logarithmic decrement, and when the latter has sunk to one-twentieth the former has fallen off one-half. All these and other facts previously obtained prove that the action of light is not _direct_, but _indirect_; and Dr. Crookes has, after repeatedly refusing to consider hasty judgments, in consequence come to the conclusion stated above, that the rotation of the light mill is the result of heat. This decision accords with the opinion of other observers. The radiometer has already entered the field of industrial science, and is used to measure the duration of exposure of photographic plates. De Fonvielle has made with it a new determination of the sun's thermometric power. He made a spectroscope with a graduated screen, which permitted the amount of light that entered the apparatus to be graduated at will. In the path of the beam he placed a radiometer, and by comparing its action in the graduated light ray, and in the light of a standard oil lamp, burning 42 grammes (11.3 ounces Troy) per hour, he found that at 4 o'clock, on June 4, 1876, the radiating force of the sun was equal to 14 lamps placed 25 centimetres (10 inches) from the radiometer. TEMPERED GLASS IN THE HOUSEHOLD. The "tempered glass," which has made the name of M. de la Bastié, its discoverer, so well known, does not prove to be always manageable. It was to have the strength of metal, and not shiver with changes of temperature. But an English lady has found that it sometimes has precisely the contrary characteristics. She purchased twelve globes for gaslights, and they were made in the manufactory of M. de la Bastié himself. But one night, after the gas had been extinguished for exactly an hour, one of the globes burst with a report, and fell in pieces on the floor, leaving the bottom ring still on the burner. These pieces, which were of course found to be perfectly cold, were some two or three inches long and an inch or so wide. They continued for an hour or more splitting up and subdividing themselves into smaller and still smaller fragments, each split being accompanied by a slight report, until at length there was not a fragment larger than a hazel nut, and the greater part of the glass was in pieces of about the size of a pea, and of a crystalline form. In the morning it was found that the rim had fallen from the burner to the floor in atoms. In all these phenomena the behavior was that of unannealed glass, of which so many curious performances have been related. THE NEW YORK AQUARIUM.