OF ARTS AND SCIENCES. 



159 



nate changes in the latter, the groups of four or five stars must take 

 but little time. 



All these things considered, it is best to use the ephemerides of 529 

 stars yearly published at Berlin, originally intended for the reduction 

 of the great zones now in progress. I think it altogether likely that 

 this will be eventually the standard time-list for astronomers gener- 

 ally in this hemisphere : at any rate, these places will be kept accurate 

 by continual observation for a good many years to come. They include 

 all stars north of 10° of south declination down to the fourth magnitude, 

 and a selected list of fainter ones to fill gaps. Many of these same 

 stars are also given in the Connaissance des Temps, the Nautical 

 Almanac, and the American Ephemeris ; but these latter have not 

 quite enough of them for rapid work. 



To illustrate how azimuth and coUimation are to be determined 

 and eliminated, I give a scheme of observations actually used at 

 Pueblo, C. T., May II, 1873, by Lieut. E. H. Ruff'ner of the United- 

 States Engineers. The scheme was, in general, agreed upon between 

 him and myself. 



On comparing this with other similar schemes, it will be noticed that 

 here are no stars observed below the pole, and but one south of the 

 equator ; in other words, all are as near the zenith as practicable. 

 INIoreover, every pair of consecutive stars include the zenith between 

 them (this is not absolutely essential, but yet very well), and hence 

 give a definite value of the clock correction if the coUimation be 

 known. 



My own scheme corresponding to the above was as follows : — 



