METEOROLOGY. 357 



farorablo circumstances given continuously since 1870 to the tri dully 

 weather niai)s of the United States and Canada should not louf? ago 

 have been supplemented by some general summary oi' the laws of storm 

 progress and weather changes with which the officers of the Signal 

 Service have become familiar. Guided by the writings of Red held, 

 Espy, Henry, and Ferrel, the deductive branch of meteorology was 

 making rapid progress — greater than could possibly have been fol- 

 lowed by those outside of the office. It was already in 1872 a matter 

 of common note that storms and also areas of high pressure, &c., pur- 

 sue successively the same or very similar paths, each one showing a 

 progressive systematic difference from its predecessor, until by a sud- 

 den change in distant surrounding circumstances the whole system was 

 broken up and a new order of things inaugurated. These similarities 

 between storm paths were frequently pointed out, both in the daily 

 weather predictions and the Monthly Weather Review, An interest- 

 ing illustration of this will be found in the daily predictions of the 

 hurricanes of August 18 and 25, 1871, as made by the present writer for 

 the official weather "Synopsis and probabilities" of the Army Signal 

 Office.] 



Van Bebber continues by saying that only in the rarest cases are press- 

 ure and temperature distributed around two storm-centers in the same 

 manner, and to this circumstance principally is it to be attributed that 

 the progress of and changes in the depressions show such extraordinary 

 variety. If the distribution of pressure and temperature is reversed, 

 then will the movement of the depression be hindered or entirely an- 

 nulled, and at the same time it itself takes a long irregular form [the 

 barometric trough of the Signal Service Bulletinsj, its longer axis per- 

 pendicular to the pressure gradients, and from its ends frequent small 

 minima break off that then follow the general current of air prevailing 

 in their region; if, however, on either side pressure or temperature pre- 

 vail, theu the direction of the movement of the storm-ceuter will be 

 thereby determined. 



From this short exi)lanation we see the great importance of these two 

 principles in their application to weather predictions. If, however, we 

 would form a correct prediction, we must extend our weather map to the 

 greatest possible extent, especially toward the west, and study expressly 

 the behavior of the great barometric maxima and minima that charac- 

 terize certain regions of the earth. We see also the importance of the 

 study of the cloud movements, especiallj- the ujjper clouds. {Z. 0. G. 

 M., XVIII, p. 447.) 



322. Teisserenc de Bort having studied the general weather condi- 

 tions attending abnormal winters, further develops his generalizations 

 relative to the " principal centers of action of the atmosphere," and 

 shows that the perturbations in the position and intensity of these at- 

 mospheric centers corresi>ond witii imi)ortaiit changes in the character 

 of the weather. If, therefore, wi- v oiild prcvlict tli'' weatlur ior a long 



