twisted nor broke. At one period it was placed on the mantelpiece, over an open fireplace, and whenever a puff of wind came down the chimney, driving the heat of the fire up toward it, the needle would then spin with amazing rapidity, reminding us of the whirlwinds that spin up from heated plains of sand or dust. Our visitors were astonished to see that needle whirling there, without any visible cause, and with no apparent attachment, for the thread by which it was suspended was so slight as almost to escape observation.
The application of the hand to the side of the jar would always cause the needle to move from a state of rest. It was very interesting to watch its motion at night, whenever the light of a candle entered the jar. It furnished a beautiful illustration of the effect which is produced upon the still night-air of summer when penetrated by the first rays of the rising sun. We may conceive that the whole atmosphere at that moment responds with infinite currents, breezes, and motions, awakened into new life from a night's rest by the heat of the sun's rays. Of course, light, as a substance, has nothing to do with the motion. It results entirely from the expansion of the air by the force of heat. Our needle moved on the same principle precisely—and on no other—that the windmill moves in a current of air.
If the glass jar were to be exhausted of its air, and the needle were then to move when struck by a beam of light, the motion might be supposed entirely due to that light; but it may well be doubted whether it is possible to produce a vacuum so perfect that it would be entirely void of gaseous substance of some kind. The planetary and starry spaces themselves are probably not entirely free from matter that would respond to the action of heat, if indeed such matter is not necessary for the transmission of light and heat. An infinitesimal portion of air, or even of vapor of mercury, spread through a large jar or receiver, would doubtless obey the same law of expansion under heat that is observed by the atmosphere in its densest conditions, and, when set in motion, would prove sufficient to move a freely-suspended needle.
As our instrument, devised for proving the rotation of the earth, did not prove that fact, but showed something else, so the radiometer, which was supposed to prove the material character of light, did not prove that fact, yet it may serve to show something else. But we do not perceive why it should be called a radiometer, any more than the windmill should be called by that name.
