82 





C. J. MERFIELD, 



hence we have 

 therefore 



dl 



da _ dr 

 a r 



os: a = d losr r 



In this manner we obtain d log r which equals d log a, 

 hence obtaining d« from equation (x) or the correction to 

 a of the tables. 



6. Observations. 



The following list of stars was observed between the 

 dates 1905 July 3 and 1905 July 25 inclusive. For most 

 pairs ten observations were obtained ; in some cases, 

 especially those at large zenith distances, as many as four- 

 teen observations were taken. During the period of time 

 occupied in obtaining the data for this investigation, about 

 512 zenith distances and 24 determinations of the nadir 

 were taken, necessitating 2384 micrometer readings and 

 C08 pointer indications. From this number only two obser- 

 vations of zenith distance had to be rejected. 



Weights depending on the definition, and varying from 

 1 to 5, were assigned to each evening's work. These were 

 used in the preliminary calculations and also in the discuss- 

 ion of the final ones obtained from the mean errors of 

 observation. 



The evening's work consisted in obtaining, as far as 

 possible, the circle readings for each star given in the list, 

 the nadir being taken just before and again at the con- 

 clusion of the observations. As stated previously, the 

 nadir readings were not here necessary, but were taken for 

 the object of deducing a value of the latitude of the merid- 

 ian circle instrument; the results of this determination 

 form the subject of a short paper to be later communicated 

 to the Society. 



