456 



NATURE 



[Sept. 27, 1877 



Table II. — Cycle of Rain/all and Sun-spols shown in the 

 Minim-um, Intermediate, and Maximum Groups. 



I regret that I have not the materials for showinor the rainfall 

 for Bombay and the Cape during the whole sixty-four years for 

 which the returns exist for Madras. 



The other point on which I venture to trouble you at present 

 has reference to a different class of atmospheric phenomena. I 

 think that it may now be affirmed that a cycle of wind- 

 disturbances exists and is coincident with, although slightly 

 lagging behind, the cycle of sun-spots. M. Pocy called the 

 attention of the French Acadcmie des Sciences to this subject 

 five years ago, and published, as far back as 1873, a list of hurri- 

 canes in the West Indies from 1750 to 1873 in support of his 

 views. Dr. Meldrum has worked out the s.ime question, as 

 regards the [East] Indi.-xn Ocean, with admirable patience and 

 success. On the publication of my cycle of Madras rainfall it 

 struck Mr. Henry Jeula, Honorary Secretary to the late 

 Statistical Committee at Lloyd's, that the subject might have 

 a practical bearing upon under-writing and marine risks. He 

 collected from Lloyd's Loss Book the statistics for the two last 

 eleven-year cycles (1876-66, and 1865-55), the only ones for 

 which materials were available ; and conjointly, we have 

 tabulated the results. We found that the percentage of losses 

 on registered vessels of the United Kingdom was 17A per cent, 

 greater during the two years of maximum sun-spots (the eleventh 

 and first of the cycles) than during the two years of minimum 

 sun-spots (the fifth and sixth of the cycles). In the same way we 

 found ihat the percentage of the total losses (calculated on the 

 eleven years) posted on Llo_)(rs Loss Book was 15 per cent, 

 greater during the two years of maximum than during the two 

 years of minimum sun-spots. We further found that the increase 

 and decrease of losses follows a cycle, closely following (although 

 for sufficient reasons somewhat lagging behind) the cycle of sun- 

 spots. These results can be tested from the succeeding tables. 



Table IV. — Percentage of Losses posted on Lloyd's Loss-Book, 

 compared with the Eleven- Year Cycles of Sun-stots and of the 

 Rainfall at Aladras. 



Again, testing the cyclic coincidence, by taking the first four 

 years, the middle or maximum four years, and the remaining 

 three years at the end, a similar result is obtained. 



Table V. — Percentage of Losses posted on Lloyd's Loss Boot, 

 compared with the Eleven-Year Cycles of Sun-spots andofthe 

 Rainfall at Madras. 



Table III.- — Percentage of Losses posted 

 covipaycd with the Eleven- Year Cycla 

 the Rainfall at Madias. 



I Lloyd's Loss-Book, 

 of Sun-spots and of 



I think, therefore, that we are justified in concluding that the 

 periodicity observed by M. Poey in the hurricanes of the 

 Antilles, and by Dr. Meldrum in the cyclones of the Bay of 

 Bengal is of such a character as to exercise a widespread effect 

 upon the commerce of the world. How far these wind-dis- 

 turbances may be eventually proved to be general throughout 

 the earth's atmosphere, or throughout any given belt of it, I aai 

 not yet prepared to offer an opinion. But that the practical 

 results of such wind-disturbances on maritime commerce are of 

 a general character the foregoing tables now place beyond 

 dispute. 



In conclusion, I beg to caution fellow-workers that no really 

 trustworthy results are to be obtained from the old plan of 

 jumbling together a number of unhomogeneous stations in a bag 

 and shaking out averages. The true method is to take certain 

 recognised factors in the rainfall, such as the monsoons, and to 

 examine whether any common periodicity exists between the 

 operation ol these factors and the sunspots. This is what I have 

 attempted to do for various pomts around the great basin which 

 stretches southwards from the Bay of Bengal, and what Mr. 

 Archibald has so carefully done in NATURE for Northern India. 

 I am now conducting a similar inquiry into the American and 

 Australian rainfall, but, as already stated, sometime must elapse 

 before the results can be presented. W. VV. HtJNTER 



AUanton House, Lanarkshire, September 20 



■n- -J- ,11 1 .1. t , •. , ' The Discovery of the Satellites of Mars 



Dividing the eleven years as nearly as the number admits by / 



three, into a minimum, an intermediate, and a maximum group, ( As some of the earlier newspaper accounts of Prof. Hall's 



we get the following results : — , discovery of satellites of Mars are said to have produced, in some 



