1S76.] Inequality of the two Semidiurnal Oscillations. 327 



winds of India and Australia, winds which are distinct from convection 

 currents, though, it may be, coexisting with and accelerating them. The 

 relation of these winds to the barometric tides is very marked, but it 

 does not seem that the differences of tidal pressure would suffice to generate 

 them, were there not a movement of the air in the same direction, arising 

 from more persistent differences of pressure. They probably also depend 

 much on local and irregular differences of pressure. 



The air thus removed in the day time from continental areas must 

 collect over the nearest areas of evaporation, with the effect of diminishing 

 the midday fall of pressure over those tracts ; and thus seem to be ex- 

 plained those apparent anomalies in the magnitude of the midday semi- 

 oscillation of the barometer, to which, in the passages quoted from Mr. 

 Buchan's memoir, he has drawn attention ; viz. in the case of the Medi- 

 terranean area and the Atlantic coast of North America. 



The direction in which this movement of the air takes place will of 

 course vary with the locality ; but there will always be, on an average, a 

 greater diurnal movement towards east coasts than towards those facing to 

 the west. This may be illustrated by the case of Calcutta and Bombay, 

 and it is more extensively illustrated by the predominant westerly direction 

 of the land winds of India, and the cold westerly diurnal winds* that blow 

 across the high plains (17,000 to 19,000 feet) of the Changchenmo and 

 Rupshu in "Western Tibet. The reason is sufficiently obvious. As the 

 great semi-diurnal waves of pressure advance from East to West, the local 

 barometric gradient of any place (in so far as it is determined by the 

 diurnal oscillation) will be expressed by a tangent to the existing phase of 

 the wave. During the hottest part of the day, viz., from 9 or half -past 9 

 to half -past 4 or 5, this gradient (which is the steepest and most prolonged 

 of the four) inclines to the eastward and increases the declivity towards 

 east coasts arising from the excess of pressure over the land. In the oppo- 

 site direction, viz., towards west coasts, it goes to diminish that declivity. 

 At night the case is reversed. The west to east barometric gradient, from 

 10 p. M. to half -past 3 or 4 A. M. is in the same direction at that tending 

 to produce an influx of air from the sea towards the land on west coasts ; 

 this however is opposed to the land wind of the coast line, which is a true 

 convection current and arises from quite different causes ; and although 

 traceable in the wind variation at Bombay, it there manifests itself only by 

 decreasing the velocity of the former. There are moreover independent 

 grounds for the influence that this compensating inflow chiefly affects the 

 higher strata of the atmosphere, while the day wind is felt in the lower and 

 more heated strata. At Calcutta the easterly (or negative westerly) 



* This I state on the authority of Dr. Cayley who assures me that ou the high 

 plains these afternoon winds are always from the West. 

 42 



