The relationships of the metal which Mendelejeff believed would fill this gap will be more easily seen by omitting the even series on either side of it, and taking only the odd series with which it will, as already mentioned, the more closely correspond.

* Cu, Ag, and Au are included in Group I. on account of their forming cuprous, argentous, and aurous oxides, but on account of their resemblance in many respects to the metals in Group VIII. they are also included in it. + This blank has now been filled up by the discovery of gallium.

As it stands between zinc with an atomic weight of sixty-five, and arsenic with one of seventy-five, while it is separated from the latter by a blank, its atomic weight must be about sixty-eight. As it is atom-analogous with Al, its salts should have a similar constitution. It should form an oxide x2O3, and a sulphide x2S3. It will be precipitated from its solution by ammonium sulphide. The metal should be easily reduced by carbon or sodium, it should have a specific gravity of 5.9, and decompose water at a red heat. As it belongs to an odd series, it should, like zinc, form volatile compounds with organic radicals, and form also anhydrous chlorides.

On the discovery of the metal gallium, it was found to agree in almost every respect with the prediction of Mendelejeff, and this fact is not interesting to chemists only, but also to pharmacologists. For the great object of pharmacology is to obtain such a knowledge of the relations between the physical and chemical characters of bodies, and their actions upon the living organism, that we may be able to predict their actions with certainty, and to know the modifications which alterations in their physical and chemical characters will produce on their physiological action.

Mendelejeff's present classification is imperfect, because we find that by it the members of some natural groups, such as those of the earthy metals, are separated from one another, although they agree in their chemical characters.

We find also that metals having similar pharmacological actions, as copper, zinc and silver, do not fall naturally together in this arrangement. But, on the other hand, we find also that by this classification, elements are brought together which do not at first seem to have any resemblance to each other, and are yet found by recent investigations to have a physiological connection. Thus mercury and calcium do not appear to resemble one another, yet Prevost has shown that, in acute poisoning by mercury, the calcareous matter disappears from the bones, and in the process of elimination by the kidneys produces calcification of these organs.1