of temperature between the surface and 
5 km. for the different quadrants in winter and the mean values for summer 
taken from Grenander’s results :— 
| Winter | Summer 
Quadrant : a eee ——————_—_—_——__— 
| N E s W. Mean Mean 
= $ ° C) ° ° lie 
Anticyclonic 5 A 243 25-0 19°4 18:9 21:0 | 274C 
Cyclonic . : - 26°5 26°9 28°7 30°6 277 29:9 C 
The mean temperatures at different heights, calculated from Gre- 
nander’s results, are as follows :— 
‘ The corresponding wind between lat. 40°-50° deduced by using St. Louis 
observations is 15 m.p.s. 
2 Sits. Wiener Akad., 1891. 
8 Arkiv fir Matematik, &c., 1905. 
a ee 
