592 



DR. GOLDSCHMIDT ON THE OBSERVATIONS 



of the same month in the succeeding year, we obtain the secular 

 variation, which has the same signs in the whole thirty-six re- 

 sults. 



We find in these values, that the secular decrease, derived 

 from the afternoon observations, is greater for all the three years 

 than that which results from the forenoon observations ; this 

 circumstance is connected with the decrease of the difference 

 between the forenoon and afternoon observations in these three 

 years ; in the two first years the contrary was the case. 



If we now combine the yearly mean values of the secular 

 change derived from both forenoon and afternoon observations, 

 we obtain the following values of the mean decrease : — 



We will investigate subsequently the law of the secular de- 

 crease of the declination, as far as our observations enable us 

 to do. 



The following table shows the mean values of the secular 

 variation for the several months : — 



