Survey of Northeast Greenland. 125 
By way of example the computation of s( and z( is here 
carried out for the first observation moment. 
By ihe argument a, is here understood the moment indicated 
in the table, which immediately precedes the moment of observation, 
midnight September 25 9: Sept. 25.5 (astronomical date). The four 
arguments for which the values of the function (5 ©, 7 €) have to be 
taken from the table thus become a:1 = Sept. 25.0, a, = Sept. 25.5, 
а:1 = Sept. 26.0 and a; = Sept. 26.5. The moment a, for which 
ihe function has bo be computed by interpolation becomes: 
h97m 
Be > rå ie. — Sept. 25.5 + 0.372, thus n — 0.372 
By means of the data given in the Nautical Almanac the above- 
mentioned table is filled in as follows: 
Argument | sq 4s( ASC д < Ark п < 
| 42) 1/ - / 11 
Sept. 25.0 | 15'21”.76 5" 78 56 le Е 91.90 
25. 5 15.98 + 0'.36 55 55.91 1932 
lea — + 5.42 и. 
26.0 10.56 4.97 + 0.45 36.03 1819 + 1.69 
20:50 05:59 17.84 
n 
9 we now 
— 
= 2 
From f(an) = Up + т Ди. = À U+1 =r 4? uo 
get for n — 0.372 
$ © = 15'15".98 — 0.372 x 5'.42 + Ча x 0.372 x 0.628 x 0”.81 
— INS 
nd = 5555.91 + 0.372 x 19’.88 + Ча x 0.372 x 0.628 x 3”.01 
= 5548.34. 
For the second observation moment in a perfectly parallel manner 
is computed п = 0.396 which in its turn gives 
$ < = 15'13”.78 and 
л < = 55'47”.86. 
The manner of computation of the difference А < + Аж indicating 
the formulæ used will appear from the table below. 
If the approximated longitude, /,, and the other elements of the 
computation had been correct, we should have found 
AC-Ax — 0. 
This, however, has not proved to be the case, and we may now 
take it for granted that the reason for this is exclusively to be found 
