78 



SCIENCE. 



[Vol. II., No. 24. 



the motions which originate radiations are not 

 confinecl to such vibrations of atoms, even if it 

 be possible that such vibrations do originate 

 radiations. And this consideration leads us to 

 what appears to be the truth of the matter, 

 which is, that the atoms themselves are in a 

 state of internal vibration. As will be seen 

 subsequent!}', this internal vibration is, no 

 doubt, accomplished under the action of internal 

 forces, which permit extremely small deforma- 

 tions onl}- of the atom b}' anj- external forces 

 which can be brought to bear upon it ; i.e., the 

 modulus of elasticitj- of an atom is very large 

 indeed, and verj- large, no doubt, when com- 

 pared with that of the molecule. Indeed, if 

 such vibrations exist within the atom itself, it 

 is not difficult to prove that the force which 

 binds the parts of an atom together (and con- 

 sequently its modulus of elasticity) is much 

 greater than the chemical force binding the 

 atoms together into a single molecule ; for it 

 has been shown, in my paper upon the internal 

 molecular energy of atomic vibration, that the 

 amount of energy which can be imparted to a 

 s.ystem like this is iuverselj- as the modulus of 

 elasticitj-. But chemical atoms are bodies 

 which we are now supposing to be in internal 

 vibration, but to which it has been found im- 

 possible to communicate energy in amount suf- 

 ficient to cause them to fly to pieces. Since 

 atoms do not become decomposed, while mole- 

 cules do under various circumstances, it must 

 be that their modulus of elasticitj^ is much 

 larger than that of molecules. 



This view accords with that of Lockyer,^ 

 who has endeavored to explain the coincidence 

 of lines in the spectra of different elements, 

 and the relation of temperature to spectra, by 

 the supposition that the so-called chemical ele- 

 ments are merely molecules which have never 

 yet been decomposed hy chemists. It must 

 be admitted that the experimental evidence he 

 adduces is of a verj- cogent character : and it 

 seems to me that the demonstration by which 

 I have shown that the mean energj- of such a 

 vibration would be extremelj* small explains 

 how such a vibration can exist without de- 

 composing the more complex atoms even at 

 the highest artificial temperatures, though 

 Lockj-er has reason to think that thej' are 

 decomposed in the hotter stars, where only the 

 spectra of the elements of low atomic weight 

 are to be found. 



Were it true that everj* degree of freedom 

 must have the same kinetic energy, we could 



I Discussion of the worliing hypothesis, that the so-called 

 (chemical) elements are compound bodies (Natui'e^ Jan. 2 and 

 Jan. 9, 1879). Necessity for a new departure in spectrum analy- 

 sis {Nature, Nov. 6, 1879). 



not admit the possibility of such a vibration; 

 for not only would such large amounts of 

 energj' be required by the degrees of freedom 

 which seem certainly to exist between the 

 atoms of complex molecules as to entirely con- 

 tradict experimental values of the specific 

 heat, but the supposition of additional de- 

 grees of freedom within each atom would 

 require an amount of energy, on the whole, 

 manj' times the actual specific heat of such 

 bodies. But when the amount of energj^ re- 

 quired bj' such degrees of freedom is nearlj' a 

 vanishing qu.antity, as I have shown, there is 

 nothing to prevent us from assuming that to 

 be the truth which spectroscopic evidence 

 makes most probable. 



AVe maj' notice, in passing, that the principle 

 upon which this paper rests, that vibrations of 

 this character can exist without absorbing an 

 appreciable amount of kinetic energy, enables 

 us to explain at the same time the extremely 

 moderate rate at which exchanges of heat take 

 place between bodies bj- radiation. They be- 

 come onljf very slowlj^ of the same tempera- 

 ture, which fact needs explanation in view of 

 the extremely rapid propagation of radiations 

 themselves. Now, according to our supposi- 

 tion, during a molecular encounter the mole- 

 cules are roughlj* shaken, and there is a deter- 

 minate distribution of energj- to be found among 

 the atoms, at its conclusion, in the form of in- 

 ternal atomic vibration, which distribution is 

 due to the circumstances of the encounter. 

 Those atoms which by chance have more energy 

 than others radiate more rapidly ; and since the 

 velocity of radiation is so great, and the 

 atomic distance so small, we maj- assume 

 that the several atoms acquire almost instan- 

 taneously an energy of internal vibration 

 sensibly equal to the mean, so that in a gas 

 this is their condition during almost the entire 

 free path of a molecule. In case the gas is 

 becoming cooler by radiation to surrounding 

 bodies, the atoms which radiate to these bodies 

 lose more of their vibratory energy than they 

 otherwise would, and thus have less mean 

 energy of internal vibration than they should 

 have under the law of distribution which de- 

 termines what fraction this energy shall be 

 of the mean kinetic energy of the molecules. 

 At the next encounter, the atoms receive their 

 proper share of the mean kinetic energy, 

 which, being partially lost by radiation, is again 

 supplied ; and so on. And because this trans- 

 formation into internal atomic vibration must 

 take place before it can be radiated, and be- 

 cause at the same time the energy of this 

 vibration is but an unappreciable fraction of 



