410 lilH.L SYSTEM TECIIMCAL JOVRXAL 



elements slmiild be systems of "atoms" of one or a few fundamental 

 materials is a thoroughly pleasing idea, although at present an un- 

 realized ideal. It is unknown how far our descendants will find it 

 expedient to dissect the atom; hut it is certain that the\- will not be 

 stopped by etymology. 



Another fact about atom-models is tliat they are not always dis- 

 placed by their successors; several may and do persist side by side, 

 each adapted to a certain set of facts and observations. E\ery atom is 

 designed in view of a very small fraction of the available knowledge 

 about properties of matter; and this applies to the latest model as 

 well as the earliest. The chemists of the nineteenth century were 

 most impressed by the immutable weight of matter and In- the laws 

 of chemical combination; hence their atoms were mereK' weighted 

 particles equipped with hooks to catch the hooks of other atoms. To 

 the physicists of fifty years ago the physical properties of gases seemed 

 the easiest phenomena to interpret, and they imagined atoms as 

 rigid elastic spheres with radii of some 10"' centimetre; by the masses 

 and motions of such atoms they explained the pressure, elasticity, 

 viscosity, diffusion and specific heats of gases. The physicists of the 

 next generation attended chiefly to the emission, the refraction, the 

 dispersion of vibratory radiations b\' luminous gases, and concci\ed 

 the atom as a framework hoUling \ibrators, like a belfry with a carillon 

 of bells. This third model is inferior to the second in explaining the 

 properties of gases, inferior to the first in explaining the laws of chem- 

 ical combination; each of the three is superior in its own field to the 

 atom-model to which this article is chiefly devoted, and which in its 

 turn is primarily adapted to a field of its own. Still other atom- 

 models have been de\ised, endowed with other properties, to account 

 for other phenomena; and it is altogether probable that many more 

 will be presented Ix'fore the eventual one is attained, if it ever is. 

 For instance, we ina\' some day behold an atom-model devised to 

 explain the conduction of electricity in solids, ver>' competent in its 

 field and quite unlike these others. In the eventual atom-model 

 the essential (jualities of all of these, and of man\' others, must be 

 happih' combined; it does not matter about the inessential ones.' 



• Now and then an article appears in a physical or chemical journal, in which 

 an oddly unconventional atom-model is proposed to interpret some such property 

 of matter as the tlU-'rmoclcctric effects, or supra-conriuctivity, or valence, or some 

 other with which the Kutherford-Bohr atom-model has not as yet been matched. It is 

 easy for a physicist to ignore sudi articles, on the ground that any model departing 

 from that of Rutherford and Hohr must l>e wrong. This is certainly a nnstaken 

 policy. .\n\ partially competent atom-model deserves to l)e examined with care; 

 its essential features nuist reai>|)ear in the eventual model. But, of course, the 

 essential feature is not always tnc conspicuous one. 



