126 



REPORT— 1882. 



several times come nearer and nearer together towards the ultra-violet, 

 while at the same time the members of each group also approach, is very 

 suo-crestive, and promises to furnish a safer basis for future research than 

 the hypothesis of harmonic ratios. As another example illustrating the 

 same fact, we may mention the absorption spectrum of iodine, where the 

 distance of fluted bands of each set decreases with the wave-length. 



There is one more fact which ought to be mentioned. The fluted 

 bands in spectra are often at fairly equal intervals from each other, 

 but a curious change and transformation seems sometimes to occur. 

 Fig. 1 is not intended to represent any particular spectrum, but simply 

 to represent this transformation ; to the left a series of bauds are seen 

 which have been denoted by I,, lo, &c. ; but to the right of I4 springs 

 up another faint band, II,, which, being repeated, gradually gains in 

 intensity until finally, by the side of Ilg, the band of the first series is no 

 longer visible. 



A third series of bands, III,, III2, &c., springs up again to the 

 right of the second series, and in its turn overpowers it in intensity. 

 Those who are familiar with fluted band specti'a will easily call back to 

 their minds the single bands like I,, the bands with weak companion 

 like I4, or the double and treble bands like I7 or I9. Thalen has 



Fig. 1. 



2 



a i 



j^m, j^u^ l^n^ ijH^ J^jr^m, IsH^, J,^,m^ u^^ 



pointed out how the absorption spectrum of iodine is composed of several 

 such overlapping spectra. The bands in this case shade off" towards the 

 red, and the additional bands always spring up towards the violet. 

 The bands approach each other very rapidly as they approach the violet 

 end. The diffei-ences between corresponding bands of the second series 

 are always smaller than between those of the first series, so that the 

 distance between two adjacent bands of the first and second series be- 

 comes larger towards the red ; the same holds good for the other series. 

 Mr. Lockyer has noticed in his jjhotographs of the spectrum, which appears 

 in the electric arc and seems to be due to nitrocarbon, that the least re- 

 frangible of the ultra-violet bands does not seem to correspond with the 

 least refrangible, but with the second band of the violet series, so that here, 

 apparently, a similar change has taken place, an additional band having 

 sprunf up in the least refi'angible side. It is much to be desired that 

 such changes should be carefully examined in each case, as they may 

 lead to most valuable results. 



The tendency to form fiutings is very remarkable. We have first the wide 

 fluting formed by the hydrogen lines ; we have next the narrow flutings 

 of band spectra, and this again, as in the case of iodine gas, approaching 

 each other indefinitely as they go towards the violet, seems to form a 

 fluting of the second order — a fluting of flutings. In the overlapping 

 bauds of different series we may even recognise perhaps flutings of the 

 third order. 



