760 



SCIENCE. 



[Vol. II., No. 45. 



seen the general action of this cause of motion, 

 it must now be applied more directly. There 

 are two causes of rain in a cj' clonic storm, — 

 one from the expansion and cooling of the moist 

 air as it enters the district of low pressure, and 

 rises in the central up-drau^ht ; the other from 

 the advance of the wind from a warmer into 

 a cooler region. The first of these will gen- 

 erallj' be nearlj' sj-mmetrical about the storm- 

 centre, and hence not productive of any 

 progressive motion : the second will as gen- 

 erally be unsymmetrical. In fig. 14, for the 

 northern hemisphere, the parallel lines repre- 

 sent normal cast and west isotherms, showing 

 the usual decrease of temperature to the north. 

 Of the several winds blowing inward to the 

 storm-centre, A and B, which advance almost 

 along the same isotherm, will not be seriously 

 changed in temperature hj their change of 

 place ; 0, which comes from a cooler to a 



who first, some fifteen years ago, called atten- 

 tion to the control of rain over storm-tracks. 

 It should be noted that the change in the 



warmer district, will consequently increase its 

 capacity' for moisture, and be a clear, cold, 

 drj'ing wind ; but D will be chilled, and must 

 produce heavy clouds and strong rain some- 

 where about the shaded part of the figure ; 

 and the storm-centre will then be transferred 

 toward the middle of this rainy district. 



Standing on the warm side of the storm, 

 the centre will appear to move nearly along the 

 isotherms to the right. Actual isotherms 

 seldom follow lines of latitude, and always 

 vary their position with the seasons, espe- 

 cially along continental borders. Thus, over 

 ■western Europe and the eastern margin of 

 the Atlantic, the summer isotherms run to the 

 north-east : so do the storms. In winter 

 the isotherms run south-eastward, and the 

 storms turn in the same direction. Figs. 15 

 and 16, illustrating this change, are based on 

 diagrams in the ' Laws of the winds,' by Le}', 



the surface. Observations on Mount Washington have shown the 

 centre of low pressure there to he about two hundred miles he- 

 hind that at sea level (Loomis),and a similar retardation has 

 been inferred in England from observations of cirrus-clouds 

 (Ley). Fig. 12 shows this to be directly connected with rain- 

 fall; for, in this unsymmetrical storm, the former horizontal 

 neutral plane is distorted, so that the centre of low pressure in 

 the upper air is clearly behind, instead of vertically above, the 

 centre on the surface of the earth. 



winter and summer prevalent winds would 

 have a similar effect on the courses of Eu- 

 ropean storms. In the United States, Pro- 

 fessor Loomis has shown that the velocity, as 

 well as the direction of advance, is closely 

 dependent on the position 

 and amount of the rain. 

 In tropical storms the ac- 

 tion of this cause of pro- 

 gression is not so clearly 

 marked ; for all the winds 

 are moist, and almost 

 equally warm. It is re- 

 ported that the rainy area 

 often extends farthest 

 ahead of the storm ; but 

 it is not at once apparent 

 why it should, for the front of the storm is oc- 

 cupied by winds from the north, which come 

 from a slightlj' cooler latitude. It maj-be sug- 

 gested, that, as their source in a region of high 

 pressure (the ' horse latitudes ') causes them to 

 move faster, it also, probably, allows them a 

 greater expansion and cooling, on entering the 

 storm-area, than is permitted in the w'inds that 

 come more slowly from the equatorial region of 

 low pressure ; but tropical storms probably de- 



'/■ 





J / 'RAlfM r 



\/ \ 



/ 



/ ;. 



pend chiefly on the prevalent winds for their 

 direction and rate of advance. In Austria none 

 of the winds are very moist, and the rainj' area 

 has no definite relation to the advance of the 



