550 SCIENTiriC RECORD FOR 1884. 



attenclinjif downward currents of air. the process from that time on may 

 become a self-sui)portiiig one, each feature continually increasing^ in in- 

 tensity until a maximum is reached, dei)endent on temperature, moist- 

 ure, and conflfjuration of the eaith's surface, when it attains a condi- 

 tion of equilibrium, and so continues until nig:htfall. Koppen's study 

 applies equally to every form of thunderstorm, tornado, and wind- 

 squall. The views developed by him in this essay are accompanied by 

 such a mass of exact data as to make it an invaluable contribution to 

 our knowledge of a subject that was however already largely eluci- 

 dated by Espy, Fourth jMeteorological Eeport, pp. 106, 1G4, 165, and 

 Prof. Joseph Henrj^ Lectures on Meteorology, Agricultural Reports, 1854 

 to 1859. In fact, we think it would not be an injustice to call the above 

 the Koppen-Espy theory of thunder-storms and squalls.] 



301. 0. Ferrari writes in reference to Lancaster, Koi)pen, and von 

 Bezold, who maintain that thunder-storms are caused by the small special 

 barometric depressions, that he himself in his previous publications had 

 really long before defended the same conclusions, and that he was the 

 first from a complete investigation to demonstrate by observation the 

 truth of these views. [Z. 0. O. M., xviii, p. 426.) 



302. Dr. G. Hellmann, iu reply to Ferrari's reclamation, states that 

 many have pabliished their views on the connection between the thun- 

 der-storms and atmospheric pressure; thus the fact that the barometer 

 suddenly rises at the outbreak of a thunder-storm and subsequently 

 falls was, as far as he knows, first recorded by J. J. Planer, in 1782. 

 Rosenthal, and especially Lichtenberg, in 1784, gave a very clear de- 

 scription of the whole process. Equally clear and apparently independ- 

 ent of each other are Grouau (1821), Strehlke (1830), and Kiimtz (1836). 

 {Z. 0. G. M., XIX, p. 44 ) 



303. M. Moller discusses the suggestion thrown out by von Bezold to 

 the effect that the winds in thunder-storms and Boen agree with the 

 direction of the gradients, and do not correspond with the law of Buys- 

 Ballot. This latter law, Moller states, afiirms that, as a result of the 

 acceleration due to the gradients, and under the influence of constant 

 friction and by reason of the deflecting force of the earth's rotation, 

 there will for equal latitudes result also equal deviations of the wind 

 from the direction of the gradients. It has many times been shown how 

 this deflection must decrease if the air passes from strong to feeble 

 gradients. On the other hand, the reverse has not been stated, namely, 

 that a length of time to be counted by many hours is necessary to com- 

 municate to a mass of air a motion deviating from the direction of the 

 gradients, provided the air was originally at rest or did not possess such 

 deviation. By a little computation Moller shows that for short distances 

 and steep gradients, such as occur near a thunder-storm, the air — im- 

 pelled bj" the difference of barometic pressure, and having both a for- 

 ward velocity and a deviation due to the earth's rotation — although it 

 obeys them both, yet appears to obey only the first of these forces, be* 



