A Possible Variation of the Solar Radiation. 79 
one considered here, is restricted to a single station, which 
we must usually take in a low altitude like that at Washington. 
It consists in making measurements.:—(1) By the actino- 
meter of the total radiation observed ; (2) Of the intensity of 
homogeneous rays in different parts of the solar spectrum 
for different altitudes of the sun. (3) From these direct 
observations we can in theory calculate the total absorption 
of the atmosphere, from the evidence gathered by the spectro- 
bolometer us to that of the several rays for different altitudes 
of the sun, and for different thicknesses of air. 
As our atmosphere and its absorption is in a condition 
of incessant fluctuation, it will always be hard to dis- 
criminate between changes due to it, and to those, if any, 
due to an actual change in the radiation of the sun itself, 
and this discrimination constitutes the particular difficulty 
of the present research. 
The present discussion w ill lead us to estimate the 
validity of any surmise as to the actuai changes in the 
solar constant, founded on such observations as those of the 
second class. 
As for the accuracy of the actinometer measures of total 
solar radiation at the earth’s surface, it is easy to obtain 
measures relatively exact, within about two per cent., or, 
perhaps, even less. But the absolute standardization of the 
instrument offers uncertainty of quite another order, so that 
I do not think that the absolute magnitudes of the results 
to be presented here are necessarily within 20 per cent. of 
the truth, or even more. 
The homogeneous rays are observed here by the bolometer, 
and the bolographic curves from which the atmospheric 
extinction of radiation is inferred, traced by the movement 
of the spot of light upon the sa leemnteie -scale, are now 
very much more satisfactor y than formerly. They represent 
an immense gain over the conditions operating when | began 
the work at Allegheny. The light spot should move only by 
an impulse from the sun; but owing to extraneous causes, it 
was at first frequently impossible to “keep it upon the scale of 
the galvanometer during so short a time as a single minute. 
The apparatus now, however, operates so well that such drift 
and tremor is relatively unknown, and the zero of the galva- 
nometer is found almost unchanged for weeks t together, a 
gain due to many causes and successive improvements during 
many years, from my own and other hands, but to which none 
have contributed more than Mr. Abbot of this Observatory. 
Measurements and reduction of a series of from five to ten 
bolographs of a single day requires, however, so much labour, 
