Let us begin by scanning the available evidence as to solar disturbances previous to the time when accurate sunspot records are available. Two rather slender bits of evidence point to cycles of solar activity lasting hundreds of years. One of these has already been discussed in Chapter VI, where the climatic stress of the fourteenth century was described. At that time sunspots are known to have been unusually numerous, and there were great climatic extremes. Lakes overflowed in Central Asia; storms, droughts, floods, and cold winters were unusually severe in Europe; the Caspian Sea rose with great rapidity; the trees of California grew with a vigor unknown for centuries; the most terrible of recorded famines occurred in England and India; the Eskimos were probably driven south by increasing snowiness in Greenland; and the Mayas of Yucatan appear to have made their last weak attempt at a revival of civilization under the stimulus of greater storminess and less constant rainfall. The second bit of evidence is found in recent exhaustive studies of periodicities by Turner[112] and other astronomers. They have sought every possible natural occurrence for which a numerical record is available for a long period. The most valuable records appear to be those of tree growth, Nile floods, Chinese earthquakes, and sunspots. Turner reaches the conclusion that all four types of phenomena show the same periodicity, namely, cycles with an average length of about 260 to 280 years. He suggests that if this is true, the cycles in tree growth and in floods, both of which are climatic, are probably due to a non-terrestrial cause. The fact that the sunspots show similar cycles suggests that the sun's variations are the cause.