94 I. P. Косн. 
leather in readiness to wipe the oculars and other parts of the instru- 
ments; likewise one may very considerably diminish the rime of the 
air exhaled by tying a silk kerchief triangularly folded over the nose 
and binding it firmly at the back of the head. The moisture is then 
condensed on the inner side of the kerchief, which becomes completely 
covered with ice. It is rather disagreeable to have this ice sheet 
dangling before one's face, but one has to put up with it. 
The danger of the instruments becoming covered with rime is 
by the way not dependent upon the temperature only. As a general 
rule it may be stated that when the moisture in the air exhaled 
condenses in the atmosphere, one must be particularly careful; if 
one cannot see the mist rising from the air exhaled, a kerchief is 
hardly necessary. 
The worst of all is when the rime settles on the horizontal axis 
of rotation under the legs of the striding level, because the level 
then indicates an erroneous inclination of the axis. This also takes 
place, as soon as the oil freezes, which happens at about —- 20° 
(Centigrade). But already at an earlier period it becomes thick and 
at the vertical movement of the telescope gathers in streaks along 
the generators of the horizontal pivot. These streaks also make the 
surface of the pivots uneven, and this, in its turn, renders the readings 
of the striding level illusory. In order to guard oneself against this 
the oil must, before any observations are made, be carefully wiped 
or washed off the pivots and the V’s; in the same manner the feet 
of the striding level must be wiped. In spite of every precaution 
ice crystals will, however, now and then settle on the pivots, and as 
the latter nearly always from there pass on to the striding level and 
remain there, it may be necessary to wipe the feet of the level every 
time it is reversed. 
The level is also in other ways influenced by the cold. Only 
by holding one’s hand over the one end of the level does one in- 
fluence the position of the bubble, in that the heat emanating from 
the hand occasions the expansion of the metal frame. Hand and 
wrist must consequently be covered with mittens and wristlets. 
As a matter of course it is not convenient to use an ordinary 
hand lantern for the purpose of reading the level. A small electric 
hand lamp is better, because the radiation of heat is here so much 
smaller, but the level also reacts upon such a lamp, if lighted for 
too long a period or at too short a distance. 
The lack of a meridian mark generally caused me to use transits 
on the vertical of Polaris for the determination of time. But by this 
very method the accuracy of the determination of time is in the main 
dependent on the accuracy with which the inclination of the axis 
