of Light and its Theory. lOS 



minated in my Treatise perfect mean spectra^ but in future I 

 shall call them spectra of the second class. On the contrary, I 

 name these latter spectra imperfect ones if their light be not 

 homogeneous ; and that is always the case when only a few 

 reflected rays act upon each other at equal distances. 



In this paper the capital letters B, C... H will denote cohur- 

 ed rays of different kinds ; B is a red ray, which lies towards 

 the extremity of the spectrum ; C is deeper in the red ; D 

 orange ; E green ; F blue ; G indigo ; and H violet. In every 

 spectrum from solar light that consists of perfect homogene- 

 ous rays fixed lines or streaks are found, which distinguish 

 themselves either by their strength, or their position, from the 

 other numerous fixed lines of the spectrum. * 



• In the meanwhile, it must not be assumed that the spectrum which 

 is obtained, when the light refracted through a prism is received on a wall, 

 or on a white surface, consists of homogeneous light. Such a spectrum 

 would consist of perfectly homogeneous colours if the surface on which it 

 is received were at an infinite distance from the prism, or if an exceeding- 

 ly small prism could be employed. In both cases, however, the light would 

 be extremely weak, and therefore no spectrum could be seen. If, for in- 

 stance, the prism is two inches wide, then the extreme red rays in the re« 

 ceived spectrum must likewise be spread into a space of about two inches. 

 The next red ones must occupy an equally large space, and they must, for 

 the most part, fall into the former ones. The same must take place with the 

 various tints of the other coloured rays. This is the reason why all the 

 coloured rays through such a prism necessarily must appear mingled toge- 

 ther, the less so, however, the further from the prism the spectrum is re- 

 ceived, or the smaller the prism, but never in so small a quantity that 

 they could present homogeneous colours which have still intensity enough 

 to be perceived. The colours of the spectrum are, according to this, so 

 much less developed the nearer we approach the white surface to the 

 prism ; and here we see at the same time the cause why we obtain entirely 

 white light when this surface is nearest to the prism, viz. because then 

 the extreme violet rays fall into the extreme red ones, and the various co- 

 loured rays are throughout intermixed. Should the white light falling 

 upon the prism proceed from a luminous surface, then the colours for the 

 same reasons could not be homogeneous, even if they were produced by 

 an infinitely small prism, or received at a very considerable distance. The 

 apparent diameter of the aperture, or that of the object from which the 

 whita rays proceed, must be exceedingly small. Thus, for instance, the 

 apparent diameter of the sun is far too considerable for the colours of the 

 rainbow to be perfectly homogeneous. If, on the contrary, the white rays 

 fall under a certain angle of inclination, and all together under the same 



