sorted out into groups showing the most orderly rhythm, and such groups are spoken of as "series." As an example, mention may be made of the spectrum of the cleveite gases, which at first sight appears most irregular, but which, in the hands of Runge and Paschen, was resolved into six series, three of which belong to helium. However, Professor Ramsay maintains that all six belong to helium. Sir Norman Lockyer1 has published the evidence in favour of the composite nature of the gases. The intensity of the lines in a series decreases with the wave length. The wave frequency of such a series of lines is given by the formula A+B/n2+C/n1 where n represents the natural numbers from 3 upwards, whilst A, B and C are constants specific for each element. This formula is due to Kayser and Runge. Balmer 2 showed that the hydrogen lines could be n2 most accurately expressed by the formula λ=A n2-4 where is the wave length, n one of the natural numbers from 3 to 15, whilst A is a constant, and has the value 3647-2 Ångstrom units according to Ames' measurements. The following table illustrates this: Roy. Soc. Proc. Vol. LVIII. p. 113, 193, Vol. LIX. p. 342, and Vol. LXII. pp. 55-58. 2 Wied. Ann. (1885), 25, p. 8. The observed wave lengths are those of Cornu,1 and are given in ten-millionths of a millimetre. This is known as the first hydrogen spectrum, and is obtained in a Geissler tube, where the pressure is not too small. For the second hydrogen spectrum the formula does not apply at all. It will be noticed that Balmer's formula is a particular case of that of Kayser and Runge. Rydberg also formulated an expression for the wave lengths of the spectral lines, which is as follows: n = a A where n is the frequency a and b are characteristic constants for a substance, m is one of the natural numbers 1, 2, 3...., and A is a constant for all series, and has the value 109721.6. In a great many cases a line, which appears to be single, is found, on using a greater dispersion, to be a double or even a triplet. Rydberg suggested that these might be remnants of flutings, the other lines of which are too faint to be seen. Kayser points 1 Journ. de Phys. [2], 5, p. 341. out that the elements in the first vertical column of Mendeléeff's table give doublets, those in the second triplets, and the third doublets.1 It appears, therefore, that elements with odd valency give doublets, whilst those with even valency give triplets, and this is confirmed by the triplets of oxygen, sulphur and selenium of the sixth group. The first elements of any group show the series strongest, and these get weaker and weaker with increase of atomic weight. The following examples will illustrate this point. Rubidium and cæsium do not show the weaker second series; gold gives no series; strontium shows a weak second series, and in barium it is absent. There is only one known terrestrial element giving single lines, and it is helium, in the principal series. It might be noted here that Kayser and Runge distinguish three series : (1.) The Principal Series; their pairs are the strongest lines in the spectrum, and are easily reversible. Their vibration difference decreases as "m" increases. (2.) The First Subordinate Series; strong, very hazy pairs of lines, having a constant vibration difference. (3.) The Second Subordinate Series; weaker, but better defined pairs of line, with a constant vibration difference. 1 Each line of a doublet or triplet is connected with the corresponding lines in the other doublets or triplets by the usual formula A-Bn 2 - Cn-4. The vibration difference is a most important constant, and seems to have a definite relation to the atomic weights. The principal series have as yet only been found in the alkali metals; all the other spectra seems to consist of secondary series, with, in some cases, a number of residual lines. From the following table it appears that, for at least the first two groups of elements, the vibration difference (d) is proportional to the square of the atomic weight. In every case the spectrum advances towards the red with increase of atomic weight. From group to group, however, the series advance towards the violet with increase of atomic weight. Kayser suggests that the melting points seem in some way to influence the series formation. The following table shows this: REESE LIBRAS OF THE UNIVERSITY OF CALIFORNIA It is obvious from what has been said that the farther we advance towards the violet, the closer do the lines approach each other, and hence the ultraviolet would seem to offer some fruitful object of research. 1 Professor Hartley 1 investigated the ultra-violet spectra of 22 elements, with most interesting results. By using prisms and unachromatised lenses of quartz or Iceland spar and dry photographic plates, he succeeded in obtaining the whole of the spectrum, from the blue to the ultra-violet, in focus at the same time. His measurements extend between 2λ 4500 and 2000. The elements examined were lithium, sodium, potassium; copper, silver, mercury; magnesium, zinc, cadmium; aluminium, indium, thallium; carbon, tin, lead; arsenic, antimony, bismuth; iron, cobalt, nickel; palladium, gold, platinum and tellurium. A summary of his results is as follows: The spectra of iron, cobalt, nickel and palladium, especially the two former, are very similar. There 1 J. C. S. 41 (1882), p. 84. |
