Feb., 1905.] 



KNOWLEDGE & SCIENTIFIC NEWS. 



35 



call it the Prime." Again, in Australia there was 

 an impression as long ago as the year 1836 that the 

 seasons underwent a variation every nine or ten years, 

 varying, however, every third series or thirty years. 



Now Bruckner has shown that there is a thirty-five 

 year period both in pressure and rainfall, the years of 

 high pressures corresponding to those of less rainfall. 

 This long-period variation is of very great importance 

 and must be reckoned with in long-period forecasting, 

 although it is not so prominent as the short-period 



Lirves to 

 variation ot pres 

 type) and the mea 

 tics, over the wl 



istrate the close relationship between the 

 i over the British Isles (Oxford taken as the 

 ainfall, from the Meteorological Office 5tatis- 

 ; of Great Britain. 



changes that have been described above. The accom- 

 panying set of curves (fig. 12) will perhaps serve to 

 illustrate this long-period variation of rainfall for a few 

 stations on the earth's surface, while the curve at the top 

 indicates the relationship between the epochs of the dry 

 and wet periods and those of the great solar variation of 

 thirty-five years described in a previous paragraph. It 

 must not be forgotten that to determine this long-period 

 variation from the rainfall records the means of several 

 yeari have to be taken together, and even when every 

 five-year values have been employed the resulting curves 

 have to be again " smoothed," as it is called. In some 

 of the cases it will be seen that the five-year means 

 render apparent the eleven-year variation, a variation 

 which seems more distinct in tropical regions, such as 

 India, than extra-tropical regions. 



A glance at the curves is sufficient to indicate the 

 existence of these variations. From a study of these varia- 

 tions at many places scattered over the earth's surface, it 

 has been found that the maxima or minima do not occur 

 at the same epochs at all places ; at present this question 

 has not been worked out, but it may possibly turn out 

 that, like the short-period pressure variation, there is a 

 give and take between two large regions on the earth, in 

 which while the maximum rainfall is occurring in one 

 region the minimum is taking place in the other. 



Fortunately for us who dwell in Western Europe, it 

 will be seen from the curves that we are entering on a 

 series of years, which, on the average, will be wet, after 

 having just experienced a number of years during which 

 the rainfall was very much below normal. The rainfall 

 of 1903 practically put an end to this long drought. It is 

 important to remember that the short period of about 

 four years is the most prominent variation of rainfall, and 

 is always at work. It is thus quite possible to have a 

 comparatively dry year when the long-period rainfall 

 variation is at a maximum, but on the average the wet 

 years will he wetter and the dry years less dry at such an 

 epoch. At the minimum of the long-period cycle the wet 

 years will be less wet and the dry years more dry. 



Enough, perhaps, has been said to show that the rain- 

 fall variations all depend on the atmospheric pressure 



changes that occur. These latter are apparently closely 

 associated with the solar cycles whether they be indi- 

 cated by spots or prominences. We are thus led to 

 deduce the most probable—and, after all, the most 

 natural —conclusion that the sun is the most important 

 factor in producing our varied weather. 



We have become acquainted with three periodic varia- 

 tions of solar activity, covering about four, eleven, and 

 thirty-five years each. The question arises — .Vre there 

 any other variations of longer period which may help to 

 complicate the solar problem, and, consequently, the 

 meteorological one as well ? 



It may be said, however, that no other periodic varia- 

 tion extending over a year has yet been traced, as the time 

 over which the observations extend is at present too short. 



The lengths of the periods which have up to now been 

 discovered have, however, such a peculiar relationship to 

 each other that perhaps a means is afforded of suggesting 

 a fourth period. If we take the length of the shortest 

 period as our unit — namely, 3-8, and multiply it by 3, we 

 obtain 11-4, which is very near our second period, which 

 is 11-3; if we again multiply 11-4 by 3, we have 34-2, 

 which again is close to 34-8, the real value, as far as can 

 yet be determined, of the thirty-five period. Now, if we 

 multiply 34-2 by 3 again, we obtain 102-6, which may be 

 the length of a new period. The above numbers, put in 



Fig. 12. Some curves showing the Iong=period variations of rainfall 

 and their relation to the 35=year solar chan.:e as indicated by the 

 vertical dotted lines, and the eleven=year solar period. 



tabular form, show the curious relationship between the 

 periods perhaps better : — 



3-8 X I = 3'8, value actually determined y8 

 3-8x3= 11-4 „ „ „ 11-3 



3-8x6= 34-3 „ „ „ 34-8 



3-8 X 9 = 102-6 „ ,, ,, ? 



To advance the knowledge of weather changes, solar 

 variations must be most carefully watched. For a suc- 

 cessful solution of the weather problem, the two sciences. 

 Solar Physics and Meteorology, must go hand in hand, 

 and the saying "unity is strength" no less applies to 

 matters scientific than it does to o'.her a.Talri of lii'e. 



