306 ANNUAL OP SCIENTIFIC DISCOVERY. 



others ; but I do not know that it has yet been put in practice 

 elsewhere. 



" The prisms, for protection and convenience of handling, are set 

 in frames of blackened brass. They are arranged around the cir- 

 cumference of a hollow cylinder of elastic gun metal, ok inches in 

 diameter, with stout flanges above and below, between which they 

 are clamped by little thumb-screws, so that they can be readily 

 removed or transposed ; it requires less than a minute to put the 

 last prism with its reflector in place of any other of the train, thus 

 reducing the dispersive power to any extent desired. 



" No particular care is required in placing the prisms, as a couple 

 of narrow flanges were cast upon the cylinder near the top and bot- 

 tom, and afterwards planed off to form true bearings for the backs 

 of the prisms. They are thus always correctly set by being simply 

 slid home before tightening the clamping screws. 



" The lower flange of the cylinder is attached to the base-plate by 

 a screw directly under the middle of the front face of the first prism. 

 Around this point, as a centre, the whole system of prisms is mov- 

 able by means of a double-threaded tangent-screw, which brings 

 the different portions of the spectrum into the field of view. 



'* The adjustment of the prisms to their angle of minimum devi- 

 ation is effected by a method devised by Mr. George Clark, which 

 is exceedingly simple, and, if not theoretically exact, answers every 

 practical purpose. The flanges between which the prisms are 

 clamped are sawed through between the prisms, and a portion of 

 the cylinder, flanges and all, equal to an arc of about 30, is cut out 

 between the first prism and the last. On closing up or spreading 

 open this gap by means of a suitable tangent-screw, the circumfer- 

 ence of the circles around which the prisms stand is correspond- 

 ingly enlarged or diminished. Probably, when the ends of this 

 opening are drawn very near together, or spread very far apart, 

 the cylinder is somewhat distorted, and a corresponding mal-ad- 

 justment of the prisms results ; but if so, the effect is very slight. 



" The instrument gives a perfect view of every part of the spec- 

 trum from below A to H ; above h, however, when all 7 prisms 

 are used, there is a loss of light occasioned by a partial obstruc- 

 tion of the apertures of the collimator and telescope by the corner 

 of the reflecting prism. 



"Were it important to secure the perfect cylindricity of the prism- 

 frame through the whole range of adjustment, it could be easily 

 done by merely fastening at the back of each prism a radial bar 

 acting upon a central pin, as in the arrangement first devised by 

 Mr. Ilutherfurd, and since adopted by Mr. Browning, in his auto- 

 matic spectroscope. 



" This plan of Mr. Clark's, doing away with all joints and hinges, 

 has the great advantage of perfect firmness and solidity in every 

 position of the instrument, an advantage hardly to be overrated 

 in an astronomical spectroscope. 



"Had it occurred to me in season I might have made the instru- 

 ment still simpler, firmer, and perfectly automatic in its adjust- 

 ment, by merely substituting for the first prism a half prism, like 



