322 



apparatus which this doctrine requires, the author regards them as 

 rather consisting of the atmosphere and the surfaces of land and 

 water with which it is in contact. Thermo-magnetic phenomena, 

 he remarks, have hitherto only been observed in metallic combina- 

 tions, but this may be owing merely to the small scale on which our 

 experiments are conducted. 



To put to the test of experiment whether thermo-magnetism could 

 be excited when the surfaces of two metals, instead of touching at 

 one point, were in symmetrical contact throughout, the author first 

 employed a compound ring of bismuth and copper, the copper out- 

 wards, and he found that to whatever point heat was applied, mag- 

 netic powers were developed, a needle being affected differently, 

 according to the different positions in which the ring was placed with 

 regard to it. After a lapse of two years from this first experiment, 

 the author resumed the inquiry with an apparatus consisting of a 

 flat ring of copper, having its inner circumference grooved and united 

 firmly by soldering and fusion to a plate of bismuth cast within it ; 

 the whole forming a circular plate 12 inches in diameter, weighing 

 119 ounces troy, which was made to revolve in its own plane. 



Heat was applied by a lamp to a given point at the circumference 

 of this plate, and a delicately suspended needle, partly neutralized, 

 was placed near it, and the deviations observed in all positions of 

 the heated point, which was made to revolve (the lamp being with- 

 drawn). These experiments led him to conclude that the effect 

 of so heating a portion of the circumference, was to create a tem- 

 porary polarity in the plate, the law of which he explains. He then 

 details a set of experiments, by which he assured himself that a 

 uniformity of action obtained wherever in the circumference the 

 heat was applied. He next instituted a series of observations for 

 determining the laws which govern the magnetic phenomena result- 

 ing from the application of heat, as above described, the results of 

 which are stated in the form of tables. Four poles appeared to be 

 produced, two north and two south ; both the north lying in one 

 semicircle, and both the south in the other, and not in alternate qua- 

 drants, and all the poles lying rather nearer to the centre than line 

 junction of the two metals. The experiments were pursued in a va- 

 riety of positions of the plate with respect to the meridian and hori- 

 zon, and with a similar general result. 



From these experiments, the author concludes, that uniformity of 

 junction of the two surfaces of a thermo-magnetic combination is no 

 obstacle to the development of transient polarity. Regarding the 

 earth and the atmosphere as such a combination, and limiting our 

 views to the intertropical zone alone, we should have two magnetic 

 poles produced on the northern, and two on the southern side of the 

 equator, similarly pointed, the poles of the opposite names being dia- 

 metrically opposed to each other. 



To apply this to the earth, it is necessary to know the times of 

 greatest heat in the twenty-four hours : this may be assumed at three 

 o'clock in the afternoon. The apparatus used by the author not 



