342 
MR. S. A. HILL ON THE WINDS OF NORTHERN INDIA, AND THEIR 
On the plateau to the south-west of the Jumna, where the hot winds are even 
stronger and more persistent than over the great plain, there is a positive increase of 
vapour in the westward direction in the air stratum with which our observations deal. 
Thus we have the following mean values for May :— 
Sutna. 
Jliansi. 
J evpore. 
•367" 
•448" 
•459" 
The dryness characteristic of the hot winds cannot, therefore, be due to the eastward 
transference of air near the ground surface from the arid region of the Rajputana 
desert; for, if this were the cause, we should and a gradual decrease of moisture on 
going north-westwards, instead of which we find an increase. 
Taking the variation of humidity in the lower atmospheric strata in conjunction 
with the diurnal variation of cloud, which, as I have shown in the paper last quoted, 
attains its maximum about the time of highest temperature, as it does in most parts 
of the world, the idea of an interchange between the upper and lower strata by local 
convection currents is at once suggested. The westerly winds of the hot weather are 
so dry, perhaps because they descend from regions where the proportion of water 
vapour in the air is normally much less than at sea-level. The diurnal interchange 
betwmen the lower atmospheric strata and those lying considerably higher was first 
suggested by Dr. Koppen,* to explain the daily inequalities of wind direction and 
velocity—inequalities which are opposite in phase over plains and on high mountain 
peaks. In the sequel it will he shown that it probably suffices to clear up all. the 
anomalies observed in the wind system of India. The explanation of the dryness of 
these winds by the descent of air from a higher stratum was previously suggested by 
Muhry, in a passage! quoted by Branford at p. 614 of his paper on the “Winds of 
Northern India” : but this was under the mistaken notion of an aerial cascade pouring 
over the edge of the Himalaya from the elevated plains of Thibet, a notion which is 
quite inconsistent with the facts of observation. 
A general idea of the average relations of pressure and wind direction over the 
Indian area will be obtained from Plate 19, on which have been charted the mean 
isobaric lines and the resultant wind directions at the more important observing 
stations, for the months of January, May, July, and October, typical respectively of 
the cold, the hot, and the rainy season, and the autumn transition period. The 
pressures used in drawing the charts are the means for ten years, published iu the 
Report on the Meteorology of India for 1884. They have been reduced to sea-level 
* ‘ Zeitschrift der Oesterreicbischen Gesellscbaft fur Meteorologie,’ vol. 14, 1879, p. 333. Koppen’s 
hypothesis has also been suggested by Blanfohd (loc. cit.) to account for the hot winds, and he adduces 
their dryness as a reason for believing it to be the true explanation. 
f ‘ Untersuchungen iiber die Theorie und das allgnmeine Geographische System der Winde,’ p. 99. 
