80 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1919. 
the personal investigation of the writer,, it appeared necessary to 
make an expedition to South America to attend to these several 
matters. The preparation for the eclipse work occupied some time 
of the director and of the instrument maker. 
Several investigations relating to the war, a brief note of which 
was mentioned in last year's report, were continued during the fiscal 
year. 
The painstaking and valuable work which Mr. Fowle has been 
doing on the revision of the Smithsonian Physical Tables should re- 
ceive some notice, although this work is being done by him outside 
of his regular hours of service for the observatory. This book has 
passed through a number of editions under his editorship and has 
attained an enviable reputation in this country and abroad for the 
accuracy and fullness of its contents. The new edition, which is now 
in press, has received unusual attention on his part, and very valu- 
able cooperation from the various scientific departments of the 
Government and of outside individuals in colleges and industrial 
corporations and elsewhere, and will be a great advance over any 
of the former editions. 
In connection with work of the Observatory, we have long wished 
to determine the solar constant of radiation by a method which 
does not involve the assumption that the transparency of the atmos- 
phere is constant over the two or three hours required for the deter- 
mination of it by the usual spectrobolometric method. We hoped 
that, seeing that the sky is brighter when the transparency is less, 
an observation by the pyranometer, or some other more suitable 
instrument, of the brightness of the sky in the neighborhood of the 
sun, combined with the usual measurements of the pyrheliometer 
and perhaps of the spectrobolometer, but only at one period of time, 
might be sufficient to determine the solar constant by a satisfactory 
empirical process based upon spectrobolometric investigations of 
former years. In the hope of getting an instrument more satisfac- 
tory than the pyranometer for this special work, a new design com- 
prising essentially two disks, one of which is shined upon through 
a graduated aperture by the sun and the other of which is exposed 
to the small region of sky desired and both connected by thermo- 
electric junction so as to enable equality of temperature of the two 
disks to be adjusted, was devised and partly constructed at Wash- 
ington. It was sent in a letter to Calama, Chile, and was finished 
by the director during his visit in Chile and is now in satisfactory 
operation, although it has not yet supplanted the pyranometer for 
the purpose in question. 
Another problem which requires a new kind of apparatus is the 
measurement of nocturnal radiation such as the earth sends out to 
space. This investigation is exceptionally difficult, for it involves a 
