Mr. W. Galhraith on some Astronomical, Sec. 127 



Article VII. 



On some Astronomical Methods of Observation. By William 

 Galbraith, A.m., Teacher of Mathematics, Edinburgh. 

 {Continued from page 45.) 



II. — REMARKS ON THE METHODS GENERALLY EMPLOYED IN 

 MAKING CIRCUMMERIDIAN OBSERVATIONS. 



When the smaller instruments of astronomy are em- 

 ployed by the method of repetition, it is of importance to ob- 

 servers to be aware of the limits within which their obser- 

 vations ought to be restrained, so as to insure the requisite 

 accuracy. This is the more to be insisted upon, as some 

 authors seem unconscious of the limits to which observa- 

 tions, under given circumstances, ought to be restricted, 

 and unacquainted with the degree of accuracy resulting 

 from the use of different tables in the hands of the public. 



The usual tables of reduction are generally formed by 

 throwing the expression derived from the principles of 

 spherical trigonometry into a series of two or three terms. 

 In general, however, when it becomes necessary to embrace 

 more than one, or at most two terms, besides the proba- 

 bility of introducing other errors, the application of a series 

 is more troublesome than the direct computation by sphe- 

 rical trigonometry, and to avoid these, it becomes neces- 

 sary to select objects which, by their situation with respect 

 to the observer, are convenient and proper for such a mode 

 of observation. 



In general, it may be remarked, that objects near the 

 zenith, though the most eligible for zenith sectors, or 

 mural circles, are disadvantageous for smaller instruments, 

 such as Borda's repeating circle, or other portable altitude 

 and azimuth circles, when the observations are repeated a 

 considerable number of times near the meridian. For the 

 use of the latter class of instruments, a considerable zenith 

 distance is necessary to obtain the requisite accuracy, for it 

 will be found, by direct calculation, that when the latitude 

 is 30°, the declination 20°, of the same name with the lati- 

 tude, and consequently the meridian zenith distance 10°, 

 that even Delambre's formula embracing these terms gives 

 results erroneous to the amount of 47" in excess, if the 

 horary distance from the meridian, when the observation 



