STARS IN COMA BERENICES. 429 



Q' = the seconds actually read on the circles at any given 

 date, then will the equation 



give 6" with the sign with which it is to be used in the formulae. 

 Tables can, of course, be constructed for any value of 6'\ 

 having as argument distances in millimeters. This has been 

 done and by their aid the measured coordinates have been re- 

 duced to the position of rectangular axes. The values of Q 

 used were as follows (Q need not, of course, be any one of the 

 readings ; it is best taken so as to make the corrections as 

 small as possible). 



Plate I Q=6(y^ 



n 43^ 



III 23^ 



IV 40^>^ 



V 58^>^ 



VI 9^^ 



VII 20^3^ 



VIII 26^ 



IX 5''y2 



^ •• • 29^^ 



XI WH 



XII 8^ 



XIII i6^^X 



XIV 36^X 



Scale-Value Corrections, Projection Errors, and Devia- 

 tion of the Cylinder from Straightness. — None of these have 

 any appreciable effect. The first is due to the fact that the 

 scale is made of German silver, while the plate is glass. 

 Changes in temperature might, therefore, give rise to unequal 

 expansion, and hence to a change in the scale-value. Dr. 

 Schlesinger ^ in 1897 investigated this question, and his results 

 show that in no case could this change affect my results by as 

 much as o".04. I have therefore felt justified in neglecting this 

 error altogether. 



' See his " Prsesepe " pp. 220-223. 



(89) 



