Oct. 3, 1878] 



NATURE 



593 



in varying proportions, of a few simplejbodies, let us 

 consider one of the homologous series of hydrocarbons ; 

 say the marsh-gas series. CH4, C2H6, CsHg, C4H10, 

 C5H12, C0H14, &c., &c. ; generally C„H2„+2. The 

 members of this series are all compounds, in varying 

 proportions, of carbon and hydrogen ; each differs from 

 the preceding by an increment of CHj. The difference 

 between the molecular weights (the weights of two vols.) 

 of the members of the series is 14. The physical pro- 

 perties of the series show a gradation from the first 

 upwards. The first is gaseous at ordinary temperatures : 

 as we ascend the series we have liquids of gradually- 

 diminishing liquidity, then solids, the melting-points 

 of which gradually increase. The chemical properties of 

 the series, so far as these have been investigated, also 

 exhibit a regular gradation. If we divide the specific 

 heats of these compounds by their molecular weights, we 

 do not obtain the same number for each — in other words, 

 the molecular heat of the members of this series varies : 

 this is a point of some importance. The specific volumes 

 — products obtained by dividing molecular weights by 

 specific gravities determined at that point at which each 

 vapour exerts the same tension, that is, at the boiling 

 points of the liquids— of the members of this series differ 

 by 22. A simple relation of some kind most probably 

 exists between the densities of the members of this 

 series, between the actions exerted by these bodies on 

 light, &c., &c. 



Take now a group of allied elements : — Oxygen (16), 

 sulphur (32), selenium (79 "S), tellurium (128). The atomic 

 weights — or the molecular weights, whichever form is 

 preferred — of the higher members of the series are 

 multiples of the atomic weight of the first : 16 X 2 = 32, 

 l6 X 5 = 80; 16 X 8 = 128. Oxygen is a gas, except 

 under conditions of great pressure ; sulphur melts at 

 about 115°, selenium at about 100°, but a modification at 

 215°; tellurium at 450° C. There is a gradation in the 

 general chemical nature of the four elements. There is 

 a simple relation between the specific volumes of these 

 four elements in the solid state ; this relation is expressed 

 by the numbers 1:3:3:4. The atomic heats of the three 

 last-named substances are the same ; the atomic heat of 

 solid oxygen, as deduced from observations carried out 

 on compounds, is rather less than the number repre- 

 senting the atomic heats of sulphur, selenium, and 

 tellurium. I might adduce other series of elements, let 

 one suffice : — 



Cacoium. 

 133 



The atomic weight of the second member of each sub- 

 section of this series is almost exactly the mean of the 

 atomic weights of the first and third j in each case the 

 number representing the atomic Weight of the middle 

 element is a very little less than the mean of the atomic 

 weights of the elements at the extremes of the series. 

 The specific volumes of the metals lithium, sodium, 

 potassium, and rubidium (the specific gravity of caesium 

 is unknown) are respectively, 11 "9, 237, 45*1, and 56'2 ; 

 these numbers are nearly in the proportion of i : 2 : 4 : 5 ; 

 there is a regular gradation, therefore, in the specific 

 volumes of the members of the present series. The 

 physical and chemical properties of the series show 

 gradations, which, so far as they have been exar^ined, 

 appear capable of tolerably simple generalisation. The 

 atomic heat of the five metals is represented by a 

 (practically) constant number. Now most of these facts 

 are quite in keeping with the hypothesis that the elements 

 which I have noticed are compounds of simpler forms ; 

 the properties of the compounds of the homologous series, 

 C„H2n+2, are in very many respects analogous with the 

 properties of the two series of so-called elements' to which 

 I have drawn attention. There is, however, one important 



difference between a certain physical property of the 

 homologous series and the same property as exhibited in 

 the elementary series — the atomic or molecular heats of 

 the elements are, with few exceptions, the same; the 

 numbers expressing the molecular heats of the members 

 of the homologous hydrocarbon series are multiples of 

 each other. 



" It seems a general law," says Berthelot, " that the 

 molecular heats of polymerised radicles are multiples 

 of each other, whereas the molecular heats of the elements 

 are, with very few exceptions, identical." The apparent 

 exceptions among the elements, we have good reason to 

 believe, will be found to obey the rule when more exact 

 investigations have been carried out. This difference 

 between the molecular heats of the elements and the 

 molecular heats of series of homologous hydrocarbons 

 lessens the probability of the elements being really com- 

 pounds, in varying proportions, of a few simple bodies ; 

 or at any rate, it leads us to. believe that, as Berthelot 

 says, the phenomena attending the decomposition of 

 the elements — supposing them to be really compounds — 

 must be different from the phenomena attending the 

 decomposition of those bodies which we know to be com- 

 pounds. Nevertheless, I think that too much weight 

 may be attached to this fact of differences in molecular 

 heats. We do know of many compound gases, gases in 

 the formation of which a very considerable amount of 

 condensation occurs, but which have almost identical 

 molecular heats. We have whole series of similarly con- 

 stituted groups of compounds having the same molecular 

 heats, i.e., provided we accept the ordinary formulae for 

 solid compounds as molecular formulae. So that the mere 

 fact that the elements have the same molecular heats 

 need not, I think, be a bar in the way of regarding these 

 bodies as compounds, provided we have other evidence 

 pointing in that direction. 



But I must now briefly consider the second form under 

 which the general question of the nature of the elements 

 presents itself, viz., are the elements compounds, in vary- 

 ing proportions, of 07ie primary form of matter? As 

 a matter of fact we know of compounds in varying 

 proportions of the same elements ; we know also of com- 

 pounds in varying proportions of one and the same 

 element. The facts which have been amassed concern- 

 ing allotropy and isomerism must be of service in any 

 attempt which may be made to answer the question we 

 are now to consider. 



It is generally possible to trace a simple relation be- 

 tween the specific volumes (^° '^ ') of the various 



members of a group of elements ; we have seen that 

 such a relation exists in the two groups already con- 

 sidered, viz., the oxygen group and the potassium 

 gi'oup. But it is stated by F. W. Clarke, of Cincinnati, 

 who has partially investigated this subject of specific 

 volumes, that so far as experiment has gone, no simple 

 relation can be traced between the specific volumes of 

 allotropic modifications of one and the same element. 

 This statement appears to me to assume an amount of 

 knowledge which we really do not possess. Clarke finds 

 for the specific volume of ordinary sulphur the numbers 

 io*4and I5'6; for prismatic sulphur he finds a number 

 varying from i6*3 to 167 ; there is no simple relation be- 

 tween this number and either of the former. But he has 

 assumed that the atomic weight of each allotrope is the 

 same, and we have no data warranting SH9h an assump- 

 tion ; the knowledge which we do possess points rather to 

 an opposite conclusion. I think I am right in saying that 

 in the case of oxygen and ozone we do possess some 

 accurate knowledge of the (relative) molecular weights of 

 two allotropes ; these molecular weights are different ; 

 hence, probably, the atomic weights of the allotropes are 

 also different. M. M. Pattison Muir 



^To be continued.) 



