XVII. Mrs. Giinther Geffers being consulted by German detec-

tives to solve a criminal case. She always "worked" in a trance.
WEATHER PROPHETS 193
also have keen eyes. Though proverbs about the wind, of which there are a vast collection, are rarely reliable, the wind is one of the best empirical criteria for judging weather changes.
Unfortunately traditional weather-lore can at best hope to tell the farmer his immediate meteorological future, when country- folk, more than anyone else, need predictions that not even scientific meteorology can provide at present. They would like to know many months in advance what the weather will be like at harvest time, whether the next winter will be particularly severe, the next summer dry, if there will be spring frosts, etc. Though most of the rules of thumb they use are based on past experience, popular weather-lore is like scientific meteorology in that the accuracy of both falls off rapidly the longer the term of the forecast.
Inasmuch as they are not based on some miraculous event in the far distant past, common-sense forecasts for the next season or for the next year work in much the same way as professional forecasts: they are either based on the principle of compensation or in the belief in cycles. There is a very widespread belief that within twelve months — ^not necessarily twelve calendar months — the total temperature is always constant, and that, for instance, a severe winter is followed by a very hot summer. Similarly with humidity — a dry winter is usually followed by a rainy summer.
The other idea, viz. that the weather is governed by periodic influences, comes to meteorology from astrology. The popular weather planet par excellence is the moon, every one of whose phases is said to have a special significance, the new moon, in particular, bringing weather changes. Careful investigations over the years have, however, shown that the moon has no perceptible influence whatsoever on the weather.
Sunspots
Meteorologists do, however, recognise the effect on the weather of other periodic phenomena. The most important of these is the sunspot cycle with a mean period of 1 1^ years. While the exact
194 CHAPTER 6
causes of sunspots are not yet fully understood,^ there are never- the-less a number of brilliant hypotheses on the subject. Formerly sunspots were considered to be solar clouds (KirchhofF), slag deposits (Zollner) or electro-magnetic storm centres (Hale), but recently they have been thought to be associated with atomic processes (Bethe). But none of these hypotheses satisfies astronomers or meteorologists completely.
Now this lacuna in our knowledge would not necessarily affect practical prediction, were it not that no one can foretell precisely when sunspots will appear with enough intensity — weaker sunspots are invariably present — to influence the earth's atmosphere appreciably. In this respect, uncertainty has become even greater than it was, for in 1843, when an amateur astro- nomer— Samuel Heinrich Schwabe — discovered the periodic rise and fall of sunspots, it was thought that he had found a perfectly periodic phenomenon. But subsequent investigations and also historical research — sunspots have been known in Europe since the beginning of the 17th century, and the first Chinese records go back to A.D. 188 — ^have made it clear that the spots are by no means as regular as, for instance, the movement of the planets. The 1 l|-year cycle is only a very rough arithmetical average, and sunspot maxima may occur at intervals of from 7-17 years — much too great a discrepancy for any kind of weather forecasting. Moreover the periods seem to be growing longer.^
Meteorologists were not alone in being fascinated by sun- spots, and all manner of people tried to relate them to all sorts of terrestrial processes — ^from plagues of insects to South American revolutions. Above all, sunspots were thought to have a direct effect on economics.
In 1878, the English economist W. S, Jevons drew attention to the fact that the occurrence of economic crises coincided almost exactly with the sunspot cycle. Clearly the two were related. The American economists Warren Persons and H. L. Moore then pursued Jevons's thesis and tried to make it more plausible, by investigating the influence of sunspots on the weather and hence on grain prices. If the sun is very disturbed and full of spots it was said to have an adverse effect on the 1 H. T. Stetson: Sunspots in Action (N.Y. 1947), p. 132. * Albert Ducroq: La Science a la conquete du passi (Paris 1955).
WEATHER PROPHETS 195
weather and hence on harvests, whereas a bright sun was said to have the opposite effect. Sunspots therefore tend to drive grain prices up as a result of poor harvests, while the absence of sunspots and the consequent good harvests tend to depress the grain market — in theory, but not in practice.
There is a much closer connection between sunspots and large-scale storms, but even then only in certain localities. While sunspots seem to be associated with storms in Siberia, Scandinavia, the West Indies, the south-eastern part of the United States, and the South Pacific, they are less so in neigh- bouring regions, and they may even have the opposite effects in other parts of the world. Some meteorologists even maintain that Europe has the worst storms when sunspots are at a minimum.
Since sunspots have proved such unreliable guides, people have looked for other periodic effects on the weather. At the moment, 22-23 and 44-46 year cycles (twice or four times the mean sunspot cycle ) are in great vogue, mainly on the evidence of certain anomalies in the annual rings of old trees, and increased periodic deposits in some African lakes. In the U.S.A. where cycle theories are particularly fashionable, they have been used to predict a severe drought for 1975, and an even severer one for 2021,1
It is due to these new cycle theories, that Bruckner's famous theory, which was once considered to be unassailable, has fallen into oblivion. On the basis of painstaking research into older weather reports, Eduard Briickner^ put forward the thesis that the European climate has been subject to periodic fluctuation for at least 1000 years, with individual cycles lasting for about 35 years, i.e. about three times as long as a sunspot cycle. In the first half of each cycle the weather was mainly warm and dry, and in the second half mainly cold and wet. Since Bruckner's individual cycle was, however, merely a statistical average, it could never be applied in practice: a current cycle might take anything from 20 to 50 years to run its full course.
1 George Kimble and Raymond Bush: The Weather (N.Y. 1944), p. 178. — I. R. Tannehill: Drought. Its Causes and Effects. (Princeton 1947), pp. VII and 175.
2 E. Bruckner: Klimaschwankungen seit 1700 (Vienna 1890) and Klimaschwankungen und Volkerwanderungen (Vienna 1912).
196 CHAPTER 6
The long summer
The most awkward thing about weather cycles is not so much their vagueness as their great number. Thus an English meteorologist noted in 1936, that there were 130 different cycle theories, with cycles varying from 14 months to 160 years. ^ With such a wealth of cycles, anything or nothing at all, can be explained. Possibly, the most complicated weather processes result precisely from the interference between two or more such cycles, but a graph of all the combined cycles would look more like a maze than like a useful chart. Cycle theories must be simple, for if there are too many independent wheels turning at different rates, the human mind, too, begins to spin in a circle. While any or all of these cycles may, in fact, exist in nature, their very interactions would make them too complex to be of practical use. It is for this very reason that meteorologists are increasingly returning to linear theories, and concentrate on actual trends rather than on the periodic return of past phenomena.
It is generally conceded today that there are, in fact, long term meteorologic trends. Our climate has undergone con- siderable changes not only in the ice age or in prehistoric epochs, but also within historical times. At this very moment there is a marked tendency towards warmer weather, for while summer temperatures have not changed a great deal, our winters have become milder.
Apparently, this tendency began in the middle of the 19th century, though, in view of the sparsity of records and system- atic observations, it is impossible to be certain. What is clear, is that the tendency was too gradual and irregular to be observed by untrained men until the beginning of the 20th century, when the winter temperatures, particularly in northern regions, increased so much that no one could mistake the fact, particu- larly since marked changes in the economic field and in our habits have resulted from it. In Western Europe, the average winter temperature is 5°F. higher than it was in the second half of the last century, and in Spitzbergen the increase is as much as 16°F. The coast of Iceland has become clear of drifting ice, and the edge of the main area of Arctic ice has receded towards the 1 Sir Napier Shaw: Manual of Meteorology (Cambridge 1936).
WEATHER PROPHETS 197
pole by some hundreds of miles. ^ But even more temperate European and N. American areas have undergone considerable climatic changes. Thus Washington has 35% fewer days of frost than it had towards the end of the last century, and the same tendency seems to govern the climate in the Southern hemisphere, at least in so far as we can tell from the rather scant data.
We must leave it to the authors of Utopian novels, to paint life in a world completely free of ice, where sunbathers can relax on the shores of the polar sea and take an occasional dip. All we can say is that this tendency towards warmer winters may well continue for another hundred or possibly another two hundred years, with marked repercussions on agriculture, emigration, fuel consumption, and so on.
But how sure can we be of this tendency when January 1940 brought a sudden spell of severe cold, and when the two sub- sequent winters were extremely cold as well.^ Some weather pro- phets became so frightened that they proclaimed the end of a 90-year-old trend, and the beginning of a new phase of increas- ingly colder weather which would most probably last for just as many years. Since then their gloomy prophecies have been silenced by the events, for, though there have undoubtedly been severe winters, milder weather has clearly been the rule during any recent 10-year period. Quite possibly this trend is one of those long cycles spanning a few centuries which can also be reconstructed from old Chinese and European weather reports. In any case, we may take it that the big thaw has set in, and that there are no signs of a new "ice age", or of a glaciation of the kind that occurred in the 17th century.
1 C. E. P. Brooks: Climate through the Ages (London 1950), p. 376.