Absorption Spectra. 227 



than can at present be claimed for it. If half-a-dozen 

 careful investigators, without specifying conditions as regards 

 quantity of substance used, &c, were asked to supply a 

 wave-length spectrum of any absorption fluid or solid, the 

 probability is that no two would be alike. And yet each 

 might be accurate. The explanation of such discrepancy 

 introduces yet another element of uncertainty in absorption 

 spectroscopy, and that an important one, namely, the 

 position and even the apparent existence of absorption 

 bands vary with the amount of the absorbing medium. A 

 glance at the accompanying diagrams (Plate I.) of the 

 absorption spectra of human blood, serves to show the 

 variation in the appearance of the spectrum when the 

 light is made to pass through varying amounts of blood. 

 In the upper spectrum, where the blood film is the thickest, 

 no characteristic absorption bands are present ; the same in 

 the second ; whilst in the third the true characteristic bands 

 of oxyhemoglobin are apparent. These continue to decrease 

 in width, as the medium becomes thinner, until, in the lowest 

 spectrum, there are no absorption bands apparent. Now, an 

 observer, according to the thickness of blood film examined, 

 would be justified in citing any of these spectra as that of 

 blood. He would, of course, probably instance either the 

 3rd, 4th, 5th, or 6th ; but the 1st and the 7th are also spectra 

 of blood, for, though no absorption bands are perceptible in 

 the lowest spectrum, the film, through which the light passes 

 which forms it, is very perceptibly coloured when examined 

 by the eye alone. I have cited blood as an example, 

 because of its general importance ; but the same is true of 

 chlorophyll, bile, and in fact any absorbing fluids. The 

 fact that the appearance of the resulting spectrum is de- 

 pendent on the amount, or thickness, of the absorbing 

 medium is of extreme importance, since quantity is a 

 factor that is always present. To ensure a certain amount 

 of uniformity in results, it is customary for the observer to 



