ALTAZIMUTH SOLAR OBSERVATIONS. 345 
with normal eyes! It is worthy of remark however that the 
precision of the estimation of tangency depends upon the direction 
of the apparent motion of the sun’s image, being more perfect 
when the tangency is estimated at the last, than when at the first 
contact, 7.¢., the advantage is for the positions illustrated in Fig. 4, 
when the apparent motions of the sun take place in the directions 
of the arrows. It is therefore desirable to secure when possible, 
similar conditions in morning and afternoon observations. 
The following matters apply to any method of altazimuth solar 
observing, and are important :— 
(i) The instrumental adjustments should be well made so that 
the outstanding corrections shall be small. 
(ii) The level tubes upon which any corrections depend should 
never be allowed to remain partly in sun and partly in shadow, 
because under such circumstances the axis of the level, upon the 
constancy of which all confidence in its indications depends, 
changes. The value of the divisions of the level fluctuate, and 
are subject to some uncertainty even under the best attainable 
conditions. 
(iii) The effect of the unequal heating of the metal which intro- 
duces sensible error in the larger instruments, should, as far as 
possible, be minimised by the scheme of the manipulation, that is 
by reversions, presentation of different sides to sources of heat, etc. 
(iv) The effect of a very slight movement of rotation of the 
stand of the instrument, due to heating, warping etc., when 
instruments are not suitably protected by observing tents, should 
be eliminated by reading the direction of the ‘referring object’ both 
Ug pierre SS 
1 If there be any astigmatism this will not be true, and if the astigmatic 
fect be serious, it may greatly prejudice measurements in one position, 
Usually the vertical meridian of the cornea, at which the observer’s 
optical system may be considered to commence, has a somewhat shorter 
radius of curvature and therefore a shorter focus, than the horizontal 
ian. Observers who aim at a high degree of precision may not find 
me 
it disadvantageous to test their vision for the measure of its astigmatic 
ection. 
