of the sun’s apparent motions.
When this discovery was effected, men made the most important and, I think, the most satisfactory step towards the determination of cyclic associations between solar and terrestrial phenomena. It is for that reason that I refer specially to the point. In reality, it does not appertain to my subject, for seasons and sun-spots are not associated. But it admirably illustrates the value of cyclic relations. Men might have gone on for centuries, we may conceive, noting the recurrence of seed-time and harvest-time, summer and winter, recognizing the periodical returns of heat and cold, and (in some regions) of dry seasons and wet seasons, of calm and storm, and so forth, without perceiving that the sun runs through his changes of diurnal motion in the same cyclic period. We can imagine that some few who might notice the connection between the two orders of celestial phenomena would be anxious to spread their faith in the association among their less observant brethren. They might maintain that observatories for watching the motions of the sun would demonstrate either that their belief was just or that it was not so, would in fact dispose finally of the question. It is giving the most advantageous possible position to those who now advocate the erection of solar observatories for determining what connection, if any, may exist between sun-spots and terrestrial phenomena, thus to compare them to observers who had noted a relation which unquestionably exists. But it is worthy of notice that if those whom I have imagined thus urging the erection of an observatory for solving the question whether the sun rules the seasons, and to some degree regulates the recurrence of dry or rainy, and of calm or stormy weather, had promised results of material value from their observations, they would have promised more than they could possibly have performed. Even in this most favourable case, where the sun is, beyond all question, the efficient ruling body, where the nature of the cyclic change is most exactly determinable, and where even the way in which the sun acts can be exactly ascertained, no direct benefit accrues from the knowledge. The exact determination of the sun’s apparent motions has its value, and this value is great, but it is most certainly not derived from any power of predicting the recurrence of those phenomena which nevertheless depend directly on the sun’s action. The farmer who in any given year knows from the almanac the exact duration of daylight, and the exact mid-day elevation of the sun for every day in the year, is not one whit better able to protect his crops or his herds against storm or flood than the tiller of the soil or the tender of flocks a hundred thousand years or so ago, who knew only when seed-time and summer and harvest-time and winter were at hand or in progress.
The evidence thus afforded is by no means promising, then, so far as the prediction of special storms, or floods, or droughts is concerned. It would seem that if past experience can afford any evidence in such matters, men may expect to recognize cycles of weather change long before they recognize corresponding solar cycles (presuming always that such cycles exist), and that they may expect to find the recognition of such association utterly barren, so far as the possibility of predicting definite weather changes is concerned. It would seem that there is no likelihood of anything better than what Sir J. Herschel said might be hoped for hereafter. “A lucky hit may be made; nay, some rude approach to the perception of a ‘cycle of seasons’ may possibly be obtainable. But no person in his senses would alter his plans of conduct for six months in advance in the most trifling particular on the faith of any special prediction of a warm or a cold, a wet or a dry, a calm or a stormy, summer or winter”—far less of a great storm or flood announced for any special day.
But let us see what the cycle association between solar spots and terrestrial weather actually is, or rather of what nature it promises to be, for as yet the true nature of the association has not been made out.
It has been found that in a period of about eleven years the sun’s surface is affected by what may be described as a wave of sun-spots. There is a short time—a year or so—during which scarce any spots are seen; they become more and more numerous during the next four or five years, until they attain a maximum of frequency and size; after this they wane in number and dimensions, until at length, about eleven years from the time when he had before been freest from spots, he attains again a similar condition. After this the spots begin to return, gradually attain to a maximum, then gradually diminish, until after eleven more years have elapsed few or none are seen. It must not be supposed that the sun is always free from spots at the time of minimum spot frequency, or that he always shows many and large spots at the time of maximum spot frequency. Occasionally several very large spots, and sometimes singularly large spots, have been seen in the very heart of the minimum spot season, and again there have been occasions when scarcely any spots have been seen for several days in the very heart of the maximum spot season. But, taking the average of each year, the progression of the spots in number and frequency from minimum to maximum, and their decline from maximum to minimum, are quite unmistakeable.
Now there are some terrestrial phenomena which we might expect to respond in greater or less degree to the sun’s changes of condition with respect to spots. We cannot doubt that the emission both of light and of heat is affected by the presence of spots. It is not altogether clear in what way the emission is affected. We cannot at once assume that because the spots are dark the quantity of sunlight must be less when the spots are numerous; for it may well be that the rest of the sun’s surface may at such times be notably brighter than usual, and the total emission of light may be greater on the whole instead of less. Similarly of the emission of heat. It is certain that when there are many spots the surface of the sun is far less uniform in brightness than at other times. The increase of brightness all round the spots is obvious to the eye when the sun’s image, duly enlarged, is received upon a screen in a darkened room. Whether the total emission of light is increased or diminished has not yet been put to the test. Professor Langley, of the Alleghany Observatory, near Pittsburg, U.S., has carefully measured the diminution of the sun’s emission of light and heat on the assumption that the portion of the surface not marked by spots remains unchanged in lustre. But until the total emission of light and heat at the times of maximum and minimum has been measured, without any assumption of the kind, we cannot decide the question.
More satisfactory would seem to be the measurements which have been made by Professor Piazzi Smyth, at Edinburgh, and later by the Astronomer Royal at Greenwich, into the underground temperature of the earth. By examining the temperature deep down below the surface, all local and temporary causes of change are eliminated, and causes external to the earth can alone be regarded as effective in producing systematic changes. “The effect is very slight,” I wrote a few years ago, “indeed barely recognizable. I have before me as I write Professor Smyth’s sheet of the quarterly temperatures from 1837 to 1869 at depths of 3, 6, 12, and 24 French feet. Of course the most remarkable feature, even at the depth of 24 feet, is the alternate rise and fall with the seasons. But it is seen that, while the range of rise and fall remains very nearly constant, the crests and troughs of the waves lie at varying levels.” After describing in the essay above referred to, which appears in my “Science Byways,” the actual configuration of the curves of temperature both for seasons and for years, and the chart in which the sun-spot waves and the temperature waves are brought into comparison, I was obliged to admit that the alleged association between the sun-spot period and the changes of underground temperature did not seem to me very clearly made out. It appears, however, there is a slight increase of temperature at the time when the sun-spots are least numerous.
That the earth’s magnetism is affected by the sun’s condition with respect to spots, seems to have been more clearly made out, though it must be noted that the Astronomer Royal considers the Greenwich magnetic observations inconsistent with this theory. It seems to have been rendered at least extremely probable that the daily oscillation of the magnetic needle is greater when spots are numerous than when there are few spots or none. Magnetic storms are also more numerous at the time of maximum spot-frequency, and auroras are then more common. (The reader will not fall into the mistake of supposing that magnetic storms have the remotest resemblance to hurricanes, or rainstorms, or hailstorms, or even to thunderstorms, though the thunderstorm is an electrical phenomenon. What is meant by a magnetic storm is simply such a condition of the earth’s frame that the magnetic currents traversing it are unusually strong.)
Thus far, however, we have merely considered relations which we might fairly expect to find affected by the sun’s condition as to spots. A slight change in his total brightness and in the total amount