256 
PROFESSOR FORBES ON THE EXTINCTION OF THE SOLAR RAYS 
quently, near the horizon, a great thickness of atmosphere having been traversed by 
the rays which reach the upper station, even the obliquity of the passage to the 
lower station does not (unless the inferior strata be particularly loaded with vapours), 
cut off anything like a corresponding portion of solar heat, and a second equal mass 
would intercept still less. 
80. Fifthly, and lastly, the maximum of intensity is sooner attained above than 
below. This arises, no doubt, mainly from the fact (amply confirmed by the hygro- 
metric curves), that the sun shines with a disproportionate intensity during the 
morning on the upper station, owingto the mass of vapours being then in the valleys. 
The solar intensity will therefore attain an earlier maximum, since after ten or eleven 
o’clock a quantity of vapour rises between the upper station and the limit of the at- 
mosphere, and therefore throws the maximum rather before noon. In the plains, on 
the contrary, where the whole atmosphere is all day between the observer and the 
sun, the maximum will incline towards the period of maximum dryness of the day, 
that is, it may be an hour or half an hour after noon. The curve of mean dampness 
VII., with its point of contrary flexure in the afternoon, entirely confirms this view, 
and the diurnal curve of temperature at Brientz, marked I., shows both inflections in 
the clearest manner. 
81. From the comparison of the two curves of solar intensity, we have deduced 
the mean loss of heat intercepted between the two stations, and we have thence con- 
cluded that, on the hypothesis of uniform opacity, about one-third of the solar heat 
is lost by vertical transmission through ihe atmosphere. It is interesting to compare 
this result with that which is deducible from the individual observations at either 
station. In that mode of viewing the subject, it appears from Bouguer’s reason- 
ing (see Art. 7), that two observations at different altitudes are, in point of rigour, 
sufficient for deducing the loss due to vertical transmission. If more than two 
values have been got, they may be combined in two series of which the means 
are taken ; or they may be treated by the method of least squares, which will give 
the most probable result, on the hypothesis of the diminishing geometrical progres- 
sion of the intensities. It is more interesting and important, however, to employ the 
superfluity of observations in testing the accuracy of the assumed law, rather than in 
giving a merely illusory degree of precision to the results of a law which may be 
wrong. For this purpose I projected, by rectangular coordinates, the intensities 
observed, the thicknesses of homogeneous air traversed (computed from the sun’s 
altitude) being the horizontal coordinate or independent variable. On the law 
commonly assumed, the points thus determined ought to lie in a regular logarithmic 
curve, which being readily prolonged by geometry or by calculation, would give the 
intensity corresponding to thickness 0, or the degrees which the actinometer B. 2. 
would show if placed wholly beyond the atmosphere. 
82. When the projection came to be made, I remarked, with much interest and 
some surprise, the admirable agreement between the insulated observations at both 
