J. P. Hall— Short Cycle in Weather. 237 



15, 1889 (G, fifth curve, Series I), the cause seems to be 

 inadequate to the effect; and again, as on Jan. 10, 1892 (E, 

 first curve, Series II), the results are not quite as great as 

 possibly one might expect. In the former case, however, with 

 only a moderate high and a shallow low, there was a great 

 intensification of the storm after it passed out on the Atlantic, 

 beyond the range of the -weather maps, but while it was still 

 near enough to affect the coast by increasing the gradient and 

 wind force ; and in the latter, with an exceptionally large anti- 

 cyclone over ]S"ew Brunswick, the fall would undoubtedly 

 have been greater had not the temperature already been 

 greatly lowered by a previous high pressure area, and by a 

 storm going out to sea south of New York City. 



So many of the apparent exceptions to the rale may be 

 reasonably explained away, that one is justified in believing 

 that the few remaining ones would disappear with more care- 

 ful analysis, or in the light of conditions on land and sea 

 contiguous to the United States but not represented on the 

 ordinary charts. There is strong reason, therefore, to believe 

 that, in the main, if not entirely, the temperature phenomena 

 here exhibited are directly related to the distribution of air 

 pressure, and that the atmosphere, owing either to the in- 

 creased operation of the cause or causes (whatever they may 

 be) which originate storms, or to the supplementary influence of 

 some other cause, is persuaded periodically to arrange itself in 

 high and low pressure areas of more than usual intensity. 



The fact that the periodicity observed in the United States 

 has also been noticed in Europe and in the Arctic regions 

 leads one to suspect that the exciting cause is cosmical and not 

 terrestrial. The 27-day period imperfectly revealed in the 

 weather corresponds nearly to that of the sun's rotation. 

 There are tremendous disturbances on that orb which appear 

 to us as spots and prominences. And a relation between these 

 solar storms and auroras and other phenomena in terrestrial 

 magnetism has long been believed to exist. So that it is 

 natural to seek for a connection between important meteor- 

 ological episodes and the rea23pearance, by the sun's revolution 

 upon its axis, of spots, faculse, prominences, or something else 

 even more permanently located upon that immense globe. It 

 is not surprising to find, therefore, that nearly 20 years ago 

 Broun suspected the existence of certain meridians* on the 

 sun which might be particularly potent in exciting auroras and 

 magnetic storms. Buys Ballot, it will be remembered, re- 

 ferred the fluctuations which he observed to " heat holes'' or 

 areas of higher or lower temperature on the solar surface. In 

 line with these ideas was the belief of Prof. Spoerer that cer- 



* Encyclopaedia Britannica; article on Terrestrial Magnetism, Section 86. 



