EARTHQUAKES. 203 



quakes occurred.* 4 These facts agree with the observations 

 made by Adolph Erman, (in the temperate zone, on the 8th 

 of March, 1829.) on the occasion of an earthquake at Irkutsk, 

 near the Lake of Baikal. During the violent earthquake of 

 Cumana, on the 4th of November, 1799, I found the declina- 

 tion and the intensity of the magnetic force alike unchanged, 

 but to my surprise, the inclination of the needle was diminished 

 about 48'.f There was no ground to suspect an error in the 

 calculation, and yet in the many other earthquakes which I 

 have experienced on the elevated plateaux of Quito and Lima, 

 the inclination as well as the other elements of terrestrial mag- 

 netism remained always unchanged. Although in general, 

 the processes at work within the interior of the earth may not 

 be announced by any meteorological phenomena or any specia. 

 appearance of the sky, it is, on the contrary, not improbable, 

 as we shall soon see, that in cases of violent earthquakes some 

 effect may be imparted to the atmosphere, in consequence of 

 diich they cannot always act in a purely dynamic manner. 

 During the long-continued trembling of the ground in the 

 Piedmontese valleys of Pelis and Clusson, the greatest changes 

 in the electric tension of the atmosphere were observed whilst 

 the sky was cloudless. The intensity of the hollow noise which 

 generally accompanies an earthquake does not increase in the 

 same degree as the force of the oscillations. I have ascer- 

 tained with certainty that the great shock of the earthquake 

 of Riobamba (4th Feb. 1797) one of the most fearful phe- 

 nomena recorded in the physical history of our planet was 

 not accompanied by any noise whatever. The tremendous 

 noise (el (/ran ruido] which was heard below the soil of the 

 cities of Quito and Ibarra, but not at Tacunga and Hambato, 

 nearer the centre of the motion, occurred between eighteen 

 and twenty minutes nffer the actual catastrophe. In the cele- 

 brated earthquake of Lima and Callao, (28th of October, 1 746,) 

 a noise resembling a subterranean thunder-clap was heard 

 nt Truxillo a quarter of an hour after the shock, and unaccom- 

 panied by any trembling of the ground. In like manner 

 long after the great earthquake in New Granada, on the 1 6th 



* I have given proof that the course of the hororary variations of tha 

 barometer is not affected before or after earthquakes, in my JRelat. hut., 

 t. i. p. 311 and 513. 



t Humboldt, Jielat. hist., t. i. p. 516-517. 



