SCIENCE. 



[Vol. XI. No. 257 



be of considerable magnitude, sufficient to be expressed in inches, 

 annually. 



In the prairie region I find twenty-four stations at which extended 

 series of rainfall measurements have been made. None which have 

 been used are less than ten years in duration, and they range thence 

 up to forty years. The sum of all these series is four hundred and 

 twenty-eight years. Each of these series was divided into two 

 equal parts, and the total rainfall of each half obtained. On the 

 theory of a progressive increase, the sum of the second half of the 

 series should be greater than that of the first half. The following 

 table exhibits the result. The first column gives the names of the 

 stations ; the second, the length of the series ; the third, the total 

 rainfall in the first half of each series ; the fourth, the same for the 

 second half ; and the fifth column, the differences between them, an 

 increase having the plus-sign, a decrease the minus-sign. 



Prairie Region. 



Chicago 



Athens 



Augusta 



Dubois 



Galesburg 



Manchester. . . 



Marengo 



Ottawa 



Peoria 



Riley 



Sandwich 



Winnebago . . . 



Wyanet 



Springfield . . . 



Dubuque 



Omaha 



Leavenworth . 

 Davenport . . 

 La Crosse . . . 

 Milwaukee . . . 

 Brookside . . . 

 Fort Madison 

 Iowa City .. 

 Muscatine. .. 



Aggregate Rainfall. 



ist half. 2d half. 



367 

 463 

 418 



569 

 306 



+72 

 —33 



—35 

 —17 



+59 

 +24 

 +83 



+46 



+18 



The results, as will be seen, have a wide range, some stations 

 showing an increase, while much the larger number show a de- 

 crease. Now, Although these series overlap one another in all sorts 

 of ways, and do not necessarily refer to the same years, still, under 

 the theory of a progressive change, they may be combined directly 

 without involving error. We may add up columns i and 2 and 

 strike a balance, and this balance shows a greater rainfall in the 

 first period by 343 inches. Dividing this by the numberof years in 

 the period, 217, it is discovered that on an average each station re- 

 ceived per year 1. 58 inches more rain during the first period than 

 during the second : in other words, instead of an increase of rain- 

 fall being produced by the increase of arborescent vegetation, there 

 has occurred, from some cause, an actual diminution. I should be 

 very slow to argue from this a deleterious action flowing from the 

 increase of forests, but it seems to militate very strongly against a 

 favorable action upon rainfall. 



In Ohio the contrary result is to be sought ; viz., a decrease in 

 rainfall owing to the destruction of forests. In this State I find 

 twelve stations, with series ranging from ten to forty-eight years 

 each, and an aggregate of two hundred and ninety-four years. The 



observations have been treated as were those in the prairie region, 

 with results as given in the following table : — 



Ohio. 



Stations. 



Years. 



Aggregate Rainfall- 





ist half. 



2d half. 







46 

 28 

 14 

 16 



48 

 26 

 40 



18 

 '4 



J044 

 513 

 329 

 329 

 203 

 166 



loos 

 475 

 807 

 412 

 353 

 27s 



96s 

 53' 

 319 

 307 

 241 

 159 

 1033 

 547 

 836 

 364 

 333 

 545 



—79 

 +18 



Cleveland 



College Hill 





+38 



Kelley's Island 





Marietta 



+28 



+72 





+=9 

 -48 



Toledo 



Urbana 



Waterville . 



—30 





It will be seen that in this case the total rainfall of the first half 

 of the series is slightly greater than that of the second half, the 

 difference being 31 inches, which, divided by the number of years 

 in the first half of the series, shows, that, along with the clearing of 

 the forests, the rainfall has diminished a trifling amount, being 0.21 

 of an inch less in each year of the second period than the first. It 

 is, of course, unnecessary to add that this change is too small to 

 have any meaning. 



In the third area, that of southern New England, there is to be 

 expected a diminution of the rainfall, consequent upon deforesting, 

 which was in progress down to, say, i860, and, in more recent 

 times, an increase due to reforesting. Prior to i860, I have eigh- 

 teen series, ranging in length from ten to forty-six years, with an 

 aggregate of four hundred years. Treated as before, the results 

 shown in the following table are obtained. Summed up, they show 

 that the aggregate rainfall in the second period was greater by 579 

 inches, or 2.9 inches in each year of the period. Deforesting, in 

 this case, seems to be accompanied by a decided increase in rain- 

 fall. 



New England. 



Amherst 



New Haven . 

 Boston . — . . 

 Cambridge . . . 



Lowell 



Lunenburgh .. 

 New Bedford 

 Waltham . . . . 

 Worcester... . 

 Fort Adams.. 

 Providence. . 



Flatbush 



Albany 



New York.. . . 

 West Point. . . 

 Brunswick . . . 

 Gardiner 



Aggregate Rainfall. 



ist half. 2d half, 



506 

 456 



267 

 493 

 978 



Differences, 



434 

 +56 

 +7 

 +S> 

 —20 



6'3 

 555 

 528 

 413 

 246 

 486 

 748 

 456 



