N54 20. 2(dx, *dx,,tdx, ,*dx, ,*dx,,) - (dx., +dx Tdxj,t'dx, ,ldx + 
22 23 24 25) 
+ (dx,,+dx 
41 a2t94*4 +dx atdx 
ahd, + 2(dx | +dx, ,tdx, 
51! dX55 dx, ,tdx,  idx,,) - 
2dy, 1 9y 45 *dy 4 4*2dy, ;-2dy, 1 dY55*dy, ,*2dy,, -2dy, 1 9Y5» + 
45) 
= 
dy 4 4*2dy 45-2dy 4, -dy ,5*dy, ,*2dy ,,-2dy, -dy , tdy, ,*2dy,, 
  
Nyc y= -Z(dx, *dx,,tdx, ,*dx, ,*dx,,) = (dx,,tdx,,+dx,2+tdx,,+dx,,) + 
+ (dx, *dx,5*dx,,tdx,,tdx,,.) + 2(dx,  tdx, 7 tdx, 3 +dx, 4 dx, 5) 
Q zo ard 
xo YoJo 25 
Q -Q am 
m m. m m. 50a 
Q - 1 
go 50a 
Q rol 
BB 25a 
Q nee 
af 50a 
er 3 Lax) sevi"... v. PN, iig i C vo 
vv] = [dx^]^* (dy^]:- —áom à BI) 5—53—5 | 50 
The standard error of unit weight is 
er FPE 
ST AZ. 
The standard error of the standard error of unit weight is 
S = = 0.118, 
  
2, Measurements of parallaxes. 
  
Two grids of high and known precision are adjusted in the plate holders of the 
Stereocomparator. The differences between the x- and y-coordinates of the 
grids are measured as x- and y-parallaxes respectively. Each parallax is 
determined as the average of for instance three settings. These parallaxes 
can be corrected with respect to the known coordinate errors of the grids.