M. Berthelot, the celebrated French chemist, has stated that, on exposing argon mixed with benzene vapour to a rain of electric sparks, he has succeeded in causing argon to combine. Its volume certainly decreases, but whether this decrease is to be attributed to true combination or not is very doubtful. The benzene is converted into a resinous mass, which coats the walls of the tube; and it is not improbable that the argon may be dissolved, or even mechanically retained, in the resinous deposit. Helium, a gas closely resembling argon in properties, may be made to enter into a similar combination with metallic platinum, if combination it can be called; but the amount absorbed in both cases is extremely minute, and the gas is evolved unchanged on heating the resin or the metal. Professor Ramsay has also made experiments on the action of a silent electric discharge upon a mixture of argon with the vapour of carbon tetrachloride; the latter decomposes, giving, not a resin, but crystals of hexachlorobenzene and free chlorine; but the volume of the argon was unchanged. It was all recovered without loss. Lastly, the rare elements titanium and uranium have been heated to redness in a current of argon with no alteration or absorption of the gas. In short, all likely agents have been tried as absorbents for argon, but in no case has any true chemical combination been obtained. These failures to produce compounds make it impossible to gain any knowledge regarding the atomic weight of argon by a study of its compounds, for it forms none. It is, indeed, in the highest degree improbable that, had compounds existed, none should have been found in Nature. There are, it is true, a few elements, such as platinum and those resembling it, which always occur native, _i.e._ in the elementary state; but even they yield to the attack of the agents tried with argon. It cannot, of course, be stated with absolute certainty that no element can combine with argon; but it appears at least improbable that any compounds will be formed. It was therefore necessary to adopt some other method in attempting to determine the atomic weight of argon,--some method dependent on its physical rather than its chemical properties, for argon, unlike almost all other elements, appears to be devoid of chemical properties.