562 
PEOFESSOE EOSCOE ON THE CHEMICAL INTENSITY OF 
Although the curves of mean daily chemical intensity showing the variation from 
hour to hour are symmetrical, the chemical action for hours equidistant from noon being 
the same, this relation appears by no means to hold good for the curves of yearly chemical 
intensity. This is distinctly seen if we compare the monthly means for the two months 
about the vernal with the two about the autumnal equinox, for 1865, 1866, and 1867. 
1865-67. 
March 1867 
April 1865 . . 
September 1865 
Mean Chem. Int. 
. 30-5 
. 97*8 
. 107-8 
. 88-9 
1866. 
March 1866 
April 1866 . . 
September 1866 
Mean Chem. Int. 
. 34-5 
. 52-4 
. 70-1 
. 94-5 
Or for 100 chemically active rays falling in the months of March and April 1865, 1866, 
and 1867 at Kew, there fell in the months of September and August 1865-66 167 rays, 
The curve, fig. 8, Plate XXII., exhibits the biennial variation of chemical intensity at 
Kew for the two years ending April 1, 1867. The yearly integral for the twelve months 
January-March 1867 and April-December 1865 is 55*7 ; whilst that for the twelve 
months of 1866 is 54*7. 
The. marked differences between the chemical intensities in spring and autumn must 
be caused by corresponding differences either in the amount of cloud or in the trans- 
parency of the atmosphere. From Table IV. (p. 561) it is seen that the mean amount 
of cloud in March 1867 and April 1865 is 5-9, and that in August and September 1865 
=4*7 ; whilst the mean cloud for March and April 1866 is 5-9, and that for the cor- 
responding autumn months=6‘8. If the number of observations made when the sun 
is shining be compared with those made when the sun’s surface is obscured by clouds, 
it is seen that of sixty-nine observations made in April 1865 the proportion between 
cloud and sunshine was as 1 to 1-9, whereas in the months of August and September, 
out of 130 observations the proportion between cloud and sunshine was 1 to 2*1. In 
1866 out of 123 spring observations the relation is 1 cloud to 0'55 sunshine, and out 
of 122 autumn observations the relation was found to be 1 cloud to 0-60 sunshine. 
Hence it appears that the effect of varying amount of cloud has been eliminated by 
the number of the observations, and that the difference in chemical intensity cannot be 
ascribed to the presence of more cloud in the spring than in the autumn. 
The only other possible explanation is to be sought in the difference in atmospheric 
transparency in spring and autumn, and the only indication which we at present possess 
of such variation in transparency is afforded by measurements of the hygrometric con- 
dition of the air, the increased transparency of moist air for the visible rays being well 
known. In March 1867 and April 1865 the mean amount of moisture was found to be 
2‘82 grains per cubic foot ; in August and September 1865 it was 4*65 grains. In March 
and April 1866 the moisture was 2-8 grains, and in August and September 4‘21 grains. 
This gives a relation of 1 to 1*65 for spring and autumn moisture 1865, and 1 to T50 
