properties of Frankland's zinc ethyle leaves no doubt of the biatomic character of zinc, for the compound Zn (CI) has the same molecular volume as ether O (C2 H3), and if the atom of zinc were taken out and replaced by one atom of oxygen, there would be no change of volume. Then half the ethyle in zinc ethyle is replaceable by iodine, just as half the ethyle in ether is replaceable by potassium. The biatomic character of this family being thus established, we can extend the conclusion to the other metals which form magnesian oxides, so called from the striking analogy of constitution of several of their salts with the corresponding salt of magnesia. In this manner we are led to adopt for iron, manganese, nickel, cobalt, and copper atomic weights corresponding to biatomic cha acters. The subsulphide of copper is thus represented by the formula Cu S, which is sufficiently similar to that of sulphide of silver, Ag S, to remove our surprise at their isomorphism. There is, moreover, in the reactions of alumina, sesqui-oxide of iron, sesqui-oxide of chromium, and sesqui-oxide of manganese, much resemblance. All these are weak bases, and their sulphates form with sulphate of potash those most characteristic salts called alums. The three first are isomorphous in the uncombined state, so that the conclusion established for iron and manganese may be extended to aluminium and chromium. But it is also arrived at by other means, for chromium in combination with oxygen and chlorine forms the well-characterized compound Cr O2 Cl chloro-chromic acid, which contains the same quantity of oxygen and of chlorine as chlorosulphuric acid in two volumes of vapour, having 52.5 of chromium in the place of the 32 of sulphur of that compound. Again, chromic and sulphuric acids exhibit a marked resemblance of properties, the former being, if anything, even more distinctly bibasic than the latter, and their normal potash salts are isomorphous, so that chromium is abundantly proved to be similar to sulphur in atomicity, and brings in evidence of its own in favour of the biatomic character of aluminium, iron, and manganese. In like mauner manganese in manganic acid is connected with sulphur in sulphuric acid, and requires a corresponding biatomic weight. The isomorphism and general analogy of per-manganate of potash with per-chlorate of potash has often been alluded to as pointing to the necessity of representing the former by a formula containing one large atom of manganese, Mn O'K: but although this formula, by assimilating the expressions for these two similar bodies, removes one difficulty, it creates at the same time another difficulty, by presenting a formula containing only one atom from the first family of elements. The speaker said he would not at present hazard any opinion regarding the propriety of removing this difficulty by doubling the above formulæ, together with that of per-chlorate of potash, although he might remark that the constitution of the basic per-iodate of soda points to the formula Io Oo Na* 3 (H2 O). An exceedingly strong ground for admitting for many heavy metals the atomic weight corresponding to a biatomic character was brought forward some time ago by Wurtz, who pointed out that adopting for oxygen the atomic weight 16, we get a half-molecule of in one molecule of various salts if we consider the heavy metals monatomic. water H2 O CH3 Other metals are susceptible of reduction by similar analogies to the class of elements which are biatomic or tetratomic, &c. Thus mercury is proved by the ethylide and methylide to be biatomic by the fact that the compound for one atom of mercury with two atoms of ethyle or of methyle, occupies the same volume in the state of vapour as the compound of one atom of oxygen with two of ethyle or of methyle Hg (C H3)2 = 2 vols., and we can take out one atom of methyle from the bi-methylide of mercury, and replace it by an atom of chlorine, bromine, or iodine without disturbing the type, Hg. The common bi-chloride of mercury has, moreover, a vapour volume corresponding to the biatomic character of the metal, and the same thing holds good of the vapour of metallic mercury itself, which has the same volume as the metal cadmium, and probably zinc, and the wellknown biatomic radicals CO, S O2, C2 H', but double the volume of the elements oxygen and sulphur. In the present state of our knowledge the speaker was not aware of any sufficient grounds for deciding which of these two constitutions of the free molecule of a biatomic element or radical is to be considered as normal and which is abnormal. On the one hand, mercury, cadmium, and all known biatomic radicals have a molecule containing one atom, while the molecule of oxygen contains two atoms, and that of sulphur two at high temperatures and six at lower temperatures. Selenium is at high temperatures like sulphur. It has been amply shown by Dr. Odling and others that tin is biatomic and tetratomic in its two chlorides, and its compounds with the organic radicals and chlorine, &c., leave no room for doubt on the point. By similar chains of evidence the remaining metals can be shown to belong to the great biatomic class containing already so many. The vapour densities of the so-called sesqui-chlorides of iron, aluminium, and chromium, as determined by Deville, show that the molecule of each of these bodies contains two atoms of metal and six atoms of chlorine, in fact the same quantity of metal as the molecule of the sesqui-oxide: this fact has been held to be an anomaly from the point of view adopted regarding their atomic weights. The speaker believed, however, that so far from being anomalous, these vapour densities are the least which can be reconciled with the conclusion that the metals permanently combine with even numbers of atoms from the first family, for if one atom of iron could on occasion combine with three atoms of chlorine to form one molecule, the law respecting it would assume the not very wise form,- that iron combines with an even number of atoms from the first family, except when it combines with an uneven number! The fact is, that the sesqui-chlorides are not exceptions to the law, as at first sight they are suspected of being. Precisely the same remarks apply to the so-called subchloride of sulphur of which the molecule is S Cl2, as required by the law. So also cyanogen C N3, acetylene CH, ethyle CH, &c. &c. Amongst exceptions, the speaker mentioned nitric oxide and calomel, both of which have vapour densities corresponding to the molecular formulæ N O and Hg Cl. Many compounds are known to undergo decomposition on evaporation, and to be reproduced on condensation; thus N HO yields the two molecules N H3 and H2O, each with its own volume, as also S O3 H2 yields SO2 and H2O. SOH and P Cl3 are also known to yield on evaporation vapour corresponding to a breaking-up into two molecules; and there are strong reasons from analogy, as well as experimental evidence, to believe such decomposition. As, however, a high authority seems inclined to doubt the decomposition, the matter may be considered as still sub judice. The existence of basic salts of mercury or copper, when apparently monatomic, is another class of apparent exceptions to the law. For if, in the sub-nitrate of mercury, the atom of metal really replaced one atom of hydrogen, just as potassium does in nitrate of potash, there ought not to be basic sub-nitrate of mercury, any more than a basic potash salt; whereas if the sub-nitrate of mercury contains, as the speaker asserted, in one molecule two atoms of metal and two atoms of the salt radical of the nitrates (N O3), then a basic salt is as natural and intelligible a compound as the basic nitrate of the red oxide. The action of ammonia on calomel confirms the molecular weight Hg2 Cl2; for the compound N H2 Hg2 Cl, formed simultaneously with sal ammonia, proves that twice (Hg CI) takes place in the reaction. [A. W. W.] ANNUAL MEETING, Monday, May 2, 1864. WILLIAM POLE, Esq. M.A. F.R.S. Treasurer and Vice-President, in the Chair. The Annual Report of the Committee of Visitors for the year 1863 was read and adopted. The amount of Contributions from Members and Subscribers in 1863, amounted to £3,423; the Receipts for Subscriptions to Lectures were £716. 12s. 6d.; the total Income for the year amounted to £5,532. 10s. 7d. On Dec. 31, 1863, the Funded Property was £30,107. 14s. 1ld.; and the Balance at the Bankers', £1,056. 14s. 9d., with Six Exchequer Bills of £100 each. Fifty-five New Members were elected in 1853. A List of Books Presented, amounting in number to 121 volumes, accompanies the Report, making, with those purchased by the Managers and Patrons, a total of 561 volumes (including Periodicals) added to the Library in the Year. Sixty-three Lectures and Nineteen Evening Discourses were delivered during the year 1863. Thanks were voted to the President, Treasurer, and Secretary, to the Committees of Managers and Visitors, and to Professor Faraday, and the other Professors, for their services to the Institution during the past year. The following Gentlemen were unanimously elected as Officers for the ensuing year : PRESIDENT-The Duke of Northumberland, K.G. F.R.S. WEEKLY EVENING MEETING, Friday, May 6, 1864.. COLONEL PHILIP JAMES YORKE, F.R.S. in the Chair. PROFESSOR ROSCOE, F.R.S. On the Metal Indium and Recent Discoveries on Spectrum Analysis. SINCE the spring of 1862, when the speaker delivered a course of three lectures in this Institution on the Spectrum Discoveries, much has been done to increase our knowledge of Spectrum Analysis, but the whole subject is still in its infancy, and the further we advance the more we find remains to be known. No less than four new elementary bodies have already been discovered by means of Spectrum Analysis: Cæsium and Rubidium, by Bunsen; Thallium, by Mr. Crookes; and Indium, by Reich and Richter, of Freiberg; whilst the foundations of Solar Chemistry, laid by Kirchhoff, have been rendered more secure by the observations of Cooke, in America; Donati, in Italy; and Miller and Huggins, in England. Cæsium and rubidium were at first only found in one or two mineral waters; they have since been shown to be widely distributed in the vegetable as well as in the mineral kingdom; they have been obtained in considerable quantities from the beet-root salt, and found in the ashes of tea and coffee, thus proving that they occur commonly in soil; whilst, quite recently, M. Pisani has found that a mineral, called pollux, occurring in Elba, contains 34 per cent. of cæsium, this metal having been mistaken for potash in the analyses which had previously been made of this substance. Thallium and its compounds have been obtained in large quantities, and their properties fully investigated by Crookes and Lamy, whilst this metal has not only been found in iron pyrites, but also in large quantities by Schrötter, in the mica of Zinnwald, and in lepidolite, from Moravia. Thallium has been shown by Boettger to occur together with cæsium and rubidium in the mineral water of Nauheim, near Frankfort; Boettger has, moreover, shown that thallium is contained in the vegetable kingdom, he has found it in the yeast of the vinous fermentation; so that thallium exists in wine; also in treacle, tobacco, and chicory. If 4 lbs. of any of these substances are employed, a sufficient quantity of thallium can be obtained as the double platinum-chloride to enable its presence to be easily detected. Professor Bunsen has informed the speaker that he has found a mother liquor from the Hartz, which contains so much thallium, that the |