30 



SCIENCE. 



[Vol. XVI. No. 389 



they are both produced by a common force, or that the spots 

 produce the effect upon the earth. The coincidences here, 

 however, are very marked, and there are practically no dis- 

 cordances. These points should be most carefully borne in 

 mind in all our studies in this line. 



Sun-Spots and Storms. 



Meldrum of Mauritius vras one of the first to study the re- 

 lation betveeen cyclones and sun-spots, and found that during 

 the three maximum periods of the spots between 1848 and 

 1871 there were nearly twice as many cyclones as during the 

 minimum periods. The evidence in 1871, however, was far 

 less than in 1848, and the more recent spot maximum of 

 1884 showed no increase in cyclones; so that this proof, 

 which has been the one that has been relied upon above all 

 others, has gradually dwindled down to practically nothing. 

 It should be noted, however, that we have no absolute proof 

 one way or the other; for during the spot maxima of 1871 

 and 1884 there may have been other forces acting which 

 tended to diminish the activity of cyclones, or to divert them 

 from the track of ships whose records Meldrum examined. 

 Poey examined the West India cyclones, going back to 1750, 

 and he thought that there were more cyclones during spot 

 maxima. This record is so long that I have deemed it worth 

 while to give it a careful study. The spot curve is remarka- 

 bly well defined, with very few irregularities. The maxi- 

 mum and minimum points are very easy to find, and one 

 can have no doubt as to the exact year of each period. The 

 cyclone curve, on the other hand, is exceedingly irregular, 

 and fluctuates back and forth across the spot curve. Com- 

 paring the cyclone and spot curves, I find that in the twen- 

 ty-four maximum and minimum periods there are eight co- 

 incidences, ten positive discordances, and six doubtful cases 

 (that is, cases which showed a flat curve for cyclones at the 

 time of either maximum or minimum spots). This is a very 

 poor showing, and certainly proves no intimate relation. 

 No one can compare these curves and consider that the rela- 

 lation is proved by them. 



If it could be shown that our cyclones were due to special 

 heat or electric action, and that sun-spots tended in the same 

 direction, there might be some hope in establishing a rela- 

 tion between them. Under the circumstances, however, it 

 is necessary either to obtain accordance or else to explain 

 away the discordances. Coincidences for a few periods are 

 very probable, and prove nothing. The weakness of this 

 cyclone research lies in the fact that only very limited por- 

 tions of the earth have been considered. The only proper 

 way would be to determine the extent of storm activity over 

 the whole globe each day of the year, and then to compare 

 this with sun-spot action. The reasons that many sun-spot- 

 ists have met with so much encouragement in their re- 

 searches have been two: 1. A "coaxing, "as Professor Young 

 puts it, of the critical point of a sun-spot period (that is, if 

 the cyclone maximum came two or three years earlier or 

 later than the spot maximum, it has been regarded as a co- 

 incidence) ; 2. A consideration of all manner of terrestrial 

 disturbances as fulfilling the prediction or establishing the 

 coincidence. All manner of tornadoes, storms, blizzards, 

 hot waves, cold waves, floods, frosts, earthquakes, etc., have 

 been drawn into their nets. It would seem that this prac- 

 tice should be regarded as very unreliable. If a spot pro- 



duces a cold wave one week or at one time, it can never 

 produce a hot wave at any other time. The determination 

 of such a relation, if there be one, is one of the simplest 

 mathematical processes that can be imagined, after the data 

 are at hand, and yet the sun-spotists are very anxious to 

 make their own computations. One of them writes, "I am 

 very much afraid you will coax the data to disprove my 

 view." It is very plain that no accurate research of this 

 kind can ever be made by any one that cannot be re- 

 peated by any other person, and the fear of sun-spotists to 

 have the verification of their theories taken out of their 

 hands is well grounded. 



Tornadoes and Sun-Spots. 



If we take such a very great territory as that of the 

 United States, and have stations at every sixty or a hundred 

 miles, then count the number of stations each dayat which the 

 wind velocity reached twenty miles or more per hour, we would 

 have a partial view of the average storm activity each day 

 of the year. Again: if we could get a record of every vio- 

 lent storm in this region, and give it a proper weight, we 

 would have a fair idea of storm activity, and could compare 

 it directly with the known and easily measured spot activity. 

 This has been done with the 2,221 tornadoes that have visited 

 this country. The following table shows the relative inten- 

 sity of tornadoes by weight, and the relative sun-spot inten- 

 sity:— 



In this table, numbers in the sun-spot column are taken 

 from the Greenwich photographs, and show the relative area 

 covered by spots in millionths of the sun's surface. The ear- 

 lier tornado records are in some doubt, as they are quite 

 meagre. The enormous increase in both tornadoes and sun- 

 spots during 1883 and 1884 is very striking, and seems to be a 

 fact, though it will require several more eleven-year periods to 

 establish the coincidence. There was an increased activity 

 in collecting tornado data in 1882, but this continued through 

 till 1887 ; so that the great increase in the two years above 

 cannot be ascribed to this cause. Moreover, the list of 2,221 

 tornadoes was made up chiefly from the same source 

 throughout all the years. There is a quite strong proof that 

 the sun-spots are due to the action of electricity. Now, it 

 has been shown that our tornadoes have an abundance of 

 electric action, so that there is no inherent improbability in 

 the supposition of a relation between these phenomena, aside 

 from a mere coincidence in their phases. 



Specific Influence of Spots. 

 If such a relation exists, some have thought that there 

 ought to be a synchronism between the appearance of spots 

 and resulting storms. The view has been strenuously sup- 



