zinc (with seven molecules of water) in quadratic prisms: they are therefore isodimorphous. We must further notice the isomorphous sulphates and selenates, which crystallise in the anorthic system with five molecules of water SO,Cu" + 5H2O copper sulphate, Other isomorphous sulphates and selenates crystallise with four molecules of water in clinorhombic prisms SO,Mn" + 4H2O manganous sulphate, And, lastly, we must not forget, in the order of compounds now before us, the numerous double isomorphous sulphates of the magnesian series, SOM".SOR2 + CH2O, which crystallise in clinorhombic prisms, and in which M" may be represented by magnesium, zinc, nickel, cobalt, iron, cadmium, or copper, and R' by sodium, potassium, or ammonium, but not by a metal of the other group. There is no known sesquioxide of the alkaline metals, but there are some which are very important and characteristic in the groups of metals of which the atomic weights have been doubled by Cannizzaro. For instance, the sesquioxides of aluminium, iron, manganese, and chromium present important cases of isomorphism. The following oxides crystallise in the rhombohedral system: A1203 corundum, Fe2O3 specular iron, (FeTi), titaniferous iron (ilmenite), Analogous formulæ have been given to all these bodies, as well as to their isomorphous compounds, amongst which we may distinguish the spinels and the alums, which crystallise according to the regular system. MgO.Al2O, spinel, FeO.Fe,O, magnetic oxide of iron, ZnO.Al2O2 gahnite, ZnO.Fe2O, franklinite. The spinels form a very natural isomorphous group, and it is a well-known fact that their metallic elements often replace each other in the same crystal, without change of form. Thus, to take a single example, franklinite forms crystals in which Zn is replaced by Fe or Mn and Fe, by Mn,. It is unnecessary here to draw up a complete list of the alums (SO1) ̧M2.SO ̧R2+24H2O, in which M is aluminium, iron, manganese, or chromium, and R potassium, sodium, or ammonium. The examples which we have just quoted show that the new system of atomic weights is in harmony with the law of isomorphism; isomorphous elements have received atomic weights which allow us to give analogous formulæ to the similar compounds in which these elements occur. In our exposition of the origin of this discovery we have already indicated the assistance which, in certain cases, may be derived from it in determining atomic weights, when for a given element considerations of a chemical order leave us to choose between several values. There are, however, a few reservations which must be made in connection with the inferences to be drawn from isomorphism in the determination of atomic weights. In the first place we must clearly comprehend the definition of isomorphism. All bodies presenting identical forms, even with a similar composition, are not necessarily isomorphous. In order to be so the elements said to be isomorphous must be able to replace each other in the same crystal, as, for example, is the case with red silver, SbS,Ag, (pyrargyrite), and with proustite, AsS.Ag, with the spinels, garnets, alums, &c. following bodies, though they possess identical forms and a similar composition, are not isomorphous, as was formerly supposed : NO,Na sodium nitrate, CO,Ca calcium carbonate (Iceland spar). NO,K potassium nitrate (saltpetre), CO,Ca calcium carbonate (arragonite). The This must not be forgotten: considering the immense number of chemical compounds and the limited number of physical forms which they can affect, it must often happen that dissimilar compounds may appear under the same form, without authorising us to consider them as isomorphous. This, for example, is the case with the two iodides of mercury Hg,I, and HgI,, which both crystallise according to the quad from the law of volumes or the law of specific heats. In the determination of molecular and atomic weights the latter laws give more efficient aid to chemistry than the law of isomorphism, although the enunciation of these laws may not be strictly accurate from a physical point of view. We have already made this remark in connection with the law of specific heats, which is an incomplete law (p. 125). The same remark applies, though in a less degree, to the law of volumes. In fact, the laws of Gay-Lussac and that of Avogadro and Ampère are dependent upon the law of Mariotte, and are in a manner forced to follow its variations. CHAPTER VI. THE NEW SYSTEM OF ATOMIC WEIGHTS RESPECTS AND THE ANALOGIES WHICH EXIST RENDERS EVIDENT BETWEEN BODIES. DUMAS-MENDELEJEFF. I. THE new system of atomic weights renders evident numerous analogies which have been discovered in chemistry, between either the elements themselves or between their compounds or reactions, thus dealing with the most varied and the most profound questions of science. It is a vast subject, which might be developed to a great length, but of which we shall here only endeavour to give a sketch. Chemistry is not merely an immense collection of facts, but more exactly the science which teaches us to classify and arrange them, and this classification should begin with the elements themselves. Attempts have, we know, for some time been made in this direction. |