DYNAMIC METEOROLOGY. 417 



of cold nortberiy wiuds iudicates the existence, of an unseen low ba- 

 rometer and severe storm to the southwest or soutlieast, toward which 

 these winds are blowing. . Perhaps the most useful rule in this resi)ect 

 is founded on tlie motions of clouds. The following generalization 

 (announced by ine in February 18713 as resulting from the study of up- 

 per and lower clouds and winds displayed on every Signal Service map 

 and quite independent of simultaneous European studies), will, it is be- 

 lieved, be found lo hold good for the whole northern hemisi)here: " The 

 general direction towards which any stratum of air is moving neaily hori- 

 zontally, is somewhat to the right hand of the direction of motion of the 

 stratum just above it, that is to say, the direction of the lower clouds 

 is to the right of that of the wind at the earth's surface; the higher 

 clouds move to IhB right of the lower clouds and the cirrus again to the 

 right of the latter ; these directions of motion may be so related that 

 the cirrus shall ))e moving in a direction nearly opposite to that of the 

 surface winds or the lower clouds. In general, when the lower currents 

 are flowing in toward a storm center the upper currents are flowing out 

 from it; or agani, when the lower currents are flowing out from an 

 area of high prt^ssure and clear, dry weather the upper currents are 

 flowing in towards it.'' It follows that when an outlying station reports 

 the motions of the cirrus cloud, we are able to infer something as to the ex- 

 istence of the area of the low barometer from which the cirri are coming. 



(I) Prediction of general storms. — When we have thus approximately 

 estimated the probable location and amount of the disturbances in den- 

 sity over all [)()rtions of the United States and outlying regions, Ave are 

 then to decide whether a storm already initiated will increase or di- 

 minish in intensity, and in what direction its center will move. 



It may be accepted as a fundamental principle that a given disturb- 

 ance will sprea(i in the direction of those regions where i)rocesses 

 are going on that favor similar disturbatices. Thus, if a mass of heavy 

 air, like the northers of the Mississippi Valley, is flowing in any direc- 

 tion under a mass of abnormally light air, it will tend to move always 

 in the direction where the lightest air is to be found, or where the di- 

 urnal changes of heat and evaporation favor the development of ab- 

 normally light a;r. Or again, if a storm is already in existence and 

 is nuiintained by the condensation of moisture in the cloud region, it 

 will be continuously moving toward the region where the coiulensation 

 is going on most actively. Tiie regions where storms are being initi- 

 ated or most liberally sujjported are generally detected (1) by observed 

 abnormal temperature and moisture ; (2) by the formation of haze, halos, 

 cloud, and rain or snow ; and (3) by a slight fall in the barometer due 

 to movements ii: the atmosphere at the level of the cloiul stratum. 

 The future movements of storm centers can generally be detected by 

 the observed changes going on in the direction of movement of the wind 

 or clouds. All of these meteorological elements must be examined by 

 careful comparisons of the preceding tri-daily maps in order to ascer- 

 U. Mis. U2 27 



