62 
How dealt with. 
Itesults. 
line some 35 miles, and the backbone itself, 45 to 50 miles. The 
last is observed very rarely in a year. Even in 1891, when 
clearness was more than 50 per cent, above the average, it was 
eight times at most. 
To make the most of the scanty returns, the nine years up 
to the close of 1891 are here considered. Values of one were 
given to such expressions as “ faint,” visible ” ; two for 
clear,” first line clear,” ‘‘fairly visible,” “distinct.” &c. ; 
and three for “very clear,” “outline very clear,” “both 
lines,” &c. For the months of partial absence allowance is 
made, but July must remain blank, with an assumed value 
between June and August. 
Thus considered, “clearness” has varied from a minimum 
represented by 20 in 1889 to 64 in 1891, with an average of 
39. Of more interest are the following monthly totals for the 
nine years:— 
Jan. P'eb. Mar. Apr. May. June. July. Aug. Sept. Oct. Xov. Dec. Year. 
Values 14 19 56 48 47 40 [421 44 48 32 3 0 391 
Order X. IX. I. 11.^ IV. VII. [VI.] V. Il.i VIII. XI. XII. 
Conclusion. 
How close do 
we approximate 
to absolute val¬ 
ues ? 
Summary. 
Thus, March and September, the equinoctial months, form 
maxima, though April and May are practically the same as 
September. The values for November and December speak 
volumes. It is curious that, though the bright sunshine of 
January is little in excess of that of December, there should 
he such a marked difference here. Finally, the year, as to 
clearness, gives three main divisions. The summer months. 
May to August, 173; the equinoctial months, March and 
April, September and October, 184 ; and the winter months, 
November to Februaiy, 36. 
Conclusion.— Having thus considered, for York, the chief 
elements upon which its weather depends, the question arises, 
does 50 years suffice to give true averages ? We may answer, 
fairly well, but by no means absolutely, especially as to rain 
We have tested this by comparing the records for each decade. 
The mean of the five will in all probability be five times closer 
to the real value. Thus we have concluded that the mean 
pressure is within on -hundredth inch of 29’905 ; the mean of 
the max. and min. thermometers within D of 47 f, the actual 
mean temperature being probably I’lower; Eainfall, 25 inches, 
