306 I. P. Косн. 
Therefore one promptly reads off on the sliding rule D, — 13.7 km 
corresponding to division 462 on the fixed rule. The correction for 
refraction and curvature of the earth now again becomes 12 m (more 
accurately 11.7m). Thus D — D, — 13.7 km. 
One will see that at the computation it is not necessary to record. 
When the index is put at the division corresponding to tan v one 
has the distance looked for in a fraction of a minute. 
Example 2. In the quite extraordinary case from Kap Bremen 
(mentioned on p. 300) the altitude is given as 1008.3 m, the zenith 
distance as 91°06’20”.3. As one is here dealing with distances of 
_more than 40km, the simplest way to find the correction for re- 
fraction and the curvature of the earth is to take the correction for 
half the distance and afterwards multiply the latter by 4. 
If H is put 1008m, v — 1506'.3, tan v = 0.0193, th ein des 
placed on 193, after which the computation becomes as follows: 
H,= 1008 m; О, = 52 km; Refr. andcurv:oftheearth— 4 < 48 = 172m. 
H,— 1180 -;D,=6) =; — = — =4 x60 — 240 
H,1948 -; D,=65 » — — — =4x67 — 268 - 
1976-662, — = — =4x69 —276 - 
Н.— 1284 -; D.—665- ; — = — | =4x70 =280- 
E1938) 66850 = —  —4x 70.2= 281 - 
- H,— 1289 - ; D,— 66.8- ; 
D = D, = 66.8 km. 
In the last mentioned example it has been necessary to compute 
six different fictitious altitudes; but it will appear that one might 
have saved a good deal of the work. The first computation, by 
which D, is determined, only takes place in order that one may get 
an approximate idea of the distance looked for, and through it of 
the fictitious altitude corresponding to the distance sought. As it is 
given that D> D,, it was not necessary to select the correction of 
refraction and curvature of the earth, corresponding to D,, but 
starting from the knowledge of the value D, one might immediately 
have tried to estimate at the value D and after that to have taken 
the approximate correction of refraction and curvature of the earth 
corresponding to the estimated value of D. When, as in the case of 
the Danmark-Ekspedition, the object is to compute several thousand 
distances, one attains considerable skill in forming an immediate 
and pretty correct estimate of the distance sought. It evidently does 
not seem to matter that one flies too high in making an estimate of 
this kind. 
The last example I will compute once more to show how the 
computation would presumably be performed by a trained computator. 
