174 ANNUAL OF SCIENTIFIC DISCOVERY. 







color. When the transmitted light is subjected to prismatic analysis, 

 there is found a remarkably intense band of absorption in the red, 

 besides certain other absorption bands, of less intensity, in other parts 

 of the spectrum. Nothing at first seemed more likely than that, in 

 consequence of some action of the ultimate molecules of the medium, 

 the incident rays belonging to the absorption band in the red, with- 

 drawn, as they certainly were, from the incident beam, were given 

 out in all directions, instead of being absorbed in the manner usual in 

 colored media. 



It might be supposed that the incident vibrations of the luminifer- 

 ous ether generated synchronous vibrations in the ultimate molecules, 

 and were thereby exhausted, and that the molecules in turn became 

 centers of disturbance to the ether. The general analogy between 

 the phenomena exhibited by the solutions of chlorophyll and of quin- 

 ine would lead to the expectation of absorption bands in the light 

 transmitted by the latter. If these bands were but narrow, the light 

 belonging to them might not be missed in the transmitted beam, unless 

 it were specially looked for ; and then the beam might be thus " epip- 

 olized," without, to ordinary inspection, being changed in its proper- 

 ties in any other respect. But in subjecting the light to prismatic 

 analysis, first with the naked eye, and then with a magnifying power, 

 no absorption bands were perceived. A little further reflection 

 showed that even the supposition of the existence of these bands 

 would not alone account for the phenomenon. For the rays pro- 

 ducing the dispersed light (if we confine our attention to the thin 

 stratum in which the main part of the dispersion takes place) are ex- 

 hausted by the time the incident light has traversed a stratum the 

 fiftieth of an inch thick, or thereabouts, whereas the dispersed rays 

 traverse the fluid with perfect freedom. This indicates a difference 

 of nature between the blue-producing rays and the blue rays pro- 

 duced. Now, as the lecturer stated, he felt very great confidence in 

 the principle that the nature of light is completely defined by specify- 

 ing its refrangibility and its state as to polarization. The difference 

 of nature, then, indicated by the phenomenon, must be referred to a 

 difference in one or other of these two respects. At first he took for 

 granted that there could be no change of refrangibility. The refrang- 

 ibility of light had hitherto been regarded as an attribute absolutely 

 invariable. To suppose that it had changed would, on the undulatory 

 theory, be equivalent to supposing that periodic vibrations of one period 

 could give rise to periodic vibrations of a different period, a supposi- 

 tion presenting no small mechanical difficulty. But the hypothesis 

 which he was obliged to form on adopting the other alternative, name- 

 ly, that the difference of nature had to do with the state of polariza- 

 tion, were so artificial as to constitute a theory which appeared utterly 

 extravagant. He was thus led to contemplate the possibility of a 

 change of refrangibility. No sooner had he dwelt in his mind on this 

 supposition, than the mystery respecting the nature of epipolized 

 lio-ht vanished ; all the parts of the phenomenon fell naturally into 

 their places. So simple did the whole explanation become, when 



