282 PROCEEDINGS OF THE AMERICAN ACADEMY. 



and finds no shifts in the region investigated. The focal length of his 

 grating was twenty-one feet, six inches. 



Such an inconsistency in the results of different investigators em- 

 phasizes the importance of taking extreme care in all the details of the 

 experiment, guarding against any possible false shifts and stating with 

 exactness all the conditions under which the investigation is made. 



The objects of the present series of experiments are, then, to attempt 

 to settle definitely the question of the coincidence or non-coincidence of 

 arc and spark lines, to discover the determining conditions in case shifts 

 do exist, and to harmonize, if possible, the results obtained by previous 

 investigators. 



Experimental Conditions ; Construction and Adjustment 

 of the Grating-Mount. 



The conditions under which the present investigation has been carried 

 out are as follows : The science hall of the college is a substantial 

 building located not less than seventy-five yards from a roadway upon 

 which there is heavy traffic ; the foundations of the building are laid 

 solidly upon glacial drift; and the room in which the spectroscope is 

 mounted is situated in the cellar. The grating is a twenty-one-foot 

 Rowland concave, of 15,000 lines to the inch, having a ruled surface of 

 six by two and one-half inches. The grating holder is of brass, and is 

 substantially built. The entire mount is of steel or iron. Seven-inch 

 channel beams form two sides of the triangle, the third is a beam of five 

 inches. The tracks are of two and three-fourths by one-half inch rolled 

 steel. The carriages are massive ; the truss-rod connecting the two is of 

 2^-inch piping ; and three draw-bars conduce to great rigidity. The 

 camera box is of iron and brass, designed by the author, and is especially 

 massive. The photographic plate is held in its place by brass screws. 

 The slit is connected rigidly to the channel beams, and the whole mount 

 is supported by rubber disks on three substantial brick piers. After 

 adjusting the instrument for any definite region of the spectrum, the 

 bases of the grating-holder and camera box were always clamped to the 

 steel tracks by nine iron clamps of various sizes. 



The camera shutter, being supported from the floor, is entirely free 

 from the camera box, and is so arranged that there is absolutely no 

 chance of mechanical jar from the operation of changing the shutter 

 between exposures. The slit mechanism is surrounded by cardboard 

 cylinders which are held free from the slit by supports connected rigidly to 

 the walls of the room. Around these cylinders is tied the black cloth which 



