July 8, 1897] 



NATURE 



227 



during the thirteen years 1881-1893, not merely with 

 regard to the diurnal, but also the monthly and annual 

 variations, together with a comparison of results at the 

 other stations. This is probably one of the first com- 

 plete discussions ever made of autographic records. 



By adding on these thirteen years to the previous 

 period of eye observations, a period is obtained suffi- 

 ciently long to exhibit any secular periodicity, such as 

 that in the mean sunspot period. 



In the case of tropical air temperature, the existence 

 of such a period is so well known that it has been 

 frequently pointed out by Blanford, Eliot, and Hill in 

 these Memoirs. 



In the case of barometric pressure, as Mr. Blanford 

 pointed out some years ago in his memorable discussion 

 of the barometric see-saw between Siberia and Indo- 

 Malaysia in the sunspot period, there is a small periodic 

 variation of the barometric pressure of such a character 

 that over India the pressure is above the normal in 

 years about the minimum epoch of sunspots, and below 

 it near those of maximum. 



Calcutta, as Mr. Archibald states, is too near the axis 

 of the see-saw (which is probably a little to the north 

 of the Himalaya) to show a very marked variation ; but 

 the following figures for the mean cycle show that it has 

 an existence, and must be reckoned with as a factor in 

 the prevalent character of the weather in different 

 years : — 



Mean Annual Barometric Pressure anomaly at Calcutta (Alipore) 

 in the eleven-year sunspot cycle from 1853 to 1893. 



Mean anomaly of annual barometric 

 pressure unsmoothed. 

 Years of cycle. Inches. 



(1) 1853 --0067 



(2) 1854 +-0007 



I4) Is::: ::: ::: X^^])^^-^--^^- 



(51 " t - '0012 



(6) % i; -t--oo42 



(7) 



(8) 



(9) 



(10) 



(II) 



1893 



- -0032 



0100 j 



sunspot. 



- -0060 



- -oon 



These minute variations of pressure might appear too 

 small to be associated with air movements and conditions 

 of any considerable magnitude. It must, however, be 

 remembered, especially by English meteorologists, that 

 the barometric oscillation, corresponding to any given 

 air-motion, is a direct function of the deflecting force of 

 the earth's rotation, and therefore of the latitude. Con- 

 sequently, a variation which would represent an insig- 

 nificant disturbance in latitude 50^ might be attended 

 with serious consequences in India. 



Thus the only marked feature of the great Madras 

 drought and famine of 1876-77 was a slight excess of 

 pressure over the whole area of drought, which nowhere 

 exceeded •04" or -05" in amount. 



The abnormal pressure conditions which enabled the 

 Indian weather department to foretell accurately the 

 last serious drought in India— that of 1891 in Rajputana 

 —were small in amount, depending on variations and 

 anomalies not exceeding 05", which is little more than 

 one-third of the amount of the regular diurnal change 

 between 10 a.m. and 4 p.m. 



The range of pressure for the mean cycle for Calcutta 

 in the present case is 02", and since the extreme range 

 of annual means is only 06", it may be fairly presumed 

 that the cyclical variation bears a sensible ratio to the 

 variations which cause effective changes in the character 

 of the years and seasons. When we further recollect 

 the well-recognised fact that in Northern India the sun- 

 spot variations affect the summer and winter oppositely 

 as regards rainfall, no doubts need be entertained that 

 NO. 1445, VOL. 56] 



the state of the solar surface coincides with variations 

 which are just as real, and probably no less practically 

 itnportant in their results, than those introduced by the 

 diurnal and seasonal position of the sun in the sky. 



It is indeed impossible to live in the tropics without 

 being sensible of the reality of the sunspot influences- 

 In years about the minimum spot epoch the mean 

 annual temperature of the air is from one to two degrees 

 higher, the range is increased, the wind force greater, 

 and the rainfall less and more irregularly distributed than, 

 about the maximum epoch. Moreover, though the total 

 sunspot effects are only a part of what occurs and are 

 often masked by certain larger variations, whose origin is 

 probably terrestrial and reactionary rather than solar 

 and direct, the relation is recognised officially as a factor 

 of practical importance, and allowed for in drawing up 

 the forecasts of the summer and winter monsoons. 



The whole subject of secular changes has received a 

 decided impetus of late from the Memoir, Vol. vi. Part 2,. 

 on " Certain Oscillatory Changes of Pressure of Long 

 and Short Period," by Mr. John Eliot, F.R.S. In this 

 Memoir attention is drawn to a remarkable series of 

 long-period waves of pressure over the Indian area, 

 which vary from six months to two years in duration. 

 The period investigated extends from 1875 to 1894. 



These waves are believed to be of the nature of 

 advances or checks in the general oscillatory system ot 

 air flow across the equator, which represents the so-called 

 south-west and north-east monsoon winds of the Indian 

 area. 



As Mr. Eliot observes : " In part they are probably 

 determined by seasonal conditions in Southern and 

 Central Asia, and in part assist in determining the 

 seasonal conditions in India, and perhaps also in Central 

 Asia." 



The extreme range of these fluctuations, measured 

 by monthly abnormals over the whole period, amounts 

 to about o'l 18", or about the same as the diurnal pressure 

 range at sea-level in tropical India. 



In discussing these waves, Mr. Eliot exhibits a similar 

 series of oppositely phased oscillations in what are 

 termed the vertical pressure anomalies between pairs of 

 stations situated at considerable elevations in the Hima- 

 laya and on the adjacent plains. 



These vertical anomalies (which are simply the vari- 

 ation at the hill stations minus the variation at the plain 

 stations) are now so successfully employed in the pre- 

 paration of the forecasts of the winter rains of Northern 

 India, that they demand a little special reference. 



In a country like India, where climatic changes far 

 outweigh ephemeral weather changes, it is found that 

 95 per cent, of the irregular ephemeral changes of 

 pressure are less in amount than the normal diurnal 

 pressure range, the air movements are slow, massive,, 

 and regular in contrast with those of higher latitudes,, 

 and due quite as much to vertical as horizontal changes. 

 It is therefore to be expected that the density or pressure 

 of the lower mass of air between the hill stations at 

 7000 feet and sea-level, will often show variations of an 

 entirely opposite character to that of the air above. In 

 other words, abnormally high pressure below argues 

 abnormally low pressure above, and vice versA. As a 

 matter of experience, this is found to be almost invari- 

 ably the rule. When the pressure at the upper levels 

 is lower than usual, the vertical anomalies would be 

 negative^ and when higher they would be positive. Mr. 

 Eliot finds, moreover, that the vertical anomaly waves 

 between stations such as Leh and Lahore, Murree and 

 Peshawar, correspond all through to the long-period 

 waves at the stations on the plains, but are exactly 

 reversed in phase. 



This shows either that the variation of flux is entirely 

 confined to the lower atmosphere, or else that the 

 vertical up and down flow remains temporarily uncom- 



