370 



NA TURE 



\_Fcb. 1 8, 1886 



molecular structure. When, however, the temperature is 

 high and the particles far apart, this structure, as revealed 

 by its spectrum, is much simpler. A process of splitting 

 up has taken place in the interval. 



While, however, this is universally allowed, tliere is a 

 'difference of opinion as to the nature of this simplifica- 

 tion, which, we are assured by the spectroscope, has 

 taken place. Thus we have already seen that in gaseous 

 water or steam we may have, at a somewhat high tem- 

 perature, a considerable variety of structures and a par- 

 tial dissociation of the various compound molecules. In 

 such a case we have at the same time portions of the 

 compound and portions of the components, thus exhibit- 

 ing a more or less complicated structure of the gas. 

 When, however, the temperature gets very high, the dis- 

 sociation is practically complete, and the compound 

 structures disappear, leaving us with molecules of oxygen 

 and hydrogen. But what will happen if we treat the 

 vapour of iodine in a similar manner ? It will be allowed 

 that as the temperature gets higher we shall have a simplifi- 

 cation of molecular structure accompanied and exhibited 

 by a great change of spectrum, but will the iodine ever split 

 up into components which bear to iodine a relation simi- 

 lar to that which oxygen and hydrogen bear to water? 



In fine, we call iodine a simple body because in the 

 conditions in which we are placed in our laboratories we 

 cannot decompose it ; but what is it in the vacuum-tube 

 and under the spectroscope ? Is it still an element, or 

 does it give any evidence of being a compound .' 



It is taken for granted that at high temperatures its 

 molecules split up, but do they split up into portions of 

 iodine or into portions of the components of iodine ? In 

 discussing this and similar questions we shall begin by 

 acknowledguig that the strongest and best proof of the 

 compound nature of any element is the exhibition of its 

 components in a separate state, while at the same time 

 we must confess that we are at present unable to do this 

 for the so-called elements. Nevertheless this inability 

 forms no ground for the assertion that the elements are 

 simple bodies, inasmuch as certain substances which we 

 know to be compound reveal their components moment- 

 arily in the spectral flame. There is a momentary dis- 

 sociation at a high temperature followed by a reconstruc- 

 tion at a low. 



A good instance of this is the yellow flame produced by 

 introducing chloride of sodium into a Bunsen's burner. 

 This yellow flame attests the existence of sodium in the 

 free state, but this existence is merely temporary, and at 

 the end of the process there is no perceptible trace of the 

 presence of this metal. Thus the only difference between 

 the experiment in which the presence of sodium is tem- 

 porarily revealed and that in which splitting up takes 

 place when the various elementary gases are brought to a 

 high temperature mny be that in the former instance wc 

 can obtain the sodium by another means, whereas in the 

 latter expernnents we cannot obtain these constituents by 

 any other means. We say may be because we know that 

 our powers are limited and can very well conceive their 

 extension in the future as we know they have extended 

 in the past. We think it, therefore, unphilosophical to 

 assume that there is any real difference between those 

 bodies which we cannot decompose and those bodies 

 which we can, unless there is some good reason for this 

 distinction apart from our inability to decompose the 

 former. Let us now, therefore, inquire whether any such 

 grounds exist. 



Our first remark is that in certain respects the elements, 

 with one or two exceptions, may be looked upon as be- 

 longing to a distinct family each member of which pos- 

 sesses the same or nearly the same atomic heat. This 

 means that the amount of heat necessary to raise through 

 a given temperature range an atom of any one element 

 is equal to that necessary to raise through the same range 

 an atom of any other element. 



This fact was first discovered by Dulong and Petit, and 

 is expressed by saying that the product of the specific 

 heat into the atomic weight, or the atomic heat, as this 

 is called, is nearly the same for all the elements. 



This peculiar law is not confined to the elements, for it 

 has been found that in all compound bodies of similar 

 atomic composition the product of the specific heat into 

 the atomic weight is likewise constant. This product is, 

 however, greater in the class of compound bodies than 

 it is for the so-called elements. For the latter the product 

 is about 6, while for the chlorides of barium, strontium, 

 calcium, &c., it is over iS, and for the carbonates of lime, 

 barytes, &c., it is nearly 22. 



Thus the distinction which the elements enjoy as a 

 family consists in the fact that their atomic heat is less 

 than that of families of compound substances. In order 

 to perceive the physical meaning of this peculiarity let us 

 imagine that we make a mixture of two substances, h. and 

 F., which have no sort of chemical attraction for each 

 other. Now in order to heat this mixture through a 

 certain temperature range, the heat required will be the 

 sum of the heats required for the two components, and 

 neither more nor less, the one being in this respect abso- 

 lutely independent of the other. 



Next let us suppose that .\ and b are both chemical 

 compounds, but that the atomic constituents of a com- 

 pound atom of .\ exercise on each other (in order to form 

 the compound atom) attractions vastly greater than those 

 u'hich the compound atoms of A so formed exercise upon 

 each other. 



Let us also imagine, that a similar law holds for B, 

 so that in fine we have to deal with the following forces, 

 some of which are strong and others weak. Thus we 

 have : — 



(a) The strong forces exercised by the various chemical 

 constituents of a compound atom of A on each other. 



(/3) The strong forces exercised by the various chemical 

 constituents of a compound atom of B on each other. 



(y) The relatively feeble forces exercised by the various 

 compound atoms of .\ upon each other. 



(3) The relatively feeble forces exercised by the various 

 compound atoms of B upon each other. 



(e) The relatively feeble forces exercised by the com- 

 pound atoms of \ upon the compound atoms of B. 



Under these circumstances there are perhaps theoretical 

 grounds for imagining that when we mix .\ and B to- 

 gether not only shall we have, as above-mentioned, an 

 independence between the specific heats of .\ and B, but 

 in addition the specific heat of a compound atom of A 

 will be found to be equal, or nearly so, to that of a com- 

 pound atom of B. 



If we now apply these principles to the so-called 

 elementary bodies, we shall, I presume, be all willing to 

 own in the first place, that (assuming for the sake of 

 argument that they are in reality compounds) the force 

 which binds their various constituents together must at 

 any rate be vastly greater than that which represents the 

 attraction of one so-called element for another. Imagine, 

 now, one atom of barium and two of chlorine to combine 

 together to form one compound atom of chloride of 

 barium, we may safely assert that the strength of the 

 chemical ties which bind together the various constituents 

 of this compound atom must be vastly weaker than those 

 which bind together the assumed constituents of the 

 element chlorine or of the element bariun^ In con- 

 formity therefore with the sucgestions we have ventured 

 to make we might expect two things to happen : 



First, the heat necessary to heat through a given tem- 

 perature range a compound atom of chloride of barium 

 ought to be nearly equal to the sum of the heats necessary 

 to heat through the same range an atom of barium and 

 two atoms of chlorine. 



Next the heat necessary to heat through this range an 

 atom of chlorine ought to be nearly the same as that 



