In historic as well as prehistoric times, as we have already seen, there have been climatic fluctuations. For instance, the twelfth or thirteenth century B. C. appears to have been almost as mild as now, as does the seventh century B. C. On the other hand about 1000 B. C., at the time of Christ, and in the fourteenth century there were times of relative severity. Thus it appears that both on a large and on a small scale pulsations of climate are the rule. Any hypothesis of climatic changes must satisfy the periods of these pulsations. These conditions furnish a problem which makes difficulty for almost all hypotheses of climatic change. According to the present hypothesis, earth movements such as are discussed in Chapter XII may cooeperate with two astronomical factors. One is the constant change in the positions of the stars, a change which we have already called kaleidoscopic, and the other is the fact that a large proportion of the stars are double or multiple. When one star in a group approaches the sun closely enough to cause a great solar disturbance, numerous others may approach or recede and have a minor effect. Thus, whenever the sun is near groups of stars we should expect that the earth would show many minor climatic pulsations and stages which might or might not be connected with glaciation. The historic pulsations shown in the curve of tree growth in California, Fig. 4, are the sort of changes that would be expected if movements of the stars have an effect on the solar atmosphere.