IN PASSING THROUGH THE ATMOSPHERE. 
255 
79. First. The rapid fall of the actinometric curves is due to the extreme rapidity with 
which the length of path increases as the sun approaches the horizon. They differ from 
the curve of air temperature, because that is a slow and gradual result of complicated 
actions ; the coolness of evening is a continuous result ; but the disappearance of the 
sun under the horizon corresponds to an instant extinction of the force of radiation. 
Secondly. With respect to the points of contrary flexure which occur about two hours 
before and after their maximum values, and which in both curves are slightly marked 
in the morning, and more intensely in the afternoon, they probably arise from the 
combination of a two-fold effect of the sun’s elevation. The one is the increased 
intensity as the sun is higher, the other is the transference of vapour from the lower 
to the higher regions of the air by the heating of the lower strata, producing the inci- 
pient condensation at a certain elevation, already alluded to as the cause of the slight 
clouds which often appear between ten and twelve o’clock. As the sun’s power dimi- 
nishes, and the vapours redescend into the less rarefied and warmer regions, they are 
in some degree redissolved in the afternoon, and the increased transparency of the at- 
mosphere (which will besides be aided by the general maximum of the temperature of 
the air occurring in the plains between two and three o’clock, and producing also a 
maximum of dryness there) checks the downward progress of the curve due to the 
increasing obliquity of the rays. Thirdly. The curve at the upper station lies wholly 
above that at the lower station, on account of the absorption of heat in every case by 
the intercepted air. Fourthly. The range at the higher station is greater than that 
at the lower. This is an evident and necessary consequence of the fact, that the max- 
imum above must exceed the maximum below, and that at sunset and sunrise they 
must both pass through zero. It might be more correct, however, to consider the 
continuous part of the curve extending to the moment before sunset from the moment 
after sunrise. In this case we might expect the difference of intensity at the two 
stations to increase very rapidly with the obliquity of the sun’s rays, so that the two 
curves, instead of approaching one another in the morning and evening as they ap- 
pear to do (Curves XII. and XIII. ), ought to separate further. It is to be recollected, 
however, that the extinction in any stratum varies with the intensity of the incident 
heat, and that being very small near the horizon, the absolute extinction will be very 
small also. Nevertheless, the relative extinction may be very great, as appears from 
the form of the right-hand branch of Curve XIV. The morning branch does not show 
the same effect, and this may be thus explained. The evening vapours are dense and 
absorptive ; in the morning the atmosphere is comparatively clear, especially amongst 
mountains. To this circumstance must be imputed the much more rapid fall of the 
Curves XII. and XIII. in their evening than their morning branch. But further, it 
will be shown presently that the law of uniformly regular extinction is not true, and 
that the loss in passing through a medium, is not only absolutely but relatively (to 
the intensity) greater at first than afterwards: that when the thicknesses are very 
great, any additional thickness intercepts but little of the radiant force ; conse- 
