SPECTRUM ANALYSIS OF BLOOD. 829 



producing an endosmosis of fluid into the corpuscles, and so causes 

 them to swell, and assume a flat or even biconvex form. 



These opposite changes of shape have been supposed to modify the 

 power of the corpuscles to absorb colored light, more being absorbed 

 when they are swollen, and less when they are shrunk. But, accord- 

 ing to Professor Stokes, this explanation is inconsistent with optical 

 principles, and the change of color is due to a modification in the 

 refractive power of the corpuscles ; in the shrunken state, their refrac- 

 tive power is increased, and, accordingly, a larger amount of reflection 

 takes place from the surfaces of contact of the corpuscles with the 

 surrounding fluid ; whilst in the distended state, their refractive power 

 is diminished, and less reflection takes place from their surfaces. But, 

 although the brilliant color, produced by the addition of strong saline 

 solutions to the blood, and the dark hue occasioned by diluting it with 

 water, may be thus satisfactorily explained, the natural alterations of 

 color produced in the blood by the respiratory changes cannot be so 

 accounted for ; though venous blood is of somewhat less specific gravity 

 than arterial, yet there is no evidence of its containing fewer salts ; 

 moreover, direct observation has failed to detect any difference in form 

 between the corpuscles of the two kinds of blood ; and lastly, the in- 

 adequacy of such a purely physical explanation is proved by the fact 

 that, even when the red corpuscles are entirely dissolved, or when 

 pure solutions of cruorin or the coloring substance of the blood, are 

 employed, precisely similar changes in color ensue, from alternately 

 agitating them with oxygen, and carbonic acid, in the former case the 

 color being brightened, and in the latter rendered dark. The nature 

 of the changes thus induced in the cruorin of the blood, has been re- 

 vealed by the photo-chemical discoveries of Hoppe and Stokes, in 

 which the so-called spectrum analysis is employed, to detect most 

 recondite changes in the cruorin. 



The formation of the prismatic solar spectrum, by passing a beam of sun- 

 light through a prism, has already been explained (p. 431). In this spectrum, 

 when sufficiently magnified, it has long been observed, that numerous, fine, 

 dark lines exist, the lines of Frauenhofer ; these are owing partly to the pres- 

 ence of vapor in the air, which refracts some of the light, but chiefly to the 

 absence, in the light examined, of luminous rays of certain degrees of refrangi- 

 bility ; the consequence of which is, that some parts of the spectrum are left 

 unoccupied by any light whatever. In the solar spectrum, Frauenhofer de- 

 scribed 80 dark bands or lines ; but 2000 are now recognized. Light obtained 

 from different sources, as by the combustion of different substances, or ordi- 

 nary light first passed through transparent bodies, solutions, or even through 

 the vapors of volatile substances, or proper gases, either colorless or colored, 

 and afterwards transmitted through a prism, also forms a spectrum ; but on 

 comparing the magnified spectra of different substances, it is found, in many 

 of them, that the dark bands differ in number, position, width, and intensity ; 

 and, moreover, that in the case of certain lights, which are colored, color bands 

 of different position, number, width, and intensity, make their appearance. 

 The yellow color band of sodium is a remarkable example of this. 



The dark bands, sometimes called absorption bands, and the color bands, be- 

 ing characteristic and constant, for certain substances, they constitute most 

 delicate means of detecting, and discriminating between, such substances. 

 This is done by the spectroscope, an instrument consisting essentially of a tube 

 with a slit at one end, a prism at the other, and a small magnifying glass with 



