PHYSICS OF THE GLOBE. 123 



To substantiate these questions, observations of other cities 

 as Washington, Philadelphia, Providence, and Paris have 

 been selected ; they all tend to show the same results. Those 

 of Paris, extending back 190 years, give a slight increase in 

 the rainfall not steadily, but with three oscillations of many 

 years each. 



This report is a continuation of a former one on, Does the 

 Clearing: of Land Increase or Diminish the Rainfall? From 

 the tables produced, it appears that the widespread public 

 impression that the clearing of land diminishes the volume 

 of rain is not founded on fact. 



Mr. Jerome J. Collins has published in Nature, for May, 

 1878, an essay on his American Storm-warnings, which will 

 attract the attention of all, and especially of European me- 

 teorologists. He also contributes the folio wins: brief sum- 

 mary of the meteorological work accomplished in 1878 by 

 the New York Herald: 



From November 15, 1877, to the corresponding date in 

 November, 1878, 51 warnings were cabled to London, and 

 distributed by telegraph to the principal commercial centres 

 of Europe. Of these 2 were sent after the 15th of Novem- 

 ber, 1877; 3 in December; 4 in January, 1878; 5 in Febru- 

 ary ; 7 in March ; 4 in April ; 3 in May ; 3 in June ; 3 in 

 Jul}'- ; 4 in August ; 4 in September ; 6 in October ; and 3 

 to the 15th of November last. Of the total number 51 34 

 have been completely fulfilled, 13 partly fulfilled, and 4 ful- 

 filled only as to some one of the conditions predicted. The 

 percentage of complete fulfilments is 66, and of complete 

 and partial fulfilments, 92. An analysis of the predictions 

 as to date of arrival, barometer, etc., gives the following 

 percentages : Date of arrival of storm-centre or depression, 

 84.5; barometer, 87.8; wind-force, 89.1; wind direction, 87.8; 

 weather, 98.3 ; locality of arrival or regions immediately 

 affected, 91.0. Of the 51 warnings cabled, several referred 

 to more than one storm or depression then in movement; 

 but 23 of the warnings referred to "storm-centres," or "cy- 

 clones," 27 to " depressions," and 1 to a "disturbance." The 

 maximum in fulfilments was reached during November, 

 1877, and January, August, September, October, and No- 

 vember, 1878. The next best months were December, 1877, 

 and February and March, 1878. The least success was expe- 



