1884.] Rains of Northern India. 15 



TliG above table confirms and justifies the description already quoted 

 from the Vade Mecum, and also the generalization just given, that the 

 area of the winter secondary rainfall maximum coincides with that in 

 which there is also a winter secondary maximum of relative humidity. 

 But it also brings into prominence some further facts, which assist in 

 throwing much light on the causes of the rainfall. In the first place, 

 it is to be noticed that the increase of the relative humidity of the later 

 months, as we proceed from Behar towards the Punjab, is due, solely, 

 to the fall of temperature ; the absolute humidity being almost constant ; 

 but the latter is decidedly lower on the high ground of Central India 

 and Rajputana, south of the Gangetic plain, than on the latter and in 

 the Punjab. These two facts, viz., the uniformity of the absolute humi- 

 dity over the riverain tract, and its decrease on the higher ground to 

 the south, indicate that it is mainly dependant on local evaporation ; 

 being, in fact, furnished by the rivers, the undried swamps left by the 

 autumnal floods, and, in no small degree, probably, by irrigation and 

 the rich vegetation of the green winter crops. In the second place, it 

 is to be observed that this riverain tract also coincides with the region 

 of lower normal pressure, to the north of the axis of maximum pressure, 

 shewn on the normal baric charts, on Plate II. And lastly, the tendency 

 to cloud formation follows, on the whole, the same laws of distribution 

 as the relative humidity of the lower atmosphere, with, however, this 

 important exception ; that, except in April and to a slight extent in 

 March, it is lower in the neighbourhood of the coast (in Lower Bengal), 

 notwithstanding the higher relative humidity of the lower atmosphere, 

 than in the Upper Provinces, where the rainfall generally originates. 



Now putting together the several facts thus independently elicited 

 from the study of our registers, we arrive, I think, at the outlines of 

 a consistent theory of the production of the winter rainfall. We have, 

 in the first instance, steady evaporation over an extensive moderately 

 humid tract, at a comparatively low temperature, it is true, but in an 

 atmosphere, the stillness of which allows of steady diffusion of the 

 vapour to high levels, and the consequent formation of cloud. The 

 slight disturbance of the baric equilibrium which follows (since the 

 vertical decrease of temperature in a cloud-laden atmosphere is slower 

 than in a clear atmosphere), is succeeded by a gentle indraught of 

 warmer and more humid air from the south ; for the Himalaya bars 

 access to northerly winds. A vortex is then rapidly formed, accompanied 

 with an increased cloud-formation, and speedily followed by precipita- 

 tion ; which takes the form of snow on the hills, and of rain over the 

 river plains. The rainfall is invariably followed by a cool wind, and 

 a wave of high barometric pressure from the west, which I can only 



