1886.] On a New Form of Direct Vision Spectroscope. 451 



of linked levers, of which one carries a graduated arc of 9-J inches 

 radius, by which the angle of rotation can be determined. Those rays 

 which suffer no deviation in passing through the train, must all follow 

 a course through the fixed prism parallel to opqr, whatever their 

 refrangibility ; but the angle of incidence at o will be different for the 

 different values of the refractive index. By turning the two movable 

 prisms into the positions shown by dotted lines the angle of incidence 

 at o will be diminished, and a less refrangible ray will follow the 

 course of no deviation. If then the first position of the prisms ABC, 

 DFE, be that for which an extreme violet ray incident in the line Im 

 suffers no deviation, all the less refrangible rays incident in the same 

 direction may be successively brought to suffer no deviation by turn- 

 ing the prisms towards the position shown by the dotted line. The 

 angles of the prisms have, of course, to be adjusted so that the 

 extreme violet ray may suffer no deviation. A simple calculation 

 suffices for this when the refractive indices of the glass employed are 

 known. We have had the fixed prism constructed with the acute 

 angles about 33|°, and the movable prisms with angles about 62°. 

 With these angles, when the movable prisms are so placed that the 

 angles of incidence at m and of emergence at o are equal (the position 

 of minimum deviation), the ray which suffers no deviation is one 

 somewhat more refrangible than K of the solar spectrum. When then 

 the prisms are turned, less refrangible rays are successively brought 

 into the field of view, but no ray much more refrangible than K can 

 be brought into the field. By increasing the acute angles of the 

 fixed prism, or by diminishing the angles of the movable prisms, a 

 longer range can be given to the instrument, but at the expense of 

 some dispersion. 



To prevent light passing directly from the collimator to the 

 observing telescope, a stop HKis placed midway between the movable 

 prisms. 



It will be observed that the fixed prism serves both as a reflector 

 and refractor, the dispersion produced by it being the same as that of 

 a simple refracting prism of 46° (or 186°— 4EBG) in the position of 

 minimum deviation. The dispersion for the extreme violet is there- 

 fore that of two prisms of 62° and one of 46° in the position of 

 minimum deviation. For less refrangible rays the position of the 

 movable prisms is not that of minimum deviation and the dispersion 

 is proportionally increased, so as to help, in a small degree, to correct 

 the inequality of dispersion of the two ends of the spectrum. At the 

 same time the symmetry of the arrangement is maintained for all 

 rays when in mid-field, and sharp definition is secured for all parts of 

 the spectrum. This is a most important character. We find that with 

 our instrument A and H of the solar spectrum are equally well seen, 

 and so are the red and violet lines of the flame spectrum of potassium. 



