August 30, 1883" 



NATURE 



429 



Hill shows, are probably due to local reacting circumstances, 

 and which afford but little hope of our ever being able to forecast 

 droughts and famines in North- West India solely from a know- 

 ledge of the state of solar maculation. 



The fact that the terrestrial effects of solar changes are con- 

 spicuous in some localities and almost totally absent in others 

 seems to many persons inc impatible with the cosmical nature of 

 the influences at work, but to those who study the subject it 

 appears, on the contrary, the only result to be reasonably ex- 

 pected both from experience and analogy. 



Thus ordinary weather is the integral of all the differentiations 

 effected during the regular seasonal changes in solar declination, 

 and it need scarcely be remarked what an endless variety of 

 conditions we have in this case, due primarily to the operation 

 of a gradual and periodic cause. Owing to diversities of super- 

 ficial character, elevation, contour, latitude, &c, we have meteo- 

 rological oscillations set up, differing from each other in phase 

 and amplitude, which, like the tides of the ocean, in some places 

 tend to exaggerate and in others to annihilate each other. So 

 also must it be, where we have solar changes which gradually 

 perform their cycle in a period of years. The forced oscilla- 

 tions they originate, though small, may in some localities, by a 

 unison of oscillations of the same phase, or an absence of oppos- 

 ing oscillations, be exaggerated above the mean amplitude, just 

 as in others they may, owing to an inequality of phase or the 

 clashing together of opposite variations, be rendered in- 

 appreciable. 



Prof. Eliot, in his " Report on the Meteorology of India for 

 1877 " (p. 3), evidently recognises this fact, when he admits the 

 probability that "at one part of the sunspot period one effect 

 of the variation of solar radiation may be to exaggerate local 

 irregularity." While therefore it is probable that we shall find 

 only a few places, where the terrestrial effects of solar spot varia- 

 tion are of sufficient magnitude and regularity to render secular 

 forecasting possible (assuming that our foreknowledge of solar 

 changes is reliable), such a fact ought by no means to be used as 

 an argument against the utility of studying the relations between 

 solar physics and terrestrial meteorology. 



Mr. Blanford, who from the first attacked the somewhat crude 

 hypothesis propounded by Dr. Hunter regarding sunspots and 

 famines in Southern India, has in his own person furnished a 

 practical protest against any such idea, since his researches on 

 the connection between barometric pressure and sunspot variation 

 have tended, not only to confirm the belief in the bond existing 

 between solar and terrestrial changes, but have also opened out 

 new collateral facls, which, if followed up, are certain to yield 

 results of the highest importance to the science of meteorology. 

 His own views on this question, which have frequently been mis- 

 understood in certain quarters, and referred to as adverse to the 

 general question, are concisely expressed in the following sen- 

 tence, which we quote from an official report recently made by 

 him to the Indian Government : — 



" While, however, I am unable to concede to the conclusions 

 hitherto placed on record, that degree of importance which has 

 sometimes been claimed for them, as affording rules of guidance 

 for the prognostication of scarcity and famine, I am fully in 

 accord with the Famine Commis-ioners as to the importance of 

 following up such clues as they afford, and of pursuing with all 

 the means at our command the investigation of the class of 

 phenomena to which they belong. It has happened again and 

 again in the past history of science, that hypotheses, which in 

 their original form were more or less erroneous, have neverthe- 

 less been most fruitful in their results. In giving system and 

 definite purpose to research they have served a most useful 

 office ; and although the course of their verification may have 

 resulted in demonstrating their error, the same process has 

 brought to light the germs of new and unsuspected truths which 

 might have long remained hidden but for the stimulus to investi- 

 gation afforded by rejected theories." 



The nature of the entire question indeed, seems to have been a 

 good deal misunderstood in this country, at least to judge from 

 the extraordinary amount of obloquy and opposition which it 

 has encountered in various quarters. 



On the one hand, it must be obvious to any one who casts even 

 a merely superficial glance at the vast changes in the physical 

 condition of the sun, indicated by the spots, prominences, &c, 

 and the dependence of all terrestrial meteorology on the quan- 

 tity (and perhaps quality) of the heat radiated from our great 

 luminary, that such changes in the former, must be reflected to 

 some extent in the latter, as indeed they are universally allowed 



to be in the case of terrestrial magnetism. On the other hand, it is 

 equally obvious to the merest tyro in meteorology, that such meteo- 

 rological fluctuations, though in many cases distinctly recognis- 

 able, are not only of small average amplitude (especially in high 

 latitudes) when compared with those which occur, as we say at 

 present, non-periodically, but take a period of years to accom- 

 plish their cycle. To imagine therefore, that such changes, even 

 if thoroughly determined, will alone enable us to forecast the 

 general weather of a season or a year, is manifestly irrational as 

 far as these latitudes are concerned, while even in India and the 

 tropics generally, we have grounds for believing that there are 

 only a few places, where the extreme range of the oscillation 

 bears a ratio to the non-periodic changes large enough to con- 

 stitute it the dominant factor of the weather. 



That when the conditions which regulate the larger and more 

 irregular changes are better understood, a knowledge of the un- 

 derlying secular meteorological changes coincident (or nearly so) 

 with the varying'phases of solar activity, will be of great assis- 

 tance in framing seasonal forecasts, it is impossible to doubt. 

 At the same time it seems strangely to have been overlooked by 

 the majority of those who have interested themselves in this 

 fascinating question, that though the sunspot variation in meteoro- 

 logical elements may alone be insufficient to form the basis of a 

 practical system of weather prophecy, it is very likely to prove 

 the key by which the entire weather problem may be solved, since, 

 when once we know the precise qualitative as well as quantitative 

 meteorological effects of a gradual secular change in the solar 

 radiation, coincident with the sunspot cycle, we shall gain an 

 immense insight into the way in which the larger and more rapid 

 oscillations are effected by the ordinary changes in solar radiation, 

 brought about by season, latitude, geographical locality, &c, 

 these latter oscillations only differing from the former, in being 

 more frequent and of greater amplitude. 



Prof. Hill in his investigation, adopts a plan which is obviously 

 necessary to any one who knows the peculiarities of Indian 

 meteorology, viz. the separate treatment of the summer and 

 winter rainfalls. 



The former season embraces the period from May to October, 

 and the latter the remaining months. An eleven-year period, in 

 favour of which there is a good deal of preliminary evidence, is 

 assumed, in order to see if there is any correspondence with the 

 analogous mean period of solar spot variation. The oscillation 

 of the summer rainfall for the whole area about its mean, is then 

 estimated for the cycle, and is found to accord generally with 

 the results for Southern India, and with Meldrum's supposed 

 universal law, in showing a direct variation with the sunspots, 

 the range being twice as great as in Southern India. At the 

 same time considerable irregularity is visible, some of the 

 stations at the border of the monsoon region giving results con- 

 trary to the average variation. The winter rainfall on the other 

 hand shows a much closer relation to the sunspots, the remark- 

 able thing about it being, that instead of varying directly, as the 

 summer fall, it varies inversely with the spotted area. 



The variation of this winter fall shown in the text is very 

 regular, and confirms a conjecture hazarded by the present writer 

 in 1877, that the similar variation which he had previously shown 

 to exist in the winter rainfall of Calcutta would be found to be 

 more decidedly marked in the sub-Himalayan zone to the north 

 of it (Nature, vol. xvi. p. 267, Meteorological Notes). 



In a postscript to the paper Prof. Hill has worked out the 

 variation from a longer series of observations, which were dis- 

 covered, apparently by accident, at the Board of Revenue in 

 Allahabad, and which, by means of registers kept in the 

 Himalayan province of Kumaon, "to which the civil disturb- 

 ances following the mutiny of 1857 did not extend," allows the 

 cycle to be worked out for the period 1844 to 1S78. The final 

 result given in the form of percentage variations from the mean, 

 is as follows : — 



Winter Rainfall of North India 

 Years of t 



Cycle. I 

 Mean per-. 



centage J— 6-8,— 0-6-3-6 - I5'5 -i7'3[+o - 8| + 27'3' + 24 7 +aV ( -5'6 -5 4 

 variation. ) I > 



For the summer rainfall the variation given in the text is as 

 follows : — 



Summer Rainfall of North India 



V C^e° f } " ' * I 4 I S I 6 



4 



■15*5 



5 6 I 7 I 8 



-17*3 +o'8 +27'3 +24 7 



10 , 11 



Mean per-\ 

 centage .+3' 6 +7'4 +9'8 +i2'6 +7*8 -s6'-iol 



variation. 



9 IO " 

 io'o - 7'o -o - 6 



