328 Mr. A. Trevel van on the Vibration of heated Metals. 



of those partial and sudden expansions which give the tremor- 

 causing impulse. 



The rocking on the narrow lead block may be induced by 

 two causes, either by some slight inequality in the weight of 

 the portions of the bar on the two sides of the ridge, or some 

 difference in the condition of the surface of that part of the 

 block which the ridge of the bar touches ; and may be thus 

 explained : 



1st, If the first-mentioned inequality exist, as soon as the 

 bar receives an upward heave, the greater weight of the one 

 side will cause it to incline to that side ; and as soon as that 

 heave ceases, and the contraction succeeds, the bar approaches 

 its original position, but will not remain in it, for the incli- 

 nation given to the preponderating side*will, on its return, of 

 course, cause the bar to incline to the opposite side. The 

 impulses which the bar receives, in this position, from the 

 renewed expansion and elongation, will not only renew the 

 upward heave, but also incline it again to the preponderating 

 side, and thus increase the lateral movement, which, like the 

 vertical, though altogether insensible at first, by incessant 

 frequent repetition accumulates and increases till the rocking 

 becomes conspicuous, and is rendered much more so by the 

 transverse rod. 



2ndly, If there exist any difference in regard to asperity in 

 the condition of the surface of that part of the block upon which 

 the ridge of the bar rests, it must necessarily follow, that the 

 impulse given to the bar on that side which is most rough 

 will be greater than on the other, and consequently the up- 

 ward heave will be so modified as to create an inclination to 

 one side. The bar thus thrown off its balance to the right on 

 ascending, will incline as far to the left on descending, and 

 there receiving the expansive impulse, it will be driven back, 

 and thus the principle of rocking will be created. The sound 

 depends upon the rapidity of the vibrations; for when slow, 

 no tone is heard. 



I think from the above-related experiments, and the con- 

 clusions and theory drawn from them, that the hitherto un- 

 known causes of many sounds are now accounted for. 



The sounds described by Humboldt as heard at sunrise by 

 those who sleep on certain granite rocks on the banks of the 

 Orinoco, also the sounds at sunrise produced by the statue 

 of Memnon, and the twang, like the breaking of a string, heard 

 by the French naturalists, as if proceeding from a mountain 

 at Carnac, are probably caused by the pyrometric expansions 

 and contractions of the heterogeneous material of which the 

 statue and the mountain consist. Similar sounds, and from the 

 same cause, are emitted, when heat is applied to any connect- 

 ed mass of machinery, and the snapping heard in a fire-grate 



