A series of remarkable experiments was performed by professor Boutigny, at the British Association at Cambridge, in 1845. He commenced by showing, that when cold water is poured on a hot metallic surface, the heat is not communicated to it; and that the water assumes a spheroidal form, and continues to roll about, upheld at a minute distance from the heated surface, without boiling. The water was poured into a hot platinum cup kept in rapid motion, and resembled a small globe of glass dancing about. There was no hissing noise nor appearance of steam, though the globule of water must, nevertheless, have evaporated rapidly; for, after gradually diminishing in size, in the course of about two minutes it disappeared. The same result takes place when any substance capable of assuming a globular form is placed on a heated surface. In proof of this, the professor placed in the heated cup of platinum, iodine, ammonia, and some inflammable substances; each of which became globular, and danced about like the globule of water, but without emitting smell or vapour, or being inflamed, until the platinum cup was cooled. Another experiment was yet more curious. Professor Boutigny heated a silver weight, of the same shape as the weight of a clock, until it was red-hot, and then lowered it by a wire into a glass of cold water, without there being any more indication of action in the water than if the weight had been quite cold. Professor Boutigny advanced no theory to account for these peculiar actions, further than that a film of vapour intervenes between the heated body and the substance, which prevents the communication of heat. The facts, however, he thought were of importance in a practical point of view, both as regards the tempering of metals, and in the explanation of the causes of steam-boiler explosions. It would seem, from experiments in tempering metals, that, if the metal be too much heated, the effect of plunging it into water will be diminished. In steam-boilers, also, if the heated water be introduced into a heated surface, the heat may not be communicated to the water, and the boiler may become red-hot, and without any great emission of steam; until, at length, when the boiler cools, a vast quantity of steam would become suddenly generated and the boiler burst.