102 
Proceedings of the Royal Society of Edinburgh. [Sess. 
which produces local clouds will also produce local modifications of 
temperature and consequently local modifications of pressure and wind 
velocity. If we ask whether such local variations of temperature and wind 
are at all probable, we have only to refer to the records of the ascents 
of registering balloons and of anemometers, or of pilot balloon ascents, 
to give an affirmative answer. 
Nothing is more noteworthy than the irregular variations in tempera- 
ture-difference as given by a pair of soundings with registering balloons, 
and the curious local irregularities of wind disclosed by pilot balloon 
ascents. Hitherto it has been customary, on quite general grounds, to 
regard them both as possibly due to the uncertainties of observation. We 
now see that they may equally well be important evidence of complication 
in the structure of the atmosphere. 
Those whose temperament inclines them that way have still the 
possibility of uncertainties in observation to fall back upon ; but the better 
plan would seem to be to arrange for simultaneous ascents of registering 
balloons and pilot balloons, so that the actual and computed distribution 
of temperature may be compared. The interesting feature of the compari- 
son would be that, if the method of computation here indicated (with its 
acknowledged uncertainties in taking mean values for p/6 and 0 instead of 
actual values) should prove serviceable, then one pilot balloon ascent gives for 
practical purposes almost as much information as three registering balloons. 
Apart from the uncertainties which have been mentioned, the con- 
clusions as to the distribution of temperature and pressure are incontrovert- 
ible by those who accept Law 1, and per contra if the conclusions are 
sustained Law 1 receives its most complete vindication. 
2 . Sounding of September 1 , 1907 . 
Westerly Wind rapidly increasing aloft. 
Table VI. 
Height. 
Distance of next higher 
isobar in kilometres. 
Distance of next higher 
isotherm in kilometres. 
k. 
k. 
k. 
k. 
k. 
4 
68 S 
oo E or W 
86 S 
119 E 
3 
77 S 
400 W 
44 S 
555 E 
2 
139 S 
294 W 
119 S 
185 W 
1 
196 S 
526 W 
43 S 
80 E 
The increase in the intensity of the pressure-distribution with height 
is clearly shown, and finds its explanation in a steep temperature gradient 
from south to north. 
