460 



NATURE 



[Sept. 6, 1888 



■comparison with the wave-length \, so that | and tj will be nearly 

 ■equal, and therefore we may write — 



r + s 



U<=* j 



I + 



tan 



(¥-!)! 



As a first approximation we may take | = 77, and then the 

 vibrations will be the same as those parallel to the axis. Since, 

 however, the centre of gravity remains fixed, the vibration must 

 be a pendulous one about this centre, which introduces a fresh 

 set of considerations. The proper vibrations of the molecule 

 would still be given by £ — o and 17 = o, but, owing to the 

 pendulous vibration, these would not completely determine the 

 motion. The difference in the action of light in different direc- 

 tions, and the corresponding fluted nature of the spectrum, would 

 appear to depend essentially on considerations of this kind. 1 



In the case of a triatomic molecule, we obtain three sets of 

 linear equations of the same form as (25) and (26), together with 

 one of the form (27) ; it is, however, unnecessary to pursue this 

 further. 



§ 8. Production of Chemical Compounds by the Effect of Light 

 and Heat. 



When an atom of any gas strikes in its course against an atom 

 of some other gas, the question which presents itself is whether 

 the two will unite to form a single molecule or not. The internal 

 equilibrium of each atom will be disturbed by the impact, so that 

 the resultant of the internal forces of the system formed by the 

 two atoms will in general have a value different from zero. Let 

 this resultant be transferred parallel to itself until it passes 

 through the centre of gravity, as is allowable from a theorem of 

 dynamics, then it will increase its velocity of translation. The 

 total energy of the system must, however, remain constant, so 

 that the energy of the internal atomic vibrations must be 

 diminished by exactly the same amount as that by which the 

 energy of the motion of the centre of gravity is increased. 

 After the impact the internal vibrations will at first be of a very 

 irregular character ; but under the action of the light rays they 

 will ultimately attain a condition of stationary equilibrium, 

 supposing such to be possible with the diminished energy. 

 When it is possible its stability will be greater, the greater 

 the diminution in the internal energy. 



Consider, for example, the formation of hydric chloride gas 

 by the action of light on a mixture of chlorine and hydrogen, 

 accompanied as it is by a measurable development of heat. 

 Both these gases exhibit strong bright lines in the blue portion 

 of the spectrum, and, in the case of hydrogen, also in the ultra- 

 violet. Vibrations of corresponding critical periods will therefore 

 easily be excited, which will greatly increase the internal energy 

 of the atoms. When an atom of chlorine now impinges upon 

 one of hydrogen, they will remain in contact for a finite, 

 though exceedingly short interval. During this interval the 

 mechanical theorem relative to the motion about the centre of 

 gravity is applicable, since there will be no external forces acting 

 on the pair of atoms during their common rectilinear motion. 

 Let it be assumed further that the energy of the molecule formed 

 by the union of the two atoms is, under the existing conditions, 

 less than the sum of their separate energies, viz. that the critical 

 vibrations of the molecule are less sensitive to the action of light 

 than those of the separate atoms, then the spherical atomic shells 

 will tend to execute resultant vibrations proper to the molecule 

 according to § 7, so that the chlorine and hydrogen will unite to 

 form hydric chloride. No energy can of course be lost, so that 

 the difference between the internal energy of the molecules 

 and that of the separate atoms will be added to that of the 

 translatory motion, and will therefore become sensible in the 

 form of heat. 



It will be noted that no special chemical affinity between 

 chlorine and hydrogen has to be assumed, but two elements may 

 be said to have a chemical affinity whenever the energy of the 

 resultant molecular vibration is, under the given conditions, less 

 than that of the separate atomic vibrations. 2 



1 Bunsen's observations (Poggendorff's Annalen, vol. cxxviii.) on crystals 

 of certain didymium salts show that there is actually a difference in the 

 absorption of light in different directions. 



2 A chemical compound may therefore be regarded as produced in a manner 

 similar to the variation of a species on the Darwinian theories of adaptation 

 and natural selection. A species undergoes variation such as to increase its 

 suitability to its environment. In exactly the same way two atoms will unite 

 to form a molecule, when they thereby become less sensitive to the influence 

 of their surroundings than they woutd be separately. Accidental conditions 

 are of no more importance in determining the formation of chemical com- 

 pounds, than the voluntary actions of individuals in determining the variation 

 of a species. 



The given conditions may depend on light, heat, or electro- 

 motive force, though the consideration of the last-named may be 

 eliminated (see § 16). An example of the action of heat is given 

 by the formation of water from hydrogen and oxygen. The 

 hydrogen burns with a blue flame. Both the elements give 

 bright lines in the red portion of the spectrum, hydrogen at 6562, 

 and oxygen at 6171, 1 so that their internal energy can easily be 

 increased by the action of heat, so that combination will take 

 place, and this is accompanied by a considerable development of 

 heat. Water being a very stable compound with respect to the 

 action of heat, we should expect it to give chiefly blue lines. 

 This has not hitherto been proved by direct experiment, but it 

 appears to be indicated by the blue Colour and intense heat of the 

 hydrogen flame. 



Since the heat of combustion which is usually developed during 

 the formation of oxides arises from a diminution in the internal 

 energy of the atoms, we should infer that (1) the stability of an 

 oxide will be greater the greater its heat of combustion ; (2) the 

 spectrum of the oxide will not extend so far towards the red end 

 of the spectrum as the spectra of the constituents. 



The former inference is confirmed by the researches of Favre 

 and Silbermann ; the latter is found to be justified for the oxides 

 of aluminium, lead, carbon, copper, and strontium (the ultra-red 

 portion of the spectrum in the case of strontium should be 

 specially noted), but it cannot be expected to hold good so 

 universally as the former. 



§ 9 Molecular Theory of Chemistry. 



In modern chemistry the term molecule is used to denote the 

 smallest mass of a substance which can exist separately. This 

 conception of a molecule is essentially different from that set 

 forth in § 7 of this paper. The chemical molecule may be 

 simply an atom, as in the cases of mercury and cadmium, but 

 this is not the case for the molecules considered by the author. 

 On the author's theory, each atom is supposed capable of separate 

 existence, which agrees with chemical phenomena when the 

 atoms are considered in the isolated, or so-called nascent condi- 

 tion, but appears to be in conflict with them in that Mariotte's 

 (Boyle's) law, and the comparison of the weights of equal 

 volumes of various elements in the gaseous state, appear to point 

 to the conclusion that their chemical molecules consist of two or 

 more atoms. 



This only applies to elements in the gaseous state and under 

 the ordinary conditions of pressure and temperature, and it is 

 quite conceivable that in high vacua and at a high temperature, 

 as for example in a Geissler tube, the atoms of diatomic molecules 

 may exist separately, a dissociation taking place similar to that 

 which is invariably found to occur in the case of chemical com- 

 pounds under similar circumstances (see § 10). The ordinary 

 hypothesis must therefore be regarded as simply expressing that 

 under ordinary circumstances the atoms of diatomic molecules 

 tend to unite in pairs to form chemical molecules. 



According to § 8, it must therefore be assumed that the diatomic 

 molecules of certain elements are less sensitive to the external 

 influences of light and heat than the separate atoms, and that 

 the internal energy of such a molecule is less than the sum of the 

 internal energies of its two constituent atoms. Suppose that £ is 

 again determined by (2) and that xi = m cos 2-irt/T, then the 

 quantities a; must be determined from the equations (25) and 

 (24). The internal energy of an atom will therefore be 



E = ■!(*«! a, 2 + m 2 a. 2 * + . . . + mj+iaj + i 2 ). 



The energy of a second atom of the same substance under 

 identical external conditions will have the same value. If the 

 two atoms are placed in contact, the new values of xi must be 

 determined from (25), (26), and (27). In this case, however, we 

 have yi — xi, ai — bi, a = ei, mi = m t so that (26) and (27) 

 become identical, and (27) reduces to (24), with the distinction, 

 however, that the quantities xi now represent the displacements 

 relatively to the common centre of gravity, instead of relatively 

 to the centre of gravity of the single atom. It therefore follows 

 that, approximately, the critical vibration periods of a molecule 

 consisting of two similar atoms will be identical with those of the 

 separate atoms. 



Now the energy of the molecule is just double that of either 

 of the constituent atoms, so that the union of the atoms cannot 

 be due to a decrease in the internal energy. It is easy to under- 

 stand, however, that when two atoms have once combined they 



■ Se: B;A. Reports, 12:4, x38s, en I :886. 



