266 I. P. Косн. 
An estimate of this kind, which takes about ten minutes, is often 
only a few tenths of a minute removed from the result, which one 
gets by a complete computation of the series of observations. 
A glance at the series of observations shows that the first pair 
of observations is comparatively far removed from the meridian and 
that the level in the course of the second pair of observations, tele- 
scope right, has shown a rather problematical reading. It is quite 
a normal occurrence for the level not to be willing to be adjusted 
at the beginning of a series of observations, and there might conse- 
quently be good sense in condemning the first two pairs. In this 
place, however, we are going, for the sake of perfection, to treat the 
whole series. 
In the subsequent computation I have taken it for granted that 
the clock correction is unknown and that the moment of the cul- 
mination must therefore be deduced from the observations them- 
selves. This may be done graphically or by interpolation!), in that 
one forms from the three sets of observations, distributed on both 
La = Be Ze b+a 
sides of the meridian а = = ,b = = с a ——, where z 
+Uu, U, + Uy’ ии, 
and u are zenith distances and UGER times expressed in seconds of 
3 UN Dam Dé с 
arc and time respectively. Result: zn, = — D + oe oo 
The graphic method as well as the interpolation presupposes 
that the diurnal arc is divided symetrically by the meridian 9: con- 
stant declination. In latitudes as high as those which we are here 
dealing with, it will therefore be most correct to check the zenith 
distances for the variation of the declination, in other words, to refer 
the declination to the same moment for the whole of the series of 
observations. It is quite immaterial which moment one chooses; 
only the declination itself must be computed for the same moment. 
From the table below the geocentric zenith distances with the 
adjunct 40, designated I, Ш and VI, have been chosen for the com- 
putation of the clock time of the culmination. One might have chosen 
others, for instance instead of III one might have used IV, only I 
should be absolutely avoided, because there is every reason to sup- 
pose that this observation is encumbered with a very gross error. 
Of the values in the table below one gets: 
— 32 +91 
== 2 — = 505 : 22. ASS ale ISK. 
a 633 : BRS b 735 - 0.0286; 
—- 0.0791 2 
Сс —- и —— — + 0.0000579. 
U, = 118381283 + 7™16s — 11545™44s controlled by 
U, = 11549™52s = 411075 — 11h45m45s, 
1) MARCUSE: Handbuch d. geogr. Ortbestimmung, p. 232. 
