78 DR. FARADAY’S EXPERIMENTAL RESEARCHES IN ELECTRICITY. (SERIES XXVI.) 
which I have written has been upon very insufficient consideration ; but hoping that 
there might be some foundation of truth in the account of the physical cause of the 
variations which I have ventured to suggest, I have not hesitated to put it forth, 
trusting that it might be for the advantage of science. The magnetic properties and 
relations of oxygen are perfectly clear and distinct, and are established by experiment 
( 2774 . 2780.) ; and it is no assumption to carry these properties into the atmosphere, 
beeause the atmosphere, as a mere mixture of oxygen and nitrogen, is shown to pos- 
sess them also (2862.)*. It varies in its magnetic powers, by causes which act upon 
it under natural circumstances, and make it able to produce some such effects as 
those I have endeavoured generally to describe. 
2964. If it be a cause, in part only, of the observed magnetic variations, it is most 
important to identify and distinguish such a source of action, even though imper- 
fectly, for the attention is then truly and intelligently directed in respect of the action 
and the phenomena it can produce. The assigned cause has the advantage of occur- 
ring periodically and for the same periods, as a large class of the effects supposed to 
be produced by it ; and if the agreement should appear at first only general, still that 
agreement will greatly strengthen its claim to our attention. It has the advantage 
of offering explanations and even suggestions of many other magnetic events besides 
those which are periodical, and it presents itself at a time when we have no clear 
knowledge of any other physical cause for the variations, but are constrained vaguely 
to refer them to imaginary currents of electricity in the air or space above, or in the 
earth beneath. 
2965. The causes, both of the original power and of its secular variations, are un- 
known to us. But if, accepting the earth as a magnet, we should be able to distinguish 
largely between internal and external action, and so separate a great class of pheno- 
mena from the rest, we should be enabled to define more exactly that which we re- 
quire to know in both directions, should be competent to state distinctly the problems 
which need solution, and be far better able to appreciate any new hints from nature 
respecting the source of the power and the effects that it presents to us. 
2966. The magnetic constitution of oxygen seems to me wonderful. It is in the 
air what iron is in the earth. The almost entire disappearance of this property also, 
when it enters into combination, is most impressive, as in the oxynitrogens and oxy- 
carbons, and even with iron, which it reduces into a condition far below either the 
metal or the oxygen, weight for weight. Again, its striking contrast with the nitro- 
gen, which dilutes it, impresses the mind, and by the difference recalls that which also 
exists between them in relation to static electricity (1464.) and the lightning flash. 
Chlorine, bromine, cyanogen and its congeners, chemically speaking, have no magnetic 
relation to oxygen. In nature it stands in this respect, as in all its chemical actions, 
alone. 
2967. There is much to do with oxygen relative to atmospheric magnetism. Its 
* Philosophical Magazine, 1847, vol. xxxi. pp. 409, 406. 
