j= k =o Fin a 19 in mere 
(4) 
Nn — 
Con ae 
= »S #£ v: cos| rm i F 
k=0 ¢ mn to 
ore =O, bo soo, 
In matrix notation, this becomes 
Oe) . = (5) 
As an example, consider table 1. In the first row is given 
the indicated temperature. In the second row, the measured tem- 
perature correction is given. The introduction of these quantities 
into equation (4) gives the set of a 7's in the third row. Putting 
these a ,'s into equation (2), gives a set of 7,'s for the specified 
T+}; Shown in the fourth row. Comparison of the fourth row with 
the second row shows that the fit of the harmonic series to the 
measured values of 7, is exact except for small round-off errors. 
This is to be expected since the number of parameters determined 
is the same as the number of points fitted. 
TABLE 1. VALUES OF Toy VS. T, 7° 
J 0 1 2 3 4 5 
a a eT a ac ie ae 
Tp. 8.62 10. 93 13.05 14.97 17. 20 19.27 
ik +. 32 +. 28 +.10 -.07 -.22 -.31 
a, . 02448964 .31270375 -.0035209249  .0020677841 -. 015928043 . 0002284838 
T - 82000001 — . 28000003 . 10000001 -. 07000018 -. 22000004 -. 31000001 
ee 
24 
