292 M. Frauenhofer on the Ref ractive and Dispersive Power 
the others, for example, the violet, going to a side at D, did not 
fall upon the prism. In like manner, from the lamp B, it was 
only the violet rays which fell upon the prism H. In this way, 
the prism received from each lamp rays of a different colour, 
setting out from the same point. If the prism H, or the aper- 
ture of the object-glass, was not too broad, some rays of the 
six lamps, for example, those between the violet and the blue, 
between the blue and the green, &c. will not fall upon the prism 
II, but will be entirely wanting. In this case, the spectrum of 
rays passing by the small aperture A, and seen by the prism 
H, and by the telescope of the theodolite* will appear as in 
Fig. 4., where I is the violet , K the blue , L the green , &c., 
and each colour will appear separate. The distances ON, NM, 
&c. will increase as the dispersive power of the glass with the 
same angle of the prism H is greater. Since these distances, 
and the angle formed by the incident ray with an emergent ray, 
may be measured by the theodolite with a great degree of accu- 
racy, it is easy, by means of this mechanism, to determine the 
index of refraction of each coloured ray for every kind of re- 
fracting substance. Above the prism A, at the distance of 1 \ 
feet, I made in the shutter another aperture, in the same ver- 
tical line with A, behind which X placed a lamp, from which 
the prism H likewise received light. The spectrum produced 
by the lamp ought then to appear by the prism before the tele- 
scope of the theodolite, and below the coloured points as P,R,Q. 
The shining orange or reddish line which appears in every 
spectrum of the light of the fire is shewn at R. This line en- 
ables us j in the present case, to be certain, that, on different days 
©f observation, we have always the same colour in the coloured 
points, which would not take place if the table on which the 
lamps are placed suffer the least change in relation to the 
prism. On this account, we ought to place the table so that the 
point N may be always found in the same vertical line with R. 
When this is not the ease* it is easy to bring it to the position 
by the adjusting screws B and C. Since the distance of the 
lamps, or rather that of the small apertures by which the light 
falls on the prism A is invariable, we are sure, on different days 
of observation, to have always the same colour in the coloured 
points. 
