48 



MICHIGAN ACADEMY OF SCIENCE. 



To show more distinctly how these observations compare with each 

 other, I have taken the mean of the two observations made in the two 

 opposite positions of the circle. These mean erroi-s onght to be the same, 

 if the observations were correct, and if the temperature had no influence 

 on them. These means are given in the following table: — 



Mean. 



4.55" 

 4.30 

 4.08 

 4.30 



+1.31 



Mean 



115° 



+6.68 

 6.47 

 6.78 

 C.39 



+6.58" 



Mean. 



175° 



+8.35" 

 8.48 

 8.75 

 8.64 



+8.56" 



