It was discovered by Dulong and Petit, early in the century, that the higher the atomic weight of an element the less heat is required to raise its temperature through a given number of degrees. This heat can be measured by dropping a fragment of the element, carefully weighed and heated to a known temperature, into a known weight of cold water, and ascertaining what rise of temperature the water undergoes, owing to the heat communicated to it by the element. These comparative amounts of heat, if water is chosen as the standard, are termed specific heats. And as the specific heats of elements have been found by experiment to be inversely as their atomic weights, the product of the specific heat of any element and its atomic weight will give a constant number. If the quantity of element weighed is one gram, and its rise of temperature one degree, the numerical value of this product is about 6·4. Now the specific heat of mercury has been found to equal 0·032; that is to say, it requires only a fraction of the value of 0·032 of heat to raise the temperature of say 1 gram of mercury through one degree, whereas the amount of heat necessary to raise 1 gram of water through one degree is represented by the number 1. Hence this number, 0·032, multiplied by the atomic weight of mercury, should yield the product 6·4; and it is seen at once that that number must be 200, for 200 × 0·032 = 6·4. This is an additional reason for believing that the atomic weight of mercury must be represented by the number 200. We come next to a confirmatory piece of evidence which greatly strengthens the view that the atomic weight of mercury must be 200; but before entering into detail let us see what an atomic weight of 200 involves. The density of mercury gas is 100, and its molecular weight must be 200. But if its atomic weight is also 200, it follows of necessity that its molecule and its atom must be identical; that unlike oxygen and hydrogen, its molecule consists, not of two atoms, but of one single atom. There is nothing strange in this conclusion; there is no evident reason why single atoms should not act as molecules, or independent particles, able to exist in a free state, uncombined with each other or with any other molecules.