30 CARNEGIE INSTITUTION 



of observation like that described further on under (2) would require 

 this work as its foundation. In fact, there is scarcely a single de- 

 partment of precise measurement in astronomy thai would not be 

 indebted to this work for a part of the data which it needs. 



The problem of observation here suggested naturally divides itself 

 into two sections. The first would be concerned with the observa- 

 tion by fundamental methods, and with the highest precision, of 

 about 600 to 800 of the principal stars ; the second would involve 

 the extension of this work by less rigorous methods to about 5,000 

 other stars — the entire work to be conducted in such a manner as to 

 be systematically consistent in all its parts, and to be a homogeneous 

 whole. 



There can be scarcely a doubt that the ideally best result would 

 be attained through the adoption of the transit and vertical circle, 

 for the observation of the principal stars at least. This, however, 

 would prove somewhat more costly in execution, and the completion 

 of the entire work, if it should be carried through with these instru- 

 ments alone, would be deferred somewhat longer than might seem 

 desirable. 



The advantage of economy would attach to the use of a meridian 

 circle for these observ^ations. This would be increased, if the use of 

 such an instrument that has already been thoroughly tested in the 

 northern hemisphere could be procured, and we believe it can. The 

 labor of a thorough investigation of the errors of graduation and 

 other errors of such an instrument is, in itself, no slight task. 

 What would be requisite for the purpose, here designed, would 

 probably be equivalent to one full year of work by four observers. 

 The degree of accuracy really attainable in the use of the meridian 

 circle should not be sensibly inferior to that for a vertical circle. 

 The distinct advantage of the latter is in the variation of method 

 which it offers. For the present we should advise the employment 

 of a meridian circle for this research, unless the proposed Southern 

 Observatory should be established on a scale which would enable it 

 to maintain for certain lines of observation the highest ideals. 



Whatever the precise methods of obser\^ation may be, the instru- 

 ment, or instruments, employed should be used for at least two years 

 in the northern hemisphere in the determination of the positions of 

 the principal fundamental stars visible there. There would result 

 a peculiar gain in precision through the comparison of observations 

 with the same instrument upon the same stars, made alternately in 

 the two hemispheres, by which certain errors of the instrument and 



