248 ROCHESTER ACADEMY OF SCIENCE. [Feb. 26, 



Other words the evidence is conclusive that magnetic perturbations 

 are not of thermo-electric origin and have nothing whatever to do 

 with heat radiations. In like manner there is no relation whatever 

 to light radiations. The photographic tracings which register auto- 

 tomatically the movements of the magnets give a complete record 

 of the times of occurrence of such storms, from which it appears that 

 they must depend upon conditions entirely unlike those concerned in 

 the radiation of light and heat. The ma.ximum effect of magnetic 

 storms is not recorded upon the side of the earth exposed to the 

 direct rays of the sun but upon that opposite. So, too, the auroras 

 accompanying magnetic storms, as seen in the Arctic regions at times 

 when daylight does not interfere with their observation throughout 

 the entire twenty-four hours, appear on the side of the earth away 

 from the sun. Thus there is no diffusion of effect such as appears in 

 the case of radiation. On the contrary there is concentration of the 

 maximum phases of any particular outbreak at a definite hour angle 

 from the sun and at certain latitudes which have reference to the 

 magnetic poles of the earth. Furthermore there are recurrences of 

 magnetic storms at the precise interval of a rotation of the sun as 

 viewed from the earth, which is advancing in its orbit in the same 

 direction that the sun is revolving on its axis. At each return at this 

 interval magnetic storms and auroras as a rule begin suddenly and 

 strongly and die out gradually, ceasing entirely most commonly within 

 two or three days. From this it appears that whatever it is upon the 

 sun that exercises these terrestrial magnetic effects has this power only 

 when in a certain very definite location relative to the position of the 

 earth. If outbreaks located promiscuously on every part of even the 

 visible surface of the sun were able to have these effects, periodicity 

 at the interval of a synodic rotation or any other regular interval 

 would be impossible. Thus the impulse that produces a magnetic 

 storm instead of being diffused indifferently in every direction from 

 its point of origin, as are light rays, is confined to one direction 

 exclusively. In short there is no analogy whatever between the 

 behavior of magnetic storms and the origin and diffusion of heat and 

 light radiations. 



Our conception of radiant energy as exhibited in the case of heat 

 and light has arisen from the analogy of atmospheric sound waves in 

 which the ultimate particles composing the conducting medium are 

 supposed to be thrown into a state of rythmical vibration. There is 

 a bounding and rebounding of elastic particles against each other 



