NATURAL PHILOSOPHY. 121 



erally negative, which accounts for the notable diminution of hori- 

 zontal intensity which precedes squalls. The precipitation of vapor 

 without rain, which often happens between eight and nine on clear 

 nights, and which is accompanied by very strong electricity, may 

 explain the magnetic perturbations which occur at that time, and the 

 diurnal electric period which corresponds with the movements of the 

 horizontal needle may belong to the same class of meteorological facts. 

 Even the aurora borealis may be included in this category, as there is 

 a continual fall of ice-needles, almost invisible, but whose existence 

 is clearly shown in the narratives of Polar voyages. Atmospheric 

 electricity on these occasions may, perhaps, be exalted by accessory 

 causes, such as the change which takes place when vapor passes to 

 the state of ice, or by the friction of wind against the little icicles in 

 a dry and very insulating atmosphere, and also by the inductive 

 action of superior regions on the falling and floating particles of ice. 

 These various subjects, M. Secchi tells us, are illustrated in his Me- 

 moires, but he does not pretend that magnetic disturbances have no 

 other causes than those indicated in the preceding remarks. 



CONNECTION BETWEEN EARTHQUAKES AND MAGNETIC PERTUR- 

 BATIONS. 



M. Lament, director of the Astronomical Observatory at Munich, 

 states that on the morning of December 26, 1861, while rectifying 

 the position of his magnetic apparatus, he remarked in all the instru- 

 ments an unusual perturbation. Their position changed rapidly and 

 irregularly, sometimes one way, sometimes another; and at the 

 same time there occurred a vertical vibration in the needles. A few 

 days after he learnt that at the very time he was making these obser- 

 vations an earthquake, which occasioned much damage, took place in 

 Greece. This, he says, proves once more, not only that the convul- 

 sions produced by earthquakes are perceived at a great distance, but 

 also that the forces which produce them modify terrestrial magnetism 

 to a certain extent. 



MAGNETISM OF HOT AND COLD* ROLLED IRON. 



Prof. Airy, in a communication to the Royal Society on the above 

 subject, states, first, that he had been desirous of examining whether 

 differences in the degree of change of subpermanent magnetism, such 

 as are exhibited by different iron ships, might not depend on the 

 temperature at which the iron is rolled in the last process of its man- 

 ufacture ; and for the purpose of experimenting, he had received 

 from the Dudley Iron Works twenty-four plates of iron, each sixteen 

 inches long, four inches broad, and one-fourth inch thick ; twelve of 

 which, after having been manufactured with the others in the usual 

 way, had been passed through rollers when quite cold. Each set of 

 twelve was divided into parcels of six each, one parcel being cut 

 with the length of the bars in the length of extension of the fibres 

 of the iron, the other being cut with the length of the bars transverse 

 to the length of extension. 



For. experimenting on these, a large wooden frame was prepared 

 capable of receiving the twenty-four bars at once, either on a plane 

 11 



