376 Lyster.— On Difficulties of Meridian Circle Work. 
Considering the point whose initial co-ordinates were v = y= UE 
1 t 1 é 
the new co-ordinates are #’ = —--“°, y/=—=-%, 
Ue 2 ve 
so that, if 45° + @ be the angle which the line joining the new 
ato, 
Wee 
a = 0:000018, gives 0 = 0’°66 approximately. 
Thus, it being assumed that the horizontal diameter of uniform 
temperature has remained fixed in position, the microscope, which 
initially read 45°, would, with the altered temperature, read the 
position as being about 2 of a second of arc less, and the dis- 
crepancy would be increased if, as is conceivable, the horizontal 
diameter, instead of remaining a straight line, bent upwards at each 
end. 
The author noticed that, with the simple law assumed, the 
error would be eliminated by the mean of the four microscopes ; 
and that when working at Dunsink, when the nights were much 
colder than the days, he had usually found a small alteration in 
the nadir point and equator point readings of the circle as the 
night progressed, amounting sometimes, between the commence- 
ment and ending of work, to as much as 2” of arc, which he 
attributed to differences of temperature of the divided circle; one 
part of the circle, about a quadrant, being subjected to the 
radiation of the naked pier, which was usually warmer than the 
night air circulating about the instrument. He thought it would 
be desirable from time to time to apply some delicate temperature 
measurer, such as a thermopile, to the different parts of the circle, 
so that materials could be collected for an investigation of the 
matter. ‘The mere collection of micrometer readings at different 
times, without any information as to the distribution of tempera- 
ture, being manifestly insufficient for the purpose. He was 
gratified to learn that the provision of a papier-maché screen to 
protect the circle was contemplated at Dunsink. 
position a y’ makes with the horizon, 0 = Z which, taking ¢,=1, 
