Spectra of Metalloids

Abstract

IN a recent number of NATURE (vol. xv., p. 401) I gave a short abstract of a paper by Messrs. Angström and Thalen, on which I should like to make a few remarks. It is known that Plücker first drew attention to the fact that one body may have different spectra, and he seemed inclined to attribute these spectra to different allotropic states of the element. Later on, however, attempts were made to give another explanation of the phenomenon. It is against these attempts that Angström and Thalen chiefly protest in their paper. They use, however, the word element, in a different sense from that in which it is generally used. An elementary body, they say, can only have one spectrum. We are aware that bodies, as iodine and sulphur, can give two spectra, but then the band spectrum is due to an allotropic state, which, from a spectroscopic point of view, behaves like a compound body. It would seem from this and other remarks, that, from a spectroscopic point of view at least, they consider an element to be not only a body which cannot be decomposed into two different bodies, but a body which cannot be resolved into any simpler molecular state. I have no objection against this if it is always clearly understood that our authors include allotropic states under the denomination of compounds. For instance, they lay great stress on the fact that a spectrum of fluted bands is always characteristic of a compound body. According to their definition of a compound this is perfectly correct, for no doubt the band spectrum belongs to a more complicated molecular state, but they cannot bring this argument forward as tending to show that Swan's spectrum of the candle belongs to a hydrocarbon, or that the fluted spectrum of nitrogen belongs to an oxide of nitrogen. The fact simply means that the molecule which gives these band spectra is to the molecule which gives the line spectra as the molecule which gives the absorption bands of iodine is to the molecule which gives the lines of iodine.