ASTRONOMY. 3G7 



in a somewhat different manner by Briinnow (Splicer. Aatr., 4 Anil. p. 

 252). 



" If tliere are within the limits of the stellar system systematic differ- 

 ences in dl, the latter is certainly not a function of p as well as of r; 

 the error resulting from this in the foregoing equations grows smaller, 

 however, as r and p grow larger in comparison with e, and therefore in 

 this respect also it is an advantage to' determine the constant of pre- 

 cession from faint stars. Moreover it is plain to see, if the supposed 

 rotation of stars in the plane of the Milky Way actually takes place, that 

 neither the precession nor the proper motion of the sun can be deter- 

 mined independently of other assumptions or conditions from the fore- 

 going equations. The equations permit three combinations only of the 

 quantities d ^, di, Q, , and i, and instead of obtaining from them the three 

 components of the sun's motion, c cos r cos <r, c cos r sin <7, c sin r, we 

 obtain only the excess of these components over the similar ones ob- 

 tained from the universal rotation.* 



•' If dl, then, is at all appreciable, we are possibly still far removed from 

 a knowledge of the real apex of the solar motion (especially if, as is 

 credible for many reasons, n is not small as compared with the average 

 p.) What has been considered in the treatment of the problem thus far 

 as the components of precession m and re, is in reality something else, 

 namely, respectively, 



/= m + cos i dl and 



(j = n -f cos Q> sin i dl. 



and aside from this the third component of rotation h = sin Q> sin i dl 



can be determined. The general precession cannot then be kept pure, 



but is mixed with that component of dl which is parallel to the ecliptic. 



"As Nyr^n and Dreyer have restricted themselves to the determination 

 of ?», they have from the first neglected the investigation of a possible 

 appreciable value of dl. This possibility can only be substantiated by 

 determining g aud h with /, which, in the equation for d a, cannot on 

 account of the quantity tang. <3, be obtained from the right ascensions 

 of equatorial stars, but on the other hand only from their declinations. 



" If dl is appreciable, a value of h must result which is not zero, unless 

 sin Q, or sin i should equal zero. But if this is the case, the coefficients 

 of dl in/ and g cannot, at the same time be zero, aud the influence of 

 dl will be shown in this, that the constant of precession obtained from 

 / will not coincide with that which may be derived from g. It is only 

 in case the plane of rotation of the fixed stars coincides with the eclip- 

 tic, that the rotation would unite entirely and without contradiction 

 with the precession, and would consequently not be recognized as a 



* The determination of motion in the line of sight by spectrum analysis gives no 

 new combinations of the unknown quantities either. If we write the last three lines 

 of the equation for d S — F cos a sin <5 + G sin u sin 8 — H cos S, we obtain the fol- 

 lowing : — dp = F cos a cos 6 -(- G sin a cos S -\- H sin S. 



