114 



DR L. BECKER ON THE SOLAR SPECTRUM. 



After all, it is difficult to say which of the two positions is the better for observing. 

 But the formulas leave no doubt that it is most advantageous to have the viewing 

 telescope as close to the collimator as possible. 



After this digression we return to the first formula, which we shall write 



sm 



(ir*) =: 



n\ 



2a cos 



. n\ C 



c- or ^ sec r 



Instead of x we introduce the angular motion of the zero point of the recording wheel. 

 Let r be the number of revolutions of the recording wheel required to turn the grating 

 from the position x = to x = x , hence 



1296000" M 00 , Mt , 1 „ 



x = ~ 150140 r=[1 ' 8872957]r ' 



180 iff 



thus giving a simple relation between r and X. 



A table was computed for the second spectrum giving X with r as argument, the 

 interval being one-tenth of a revolution of the recording wheel. On account of the 



relation 



x 2 = C - x 1 or t. 2 = c — t x , 



one table suffices for both positions of the grating. The following is an abstract of the 

 table employed : — 



±r. 



X in A.U. 



1. Diff. 



±r. 



A. in A.U. 



1. Diff. 



±r. 



k in A.U. 



1. Diff. 



-380 

 -370 

 -360 

 -350 

 -340 

 -330 

 -320 

 -310 



6055-92 

 5995-76 

 5935-52 

 5875-19 

 5814-78 

 5754-29 

 5693-71 

 5633-05 



60-16 

 60-24 

 60-33 

 60-41 

 60-49 

 60-58 

 60-66 



-310 

 -300 

 -290 

 -280 

 -270 

 -260 

 -250 

 -240 



5633-05 

 5572-32 

 5511-52 

 5450-63 

 5389-67 

 5328-63 

 5267-52 

 5206-33 



60-73 

 60-80 

 60-89 

 60-96 

 61-04 

 61-11 

 61-19 



-240 

 -230 

 -220 

 -210 

 -200 

 -190 

 -180 



5206-33 

 5145-06 

 5083-73 

 5022-33 

 4960-86 

 4899-32 

 4837-71 



61-27 

 61-33 

 61-40 

 61-47 

 61-54 

 61-61 



All the standard lines were first expressed in terms of r, and then compared, in every 

 series of observations, with the observed revolutions of the recording wheel. Had the 

 apparatus been perfect, the differences between calculated and observed quantities would 

 have been constant for the same set, the difference being the correction of the zero point, 

 after allowing for which the foregoing table would have supplied the wave-lengths by 

 interpolation. Owing, however, to irregularities in the mechanism, changes of tempera- 

 ture, &c, the zero point changed usually by a few hundredths of a revolution from line 

 to line. So long as these changes were moderate in quantity it was easy to allow for 

 them by a simple graphic process, but in certain positions of the worm-wheel the correc- 

 tions changed very rapidly. The reducing curve then became so steep as no longer to 



