June 1 6, 1881] 



NATURE 



151 



And first of all let me speak of the sun's influence on the 

 magnetism of the earth. 



Suppose that the chief observa cries of the world have each a 

 vault, and that in this vault a magnetic needle is delicately 

 suspended. We may imagine the sun to be shut out altogether, 

 the only light being that of a lamp which enables us to record, 

 either photographically or otherw ise, on a magnified scale any 

 small oscillations of the needle. The vault may be supposed to 

 be sufficiently deep down to be practically uninfluenced by the 

 heat of the sun, so that it will exhibit no difference in tempera- 

 ture between noon and midnight. Finally there must be no 

 iron or steel .ibout the place, or anything which might affect the 

 needle. Now under these circumstances you would naturally 

 imagine that the needle wonld be perfectly stationary, always 

 pointing in the same direction. Such, however, is not the case — 

 it does not move very greatly, but nevertheless it does move, and 

 its position depends on the hour of the day, or, in other words, 

 upon the sun. '1 he sun cannot heat the chamber in the least, 

 nevertheless it can influence the magnet, and we might even tell 

 in a rough w ay the hour of the day by noting the position of 

 the needle. In this country the needle attains one extreme in 

 its daily progress about five or six in the morning, and the other 

 about one or two in the afternoon, and the difference in position 

 of these two extremes is called the diurnal range of magnetic 

 declination. Here then we have a magnetical phenomenon which 

 depends upon the sun, and which does not take place simul- 

 taneously at the various observatories of the earth, inasmuch as 

 the sun travels from east to west, so that when it is sbc in the 

 morning at one place it may be midnight at another. 



In the next place, we have abrupt magnetical changes analogous 

 to the well-known abrupt meteorological changes, and bearing the 

 appropriate name of magnetic storms. 



A magnetic storm is not a mere local outbreak, but is felt 

 simultaneously at all the various points of the earth's surface. 

 The various needles in the various vaults of which we have now 

 been speaking will all be affected at the same moment of time, 

 and will be found to be oscillating backwards and forwards in a 

 disturbed state. It thus appears that diurnal ranges and 

 magnetic storms are two distinct phenomena. 



To begin with diurnal declination ranges. These have, as their 

 very name implies, a connexion with the hour of the day, and 

 hence with the position of the sun. Again, in middle latitudes 

 declination ranges are greatest in summer wlien the sun is most 

 powerful, and least in winter. Lamont was the first to observe 

 the signs of a long period inequality in the yearly means of the 

 Munich diurnal declination ranges, and in 1852 Sir Edward 

 Sabine succeeded in showing that this inequality followed that 

 of sun-spots previously discovered by Schwatie, maximum ranges 

 corresponding to years of maximum sun-spots, and minimum 

 ranges to' years of minimum sun-spots. In the same year 

 Dr. Wolf and M. Gautier independently remarked the same 

 coincidence. 



But there is more than a mere general correspondence between 

 these two phenomena, for it is believed that all inequalities of 

 sun-spots, whether of long or short period, are accompanied 

 by corrtsponding changes of declination range, a large range 

 invariably accomp.anying a large number of spots. Perhaps I 

 ought to say a large range following a large outbreak of spots, 

 for the solar phenomenon leads the way and the magnetic change 

 follows after it at a greater or less interval of time. I may add, 

 likewise, that we have some evidence which leads us to suspect 

 that particular states of declination range, like particuLar slates 

 of weather, have a motion from west to east, the magnetical 

 weather moving faster than the meteorological. From a pre- 

 liminary investigation which I have made, I even think there 

 may ultimately be a possibility of forecasting meteorological 

 weather by means of magnetic weather five or six days before. 

 It will be noticed, that as far as declination range is concerned, 

 we have no evidence of a direct magnetic action of the sun upon 

 the earth, but we have, on the other hand, evidence that the 

 magnetic effect, like the meteorological, lags behind the cause in 

 such a way that we are inclined to attribute the magnetic as 

 well as the meteorological phenomena to the heating effect of the 

 stm's rays. 



Let us next take magnetic storms. These, as we have seen, 

 affect the various stations simultaneously, so that the magnetism 

 of the earth appears to change as a whole, and in this respect 

 they are very different from the ordinary diurnal oscillations of 

 the needle. Nevertheless, equally with declination range-, mag- 

 netic storms appear to depend on the state of the sun. In 1852 



Sir Edward Sabine showed that in those years when there are 

 most sun-spots there are most magnetic storms, while, on the 

 other hand, years of minimum sun-spots correspond to a 

 minimum number of such storm--. The late John Allan Broun, 

 an eminent magnetician, has given reasons for believing that the 

 greater magnetic disturbances are apparently due to actions 

 proceeding from particular meridians of the sun ; this when 

 verified will be a fact of the greatest importance. 



Ag.-iin, Prof. Loomis of America, from a discussion of 

 135 jcases of magnetic disturbance, concludes that great dis- 

 turbances of the earth's magnetism are accompanied by unusual 

 disturbances of the sun's surface on the very day of the magnetic 

 storm. It might at first sight be thought from this last observa- 

 tion that a magnetic storm is due to some direct magnetic 

 influence propagated from the sun to the earth, and accom- 

 panying a rapid development of spots-, the influence being thus 

 very different from that which may be supposed to cause varia- 

 tions in the magnetic range. But I do not see that this result 

 follows from Prof. Loomis' observations. There is, I think, 

 evidence that the earth before a magnetic storm is in a critical 

 magnetic state — out of relation to its surroundings — and hence 

 a sudden solar outburst may be the immediate occasion of its 

 starting off. But I fail to see any evidence that the influence 

 received from the sun on such occasions is different in kind to 

 that which afl'ects magne' ic ranges. For we know that magnetic 

 storms occur most frequently about the equinoxes, or at those 

 times when the sun is crossing the equator. Now were a 

 magnetic storm produced by a magnetic influence immediately 

 proceeding from the sun, it would be difiicult to understand why 

 there should be any marked reference in magnetic storms to 

 certain months of the year. 



When the magnetism of the earth is in a disturbed state this 

 may of course be rendered visible by means of the oscillations 

 of a delicately suspended magnetic needle. Nevertheless there 

 are associated phenomena of a very conspicuous character which 

 vividly impress us with the reality of the occurrence. One of 

 these is the aurora — displays of which invariably accompany 

 considerable magnetic storms, on which occasions they may be 

 witnessed over a large portion of the globe. 



Another of these is the earth currents which on such oc- 

 casions aflect all telegraphic lines connected with the earth. These 

 earth currents are automatically registered at Greenwich by the 

 Astronomer Royal, and their peculiarity is that during magnetic 

 storms they are very violent, passing rapidly and frequently 

 backwards and forwards between positive and negative. 



We come now to the meteorological effects produced by the 

 variable state of the sun's surface. More than ten years ago 

 Mr. Baxendell of Manchester pointed out that the convection 

 currents of the earth were apparently altered by the state of the 

 sun's surface, and since that time this peculiar connexion between 

 the sun and the earth has been investigated on an extensive scale 

 by observers in va.-ious portions of the globe. Dr. Charles Mel- 

 drum of the Mauritius Observatory was one of the first pioneers 

 in this important branch of inquiry. In 1S72 he showed that 

 the rainfalls at Mauritius, Adelaide, and Brisbane were greater 

 generally in years of maximum than in years of minimum sun- 

 spots Shortly afterwards Mr. Lockyer showed that the same 

 law held for the rainfalls at the Cape of Good Hope and Madras. 

 Dr. Meldrum afterwards took twenty-two European observatories, 

 and found that the law held in nineteen out of the twenty-two. 

 It would however appear from the observations of Governor 

 Rawson that at Barbadoes, and perhaps other places, the maxi- 

 mum rainfall does not coincide with the maximum sun-spot 

 years. As locality has a very great influence upon rainfall, it 

 might be supposed that by measuring the recorded depths of 

 water in large rivers and lakes we should be able to integrate the 

 rainfall over a large area, and thus avoid irregularities tl"e to 

 local influence. This too has been attempted. In 1S73 Gustav 

 Wex examined the recorded depths of water in the Elbe, Rhine, 

 Oder, Danube, and Vistula for the six sun-spot periods from iSoo 

 to 1S67, and came to the conclusion that the years m which the 

 maximum amount of water appeared in these rivers were years 

 of maximum sun-spots, while the minimum amounts of water 

 occurred during the years of minimum sun-spots. 



In 1874 Mr. G. M. Dawson, in America, analysing the fluc- 

 tuations of the great lakes, came to a similar conclusion. Ihis 

 leads me to a very practical and important part of the subject. 

 In countries such as ours we often sufi^er from excessive rainfall, 

 and are rarely incommoded by excessive heat ; but in hot countries 

 such as India a deficient rainfall means a deartli, or even a famine. 



