Note 37 — Chapter XIII · POTASSIUM, RUBIDIUM, CÆSIUM, AND LITHIUM. SPECTRUM ANALYSIS

The importance of the spectroscope for the purpose of chemical research was already shown by Gladstone in 1856, but it did not become an accessory to the laboratory until after the discoveries of Kirchhoff and Bunsen. It may be hoped that in time spectroscopic researches will meet certain wants of the theoretical (philosophical) side of chemistry, but as yet all that has been done in this respect can only be regarded as attempts which have not yet led to any trustworthy conclusions. Thus many investigators, by collating the wave-lengths of all the light vibrations excited by a given element, endeavour to find the law governing their mutual relations; others (especially Hartley and Ciamician), by comparing the spectra of analogous elements (for instance, chlorine, bromine, and iodine), have succeeded in noticing definite features of resemblance in them, whilst others (Grünwald) search for relations between the spectra of compounds and their component elements, &c.; but—owing to the multiplicity of the spectral lines proper to many elements, and (especially in the ultra-red and ultra-violet ends of the spectrum) the existence of lines which are undistinguishable owing to their faintness, and also owing to the comparative novelty of spectroscopic research—this subject cannot be considered as in any way perfected. Nevertheless, in certain instances there is evidently some relationship between the wave-lengths of all the spectral lines formed by a given element. Thus, in the hydrogen spectrum the wave-length = 364·542 m2/(m2 - 4), if m varies as a series of whole numbers from 3 to 15 (Walmer, Hagebach, and others). For example, when m = 3, the wave-length of one of the brightest lines of the hydrogen spectrum is obtained (656·2), when m = 7, one of the visible violet lines (396·8), and when m is greater than 9, the ultra-violet lines of the hydrogen spectrum.