104 M. Fraunhofer on the Laivs 



I denote the angles at which the rays B, C... of light, mo- 

 dified by the parallel wires, diverge from the axis in the first 

 spectrum which is nearest to the axis, with B^, C^ ..., in the 

 second spectrum with B% C"..., in the third spectrum from the 

 axis with B'", C... and so on. 



From the experiments which are described in detail in my 

 second Treatise, it results, that, if we denote the width of single 

 intervals in the wires with 7, and the thickness of the single 

 wires with 5, expressed in parts of an inch (Paris measure,) 

 we find with all the systems of wires that the arcs of these cir- 

 cles are as follows, supposing the radius = 1 : — 



^, _ 0,0000254.1 And further : 



C' = 



7+ ^ 

 0.00002425 B'' = 2B' 



y+d 



C" = 2C 



angle upon a prism^ then the extreme red ones, after the refraction in the 

 whole width of the prism, likewise diverge all under one and the same 

 angle ; and this must be the case with each of the succeeding kinds of 

 coloured rays. But a perfect object-glass has the property of uniting 

 in one point all the rays which fall parallel to each other under any angle in 

 the focus ; and hence such a prismatic spectrum in a perfect telescope must 

 consist of perfectly homogeneous colours, provided that the white rays pro- 

 ceed from an object of which the apparent diameter is exceedingly small. 

 These circumstances, which in themselves contain nothing mysterious, and 

 very simply follow from the nature of the subject, are often in optical ex- 

 periments too little, or not at all observed ; thus frequent deceptions have 

 occurred in these experiments which have led to erroneous conclusions^ 

 Wliat has been here said applies not only to the phenomena of refraction, 

 but also to those of inflexion, and the mutual action of bent rays on each 

 other, and is the cause why, when the light modified through parallel wires 

 is received on a white surface, nothing is seen of those phenomena which 

 are observed in it by means of a telescope. It will be easily seen that a lens 

 cannot supply the office of a telescope. If we wish on a wall or white sur- 

 face to have a spectrum, the colours of which should consist of perfectly 

 homogeneous light, it would be requisite to receive the light proceeding 

 under a proper angle from a very good prism, with a very good object- 

 glass of long focus, and to place the white surface accurately in its focus. 

 In such a spectrum the fixed lines or streaks would be likewise visible ; the 

 prism and the object-glass, however, must not be small, in order to have 

 sufficient clearness, while the coloured light is sent forth from the white 

 surface in all directions, and only a small part of it can strike the obser- 

 ver's eye. 



