154 REPORTS ON INVESTIGATIONS AND PROJECTS. 



Unfortunately, the conditions of observation at San Luis at zenith-distances 

 much over 65 ° seem to be very unfavorable, both on account of the defini- 

 tion and the unsteadiness ; in fact, the unsteadiness at all zenith-distances, 

 according to the unanimous testimony of the observers having experience of 

 both places, Albany and San Luis, is distinctly more troublesome at the latter. 



At the same time, the accuracy of the observations at ordinary zenith- 

 distances does not seem to be so much impaired as one might have supposed. 

 This appears to be substantiated by the preliminary reductions already made 

 at Albany, and is in accordance with experience elsewhere, except when very 

 special accuracy in observation of a very few stars is attempted. 



As a general check on the direct observations, about 400 observations by 

 reflection have been secured by the fundamental observers ; but they are not 

 included in the foregoing count. Owing to the prevalence of wind at San 

 Luis the observers are not sanguine as to the quality of these observations. 



COMPUTATIONS UPON OBSERVATIONS. 



Computation of reductions to apparent-place and of tabular correction for 

 refraction have been occupying a large part of the attention of the computing 

 staff at Albany since the beginning of the arrival of the instrumental record, 

 to which allusion has been made. None of this could be sent until several 

 months of observation had been accumulated. 



The Director has given a great deal of attention to test-reduction, as well 

 as to parts of the reduction that are semi-definitive. None of the systematic 

 treatment of the work can be considered definitive until practically the whole 

 of the observations are under consideration. The tests have gone far enough, 

 however, to indicate that the instrument, from month to month, is very steady 

 indeed, and that, although some of the diurnal changes are distinct, they are 

 also in general very small. The determination of the level by the nadir obser- 

 vations seems to be precise in a gratifying degree, so that it is possible to 

 detect a very minute diurnal variation which could not have been very cer- 

 tainly shown with less precise observations. The standard collination appears 

 to be very permanent except when adjustments have been purposely made. A 

 small change with the temperature is closely marked and precisely determined. 



In relation to the azimuth it is not possible to decide upon the true char- 

 acter of its variations. The changes in this coordinate appear to be larger 

 and less regular in the nadir, though not pronounced from month to month. 

 Much labor has been expended by the Director personally upon this subject. 

 His investigations have proceeded far enough to indicate that the outlook for 

 systematically accurate results in this respect is good when all the material 

 of special observation for systematic constants shall be at hand. 



As stated previously, the observations have been tested extensively for 

 their quality in the differential sense. In this respect they appear to corre- 

 spond well with the standard toward which we aimed. This standard did 

 not propose competition with special observations where the highest accuracy 

 has been attained upon comparatively few objects. It aimed at attaining, in 

 the differential sense, good rank among the best modern collections of obser- 



