H. Crew — Rotation of the Sim. 153 



sun exactly on the slit, and to compensate for any slight error 

 in placing the heliostat or in the rate of its clock. The levers 

 by which these motions were accomplished were placed con- 

 venient to the observer at the eye-piece. The collimator and 

 telescope each had an achromatic objective of 6^ inches 

 diameter, made by Prof. Hastings ; the former had a focal 

 length of 1 feet and 3 inches, the latter of 1 feet and 10 inches. 

 The angle made by their optical axes was 12°. Both were 

 firmly attached to a long heavy cast iron frame. For the pur- 

 pose of holding and rotating the grating, this iron frame car- 

 ried upon it the tripod, circle (14 inches in diameter), and plat- 

 form of a spectrometer. This whole system rested on a solid 

 brick pier supported on heavy beams. These beams rested 

 upon two partition walls of the new physical laboratory, mak- 

 ing altogether a rigid, convenient, and accurate instrument, in 

 which the relative position of the collimator, telescope, and 

 grating, could be maintained perfectly constant. 



The eyepiece of the telescope was fitted with a micrometer 

 screw by Grnnow ; in its focal plane was fixed a very thin ver- 

 tical glass scale, ruled in half millimeters. By this means the 

 width of the narrow band of the spectrum during any one 

 reading was measured, and a slight correction corresponding to 

 this overlapping of the sun's image on the slit was introduced 

 into the final value of each set of observations. The grating, a 

 plane one ruled by Prof. Rowland, was 4 inches in length and 

 had 14436 lines to the inch. The spectrum of the fourth order 

 on one side gave superb definition, widely dividing b 3 and h t 

 and the lower component of E. The definition in the fifth 

 order was, of course, not so good — though better than most of 

 the Rowland gratings ruled with this number of lines ; but in 

 this and higher orders it was found that about as much was 

 lost by poorer definition as was gained by increase of displace- 

 ment. 



To determine the angle which the slit of the spectroscope 

 makes with the projected solar axis, the following method was 

 suggested by Prof. Rowland. Between the slit and the con- 

 densing lens was inserted a Steinheil prism mounted in a brass 

 tube with a divided head. This tube was so placed as to have 

 its axis in the prolongation of the optical axis of the collimating 

 telescope, and was made so as to rotate about this axis carrying 

 with it the prism. The prism was placed with its refracting 

 edge perpendicular to the ray. Before the mirror of the 

 heliostat was suspended a fine wire plumb line. The image of 

 this plumb line was brought to focus on the slit plate by a 

 spectacle lens temporarily placed between the prism and the 

 condensing lens. The angle between this image and the slit 

 could now be measured by rotating the prism ; the zero position 



