MATHEMATICAL SECTION. 41 
17TH MEETING. FEpRuARY 10, 1885. 
The Chairman, Mr, G. W. Hirt, presided. 
Present, eleven members and one guest. 
In the absence of the Secretary the reading of the minutes of the 
last meeting was deferred, and Mr. R.S. Woopwarp designated as 
Secretary pro tem. 
Mr. KuMMELL made a communication on 
AN ARTIFICE SOMETIMES USEFUL FOR THE ADJUSTMENT OF 
CONDITIONED OBSERVATIONS. 
[Abstract.] 
The general process consists in multiplying equations of condition 
by such factors as will extinguish side-coefficients in the normal 
equations for correlates. This was shown to be possible in an infinite 
number of ways. One such way leading to linear equations for the 
multipliers was shown to require for the extinguishment of all the 
side-coefficients the solution of the normals, 7. e., the very work 
which was to be evaded. The method would, however, be advan- 
tageous for the partial extinguishment of large side-coefficients, and 
the normal equations could thus be solved with advantage by Gauss’ 
indirect method. A useful symmetrical rule was given for extin- 
guishing the side-coefficients for a pair of conditions. Illustrations 
of this rule in the adjustment of simple geometrical figures were 
given, beginning with a simple figure of two triangles and extend- 
ing to a complete pentagon. 
Mr. Hixt remarked that Jacobi had proposed a similar method 
for removing side-coefficients. Further remarks were made by Mr. 
Woopwarb. 
The next communication was by Mr. Gustave L. RAVENE on 
THE THEORY OF MERCURY. 
[Abstract. ] 
The method here used of computing the secular variation of the 
elements of an orbit is dueto Gauss. 
