l66 REPORTS ON INVESTIGATIONS AND PROJECTS. 



The remainder of the reductions from apparent to mean place have been 

 finished, and all the corrections entered upon the duplicate sets of right- 

 ascension sheets. The pivot corrections have been entered for both sets of 

 computations. Definitive collimations have been determined for the San 

 Luis observations, and these collimations have been entered on one set of 

 right-ascension sheets as far as series 398. For the definitive determination 

 of the collimation it has been assumed that the collimation is constant over a 

 considerable period of time (not including the change due to temperature), 

 and that such a constant more truly represents the actual collimation than 

 the individual determinations. On this assumption the probable error of a 

 single determination amounts to ±?0I2, the collimation in general exhibiting 

 an excellent agreement. The temperature coefficient was found to be — ^ 005 

 clamp £ at -)- 15° C. The temperature used is that of the thermometer at- 

 tached to the barometer. It has previously been found that there is a lag in 

 instrumental constants, and as the barometer is inclosed in a case, the lag of 

 the change in temperature in the barometer case seems to somewhat approxi- 

 mate the lag in the collimation. 



The level constant, determined from nadir readings, generally represents 

 a smooth curve for any one series of observations. Although there is evi- 

 dently a mean effect of temperature on the level, it can not be employed 

 because of the many anomalous series. 



The correction for the effect feet north or south and that for the correc- 

 tion of Eye and Ear minus Chronograph have been applied to both sets of 

 right-ascension sheets. 



Nothing has been done as regards the application of the magnitude equa- 

 tion, as we are awaiting the magnitudes now being determined at San Luis 

 before entering these corrections. 



AZIMUTH AND CLOCK CORRECTIONS. 



A considerable portion of the efforts of the staff of the Department of 

 Meridian Astrometry during the past year has been devoted to an indepen- 

 dent derivation from the transits of the San Luis observations of clock cor- 

 rection and azimuth, so far as it could be carried at the present time. After 

 correction of the transits for personal equation for magnitude, pivot correc- 

 tion, collimation, and level, the azimuths were derived from transits, at suc- 

 cessive upper and lower culminations, of stars within 13° of the South Pole. 

 The list of azimuth stars included 17 of this description. Each fundamental 

 observer was on duty for one week and results for one culmination were 

 combined with transits by the same observer. 



In reducing the observations, differential corrections were applied to a 

 given transit of a circumpolar star from one culmination to another, preced- 

 ing or following by 12 hours, so that the azimuths are independent of any 

 adopted right ascension of the circumpolar stars. In order to establish a 

 period within which the instrumental azimuth could be regarded as fixed, it 



