A new and most valuable method of determining the molecular weights of non-volatile as well as volatile substances has just been brought into prominence by Prof. Victor Meyer (Berichte, 1888, No. 3). The method itself was discovered by M. Raoult, and finally perfected by him in 1886, but up to the present has been but little utilized by chemists. It will be remembered that Prof. Meyer has recently discovered two isomeric series of derivatives of benzil, differing only in the position of the various groups in space. If each couple of isomers possess the same molecular weight, a certain modification of the new Van't Hoff-Wislicenus theory as to the position of atoms in space is rendered necessary; but if the two are polymers, one having a molecular weight n times that of the other, then the theory in its present form will still hold. Hence it was imperative to determine without doubt the molecular weight of some two typical isomers. But the compounds in question are not volatile, so that vapor density determinations were out of the question. In this difficulty Prof. Meyer has tested the discovery of M. Raoult upon a number of compounds of known molecular weights, and found it perfectly reliable and easy of application. The method depends upon the lowering of the solidifying point of a solvent, such as water, benzine, or glacial acetic acid, by the introduction of a given weight of the substance whose molecular weight is to be determined.

The amount by which the solidifying point is lowered is connected with the molecular weight, M, by the following extremely simple formula: M = T x (P / C); where C represents the amount by which the point of congelation is lowered, P the weight of anhydrous substance dissolved in 100 grammes of the solvent, and T a constant for the same solvent readily determined from volatile substances whose molecular weights are well known. On applying this law to the case of two isomeric benzil derivatives, the molecular weights were found, as expected, to be identical, and not multiples; hence Prof. Meyer is perfectly justified in introducing the necessary modification in the "position in space" theory. Now that this generalization of Raoult is placed upon a secure basis, it takes its well merited rank along with that of Dulong and Petit as a most valuable means of checking molecular weights, especially in determining which of two or more possible values expresses the truth. - Nature.

[Continued from Supplement, No. 642, page 10258.]