218 PROCEEDINGS OF SECTION A. 
applications of his hand to the instrument. The one temperature 
recorded was 35°; and if this was the air temperature, it is very 
certain that that of the telescope was much higher, after being 
touched as many as fifty times during the 20 minutes occupied 
by the work. 
In connection with the remarks on temperature, it is a some- 
what significant fact that the values given in the preceding 
paragraph seem to bear a direct relationship to the time occupied 
over the observations, the lowest value being given by the set 
which was the most rapidly made, and the highest by that which 
took the longest time. Arranged in order of time occupied, they 
are— 
Star 2452 ... R=58":552 Star 8914 .... R=58":672 
3)  L6307 2 R=587-634 > 0103 2.) "R58 728 
” he Wee R= 984629 os) 27d) San . R—OSe:806 
So uniform are the progressions in time and value that it is 
difficult to resist the conclusion that there is a connection between 
them, and this is probably through the changes of temperature 
caused by the proximity and touch of the observer. 
DECLINATIONS OF STARS. 
The effect of errors in the adopted declinations of the stars 
observed is clearly seen in the latitude results, and wherever more 
reliable declinations have been applied they have been much 
improved. In connection with this it is apparent that a great 
work remains to be done, in combining the positions given in the 
various star catalogues; and attention may be drawn to the 
reduction, by Dr. H. 8. Davis, of the declinations and proper 
motions of stars required for the observation of variation of 
latitude at New York.* Dr. Davis has brought together for some 
fifty-six stars the work given in 130 catalogues, and so has 
obtained declinations of enormously greater weight than could be 
assigned to those given by any one observatory. For instance, in 
the determination of » Geminorum, the data afforded by no less 
than 1,748 observations has been availed of, yet in the previously 
published catalogues the largest number brought together in a 
form available for general use was 179. Though the observatories 
all over the world have been at work gathering information for 
many years, it yet remains to have the results combined. Rigorous 
reductions of this kind no doubt involve a large amount of labour, 
and it can hardly be proposed to at once enter on such a reduction 
of all the star places. A first step, though, might be made 
by subdividing the heavens into areas of, say, 5 degrees by 5 
degrees, selecting the star in each subdivision which has been best 
* See Contributions from the Observatory of Columbia College, New York. No. 8. 
J. K. Rees, Director. Similar work has also been done by Dr. Auwers. 
