6o 



NATURE 



{Nov. 22, 1877 



which correspond closely in their maxima and minima periods 

 with those of sun-spots. {2) A more comprehensive survey of 

 the European rainfall has so far failed to establish this 

 correspondence. Dr. Jelinek's examination of fourteen sta- 

 tions, from 1833 to 1869, showed that the coincidence held 

 good in fifty-two cases, bat failed in forty-two. While 

 frankly accepting this as evidence against a real coinci- 

 dence, it should be remembered that a general law such as 

 a common periodicity in sun-spot activity and terrestrial rainfall 

 will be subjected to, and sometimes overruled by, the local 

 surroundings of individual stations. (3) On the other side, 

 Gustav Wex, from the recorded depths of ttie El^ie, Rhine, Oder, 

 Danube, and Vistula, for six sun-spot cycles (1800-1867), found 

 that the maximum amount of water occurred during periods of 

 maximum sun-spots, while the minimum levels were reached in 

 the periods of minimum sun-spots. The evidence, as regards 

 Europe, is, therefore, conflicting ; and it is safer for the present 

 to reckon it as against a well-marked common periodicity. I 

 hope at no distant date to submit the results of a new and more 

 exhaustive examination of the European rain- register?. 



I now proceed to the North American rainfall. Here, as in 

 Europe, the question is complicated not merely by disturbing me- 

 teorological influences, such as the Gulf Stream, but by the uncer- 

 tain value of the rain-returns. These are causes which even at a 

 carefully supervised station render it difficult to estimate the 

 number of inches yielded by long-continued or very violent snow- 

 storms. At badly supervised stations, or in the case of private 

 gauges where the supervision is apt to be of a still more hap- 

 hazard character, these difficulties often suffice to render the 

 returns quite worthless. Yet it is the latter class of records on 

 which we have chiefly to depend in an attempt to deal with the 

 American rainfall during a long series of years. Nowhere does 

 meteorology now receive more careful and scientific study than 

 in the Western Continent, but in many of the most valuable 

 series the element of time is still necessarily wanting. The 

 tvidence hitherto received from America has, on the whole, 

 been favourable to the existence of a common periodicity. Mr. 

 Dawson, Geologist to the British North American Boundary 

 Commission, found a correspondence, although by no means an 

 absolute one, between the fluctuations of the great lakes and the 

 sun-spot periods. This question has been lately revived and 

 interpreted afresh by a distinguished meteorological observer in 

 Northern India. Prof. Brocklesby's contributions to the Ame- 

 rican Journal of Science also point to a connection between 

 variations in the sun-spot area and annual rainfall. 



It was with a knowledge of these statements that I undertook 

 a systematic inquiry into the American rain-returns. I ought at 

 once to say that the result of that inquiry altogether fails to 

 establish the existence of a common cycle, so far as concerns the 

 temperate zone. I divided the American stations into four 

 groups. The first group consisted of eleven stations in east 

 coast or Atlantic States, lyinjj between 40° and 45° N. latitude. 

 The second group consisted of seven stations in Inland States, 

 from 38° to 48°. The third group was intended to consist of 

 s'ations in the West Coast or Pacific States, but I have obtained 

 the returns (and those for a period altogether too brief) for only 

 a single West Coast Station, San Francisco. I give them, how- 

 ever, for what they are worth. The fourth group consists of 

 three coast-stations in the Southern States, between 30° and 

 33°; or just above the sub-tropical region with which Mr. 

 Archibald's returns for the Bengal stations deal. 



The results of the examination of the four American groups 

 may be summarised thus : (i) Taken as a whole, the returns from 

 the twenty-two stations do not exhibit any common periodicity 

 between the rainfall and the sun-spots ; nor do they disclose an 

 eleven year's cycle corresponding to the one which I have shown 

 to exist in the rainfall (at Madras and elsewhere) gathered from 

 the Indian Ocean. (2) That as regards the three northern 

 groups, stretching across the continent from 38* to 48' N, lat., 

 the rainfall, so far as any symptoms of periodicity can be detected 

 at all, tends to vary inversely with the sun-spois ; but that it is 

 impossible to discover any real periodicity whatever. (3) On the 

 other hand, that as regards the southern group, between 30° and 

 33°, there are symptoms of a periodicity tending to coinciae with 

 the sun-spot variations ; but that these symptoms are not suffi- 

 ciently uniform in the small number of southern stations which I 

 have examined, to justify any conclusion. 



The calculations on which these results are based would 

 occupy many pages, but their general line may be indicated in a 

 few sentences. Thus the mean rainfall at the twenty-two stations 

 during the years of maximum sun-spots for which the records 



have been obtained, was 37J inches, while during the years of 

 minimum sun-spots it was 39. The years of maximum sun-spots, 

 together with the years immediately preceding, had a mean fall 

 at the twenty-two stations ot 40*2 inches ; wtiile the minimum 

 years of sun-spots, taken together with the years immediately 

 preceding, had an almost exactly equal rainfall of 40'i inches. 

 In the northernmost grouo of eleven /Vtlantic stations the mean 

 rainfall of the years of maximum sun-spots was 39 inches, agaiist 

 an average of 41 inches in years of minimam suti-spots ; in the 

 second group of seven inland stations (38" to 48") the mean rain- 

 fall of the years of maximum sun-spots was precisely equal to 

 that of years of minimam sun-spots, being 33I inches in both ; 

 in the third group, San Francisco, the mean rainfall years of 

 maximum sun-spots was 21 inches against 23^ inches in 

 minimum years ; in the fourth group of three southern stations 

 (30° to 33°) the returns for the minimum and maximum years are 

 broken ; but taking these years and the preceding ones together, 

 the mean rainfa'l of the years of maximum sun-spots with the 

 years immediately preceding was 51 inches, against 48 J inches 

 in the years of minimum sun-spots and immediately preceding 

 ones. 



The returns have also been examined by another method. I 

 have shown elsewhere that the rainfall at Midras, and other 

 stations around the Indian Ocean, follows a well-marked cycle of 

 eleven years, with a miximum, minimum, and intermediate 

 period, corresponding with the maximum, minimam, and inter- 

 mediate period of sun-spots. The American stations not only fail 

 to show such a correspondence, bat as regards the three northern 

 groups so far as any symptoms of periodicity exist, they point in 

 the opposite direction. The fourth or southern group of stations, 

 on the other hand, so far as they disclose a periodicity, tend to 

 coincide with the periodical variations in the sun-spots. The 

 following table will show this. The Madras rainfall in the 

 tropics discloses a cycle closely corresponding with the eleven 

 cycle of sun-spots ; speaking generally, the American rainfall 

 in the temperate zone discloses no such cycle ; but the southern 

 stations begin to furnish symptoms of such a cycle. 



Table of Madras and American Rainfall Compared with the 

 Eleven Years' Cycle of Sun-spots 



Note. — The sun-spot figures represent the relative numbers, reduced from 

 Wolf. I he rainfall is expressed in inches. The San Francisco returns deal 

 with only tweuty-one years, or not. quite two complete cycles; much too 

 short a period for any definite conclusion. 



The records of the twenty-two American stations extend over 

 brief periods compared with the Madras returns. Several of 

 them disclose breaks or gaps ; few of them have been kept with 

 the minute care bestowed by the professional astronomical staff^ 

 on the rain gauge at the Madras Observatory, and the value of 

 most of the eighteen northern ones is rendered in some degree 

 uncertain by snow-storms. It is probable, moreover, that better 

 and much more complete returns are available to American 

 meteorologists than I possess for the twenty-two stations which 



