By the end of a week he had begun to suspect that his magnet analogy went farther than he was at first prepared to hold. The action of a magnet was along curved lines of force. So he asks:-- “Can induction through air take place in curves or round a corner--can probably be found experimentally--if so not a radiating effect.” After ten days more he has made another step. “Electricity appears to exist only in _polarity_ as in air, glass, electrolytes, etc. Now metals, being conductors, cannot take up that polar state of their own power, or rather retain it, and hence probably cannot retain developed electric forces. * * * * * “Metals, however, probably hold it for a moment, as other things do for a longer time; an end coming at last to all.” This, it will be observed, is nothing more or less than Clerk Maxwell’s theory of conduction as being the breaking down of an electrostatic strain. In January, 1836, followed the famous experiment of building a twelve-foot cube, which when electrified exteriorly to the utmost extent, showed inside no trace of electric forces. The account in the unpublished MS. of the laboratory book is, as is the case with so many of these middle-period researches, much fuller than the published _résumé_ of them in the “Experimental Researches.” All through 1836 he was still at work. Even when on a holiday in the Isle of Wight, in August, he took his notebook with him, and writes:-- “After much consideration (here at Ryde) of the manner in which the electric forces are arranged in the various phenomena generally, I have come to certain conclusions which I will endeavour to note down without committing myself to any opinion as to the cause of electricity, _i.e._ as to the nature of the power. If electricity exist independently of matter, then I think that the hypothesis of one fluid will not stand against that of two fluids. There are, I think, evidently, what I may call two elements of power of equal force and acting towards each other. These may conventionally be represented by oxygen and hydrogen, which represent them in the voltaic battery. But these powers may be distinguished only _by direction_, and may be no more separate than the north and south forces in the elements of a magnetic needle. They may be the polar points of the forces originally placed in the particles of matter; and the description of the current as an axis of power which I have formerly given suggests some similar general impression for the forces of quiescent electricity. Law of electric tension might do, and though I shall use the terms positive and negative, by them I merely mean the termini of such lines.” Right on until November 30th, 1837, this research was continued. The summary of this and the succeeding researches of 1838 on the same subject, drawn up by Professor Tyndall,[45] is at once so masterly and so impartial that it cannot be bettered. It is therefore here transcribed without alteration.