Decemiiek 14, 18S3.I 



SCIENCE. 



759 



loooo mn ^- 



variable upper currents ; the supply- of wnrm, 

 moist nir, and consequent occurrence of heavy 

 rain ; the relative strength of the iublowing 

 winds ; and a certain ettect of the earth's 

 rotation. All these causes of progression are 

 variable in amount, and in relation to one an- 

 other ; and it is therefore natural to find their 

 resultant inconstant. 



The first-named cause is the most evident, 

 the most powerful, and was the first recog- 

 nized. The general or planetary circulation 

 of the winds will require that any disturbance 

 in the moving atmosphere shall partake of its 

 motion, and bo carried along in the direction 

 of the current within which it is generated. 

 Thus a storm arising in the equatorial calms 

 is carried westward as soon as it attains suf- 

 ficient height to reach the upper current, which 

 must there move from east to west. No equa- 

 torial cjclonc lias ever been observed moving 

 eastward. On approaching the western shores 

 of the ocean, a part, at least, of the general 

 winds, turns toward the poles, as may be seen 

 on any wind-chart, and in 

 latitude 2.3° or 30° passes 

 from the region of the 

 tropical winds into the 

 system of the prevailing 

 westerl_v winds of temper- 

 ate latitudes. The storms 

 have a strikinglj- similar 

 course, and, on the western 

 side of the oceans in these 

 latitudes, never move to- 

 wards the equator. Their 

 further progress, and that of the many storms 

 of the temperate zones, is easterly, with a lean- 

 ing towards the pole while crossing the oceans, 

 and a variable north-easterlj- or south-easterly 

 advance on the continents. No storm has 

 been found crossing the North Atlantic from 

 east to west, or moving from our Atlantic 

 coast to the plains beyond the jMississippi. 

 Additional evidence of this style of bodily 

 transference of storms will be given in con- 

 sidering the relative strength and the direction 

 of their spiral winds on different sides of the 

 centre. 



The importance of the condensation of vapor 

 and consequent rainfall in decreasing the cool- 

 ing of the central up-drauglit, and so increasing 

 its strength, has already been shown. In the 

 explanation of tliis process, it was tacitl}- as- 

 sumed that all the surface-indraught was equally 

 warm and moist, so that condensation and 

 rain would occur symmetricall3- about the cen- 

 tre of low pressure. It will now be seen, that, 

 when a storm-centre is supplied from areas 



of unequal warmth and moisture, symmetrical 

 cloud-forming and I'ain-falling on all sides will 

 be impossible ; there will bo more r.ain, and 

 hence less cooling, on one side than on the 

 other ; and just as the liberation of ' latent 

 heat' aided in the formation of the first cen- 

 tral barometric depression, so it will now tend 

 to displace this centre to the side where the 

 greatest amount of rain falls. If no other 

 cause but this acted, the storm would advance 

 regularly toward the region of heaviest pre- 

 cipitation : but this advance will not be like 

 the bodily transference of the rotating winds 

 ctfected by the general atmospheric currents ; 

 it will be rather the abandoning of one cen- 

 tre of attraction as a stronger one is cre- 

 ated beside it, — the continual filling-up of 

 one depression, and production of another. 

 This may be illustrated by a modification of 

 fig. 8, given here in fig. 12, in which the dotted 

 lines show the gradients and winds established 

 at a certain peiiod of the storm. Let it be 

 supposed that warmer, moister winds enter 



on the right, and cooler, drier winds, on the 

 left. Where cooler, the air will be contracted, 

 and the isobaric surfaces depressed : where 

 warmer, from its own warmth, as well as from 

 that of the condensing vapor, the air will lie 

 expanded, and the isobars elevated, as shown 

 in full lines in the figure. The gradients will 

 then be unsyrametrical about the original cen- 

 tre ; and the previous motion of the winds will 

 be accelerated at some points, retarded or re- 

 versed at others. As a result, the pressures 

 at the surface will be changed from their pre- 

 vious arrangement to a new one, shown in fig. 

 1.3, in which the region of least pressure has 

 moved to the side of the warmer winds and 

 heavier rains. Any further inflow of the sur- 

 rounding air must now be to the new low- 

 pressure centre : in other words, the storm has 

 advanced to the right. The process will be 

 continuous as long as the winds on opposite 

 sides of the storm are unlike.' Having thus 



* ¥\g. 12 may serve further to explain the retarded arrival of 

 the centre of low prcsBurc at altitudes of a niite or more above 



