304 DISPUTED EFFECTS OP FORESTS 



establishing the statistics of storms, it is because iu the forests there 

 are neither notable devastations committed by them nor inhabi- 

 tants to report them. 



" The department of the Gironde furnishes a striking case in 

 support of this explanation. According to the storm chart of this 

 department (Atlas meteor ologique, 1868) the hail would seem to fall 

 by preference on the banks of the Garonne, or in the rich canton 

 which extends between that river and the Dordogne, but particularly 

 on the inhabited places; and, on the contrary, that the desert landes 

 were entirely spared by the hail. 



"But it is well known that the landes are traversed by violent 

 storms ; but the shepherds of the landes do not send bulletins to the 

 Observatory !" 



So much in regard to storm-charts such as are cited by M. Renou. 

 In regard to the kind of storms referred to by him, they are such as 

 were not likely to deviate from the cause in consequence of any such 

 difference on the surface of the earth as a forest might present. 



Of the hailstorm cited by M. Renou, Sir John Herschel writes ; — 

 "In the hail-storm of July 13th, 1788, which passed across France 

 from south to north, two such tracks were marked, of 175 and 200 

 leagues in length respectively, parallel to each other, the one four 

 leao-ues broad, the other two, and separated by a track five leagues 

 in breadth, in which only rain fell. A similar character is very com- 

 mon though not to such an extent. Such linear hail-storms are 

 always attended with violent wind, sudden depression of the baro- 

 meter, indicating a great commotion in the air, and probable mingling 

 of saturated masses of very different temperature. 



And in writing on the Hydrology of South Africa, I had occasion to 

 state : " According to views advanced by Espy, a distinguished 

 American meteorologist, the two parallel lines along which the hail- 

 storm advanced in the case mentioned by Sir John Herschel can be 

 satisfactorily accounted for. According to one of the laws regulating 

 storms, many of these advance, as does a wheel, rotating and pro- 

 crressino- horizontally — as a wheel does vertically in advancing along 



a road as is made visible in a small advancing whirlwind of dust, or 



leaves, or straw, or chaff. The cyclone, a whirlwind of such extent 

 as to create a storm, is fed with air as it advances; this, raised to a 

 gi-eat height, expands, cools, precipitates the moisture it contained, 

 and throws this off in a frozen state — as does a twirled wet mop 

 throw off drops of water, — and by gravitation it falls. As the rotaiing 

 ■whirlwind advances, the motion of the air forming the forefront will 



