In order to prevent misunderstanding of the above, I must add that I have therein studiously assumed a negative position in reference to all hypothetical conceptions of the nature of heat, light, etc., and their modes of transmission, simply because I feel satisfied that the subject has hitherto been obscured and complicated by overstrained efforts to fit the phenomena to the excessively definite hypotheses of modern molecular mathematicians. The atoms invented by Dalton for the purpose of explaining the demonstrated laws of chemical combination performed this function admirably, and had great educational value, so long as their purely imaginary origin was kept in view; but when such atoms are treated as facts, and physical dogmas are based upon the assumption of their actual existence, they become dangerous physical superstitions. Regarding matter as continuous, _i.e._, supposing it to be simply as it appears to be, and co-extensive with the universe, in accordance with the experimental evidences of the unlimited expansibility of gaseous matter, we need only assume that our sensations of heat, light, etc., are produced by active conditions of such matter analogous to those which are proved to produce our sensations of sound. On this basis there is no difficulty in conceiving the rationale of the reaction which produces the repulsion of the radiometer. I may even go further, and affirm that it is impossible to rationally conceive radiation producing any mechanical effects without mechanical reaction. If heat be motion, and actual motion of actual matter, mechanical force must be exerted to produce it, and a body which is warmer on one side than the other, _i.e._, which is exerting more outward motion-producing force on one side than on the other, must be subject to proportionally unequal reaction, and, therefore, if free to move, must retreat in a direction contrary to that of its greater activity. Regarded thus, the residual air of the radiometer does act, not by collisions of particles between the vane and inside of the glass vessel, but by the direct reaction of the radiant energy which would operate irrespective of vessels, _i.e._, upon naked radiometer vanes if carried halfway to the moon, or otherwise freed from excess of atmospheric embarrassment. The recent experiments of Mr. Crookes, showing retardation of the radiometer with extreme exhaustion, seem to indicate that heat-rays, like the electric discharge, demand a certain amount of atmospheric matter as their carrier. I cannot conclude these hasty and imperfect notes, written merely with suggestive intent, without quoting a passage from the preface to the “Correlation of Physical Forces,” which, though written so long ago, appears to me worthy of the profoundest present consideration.