OF ARTS AND SCIENCES. 



135 



One observer being stationed at the galvanometer, and a second at 

 the instrument out of doors, at a signal from the first, the latter brings 

 the image of the moon into the opening of the thermograph, and the 

 deflection occasioned is noted and recorded by the observer at the gal- 

 vanometer. In this manner a set of ten readings is taken, and, the 

 altitude of the moon being measured at the beginning and the end of 

 the set, the mean of the ten deflections is taken as the deflection cor- 

 responding to the mean altitude. About six minutes are required to 

 make a set of the readings. 



The following series, selected at random from the observation-book, 

 will serve as an example of the results obtained. 



Date. — Febroary 24, 1888. 



The probable error of a single observation, when the moon is the 

 source of heat, is considerably greater than is the case with an arti- 

 ficial source, or even with the sun. The cause is, in part, due to in- 

 visible clouds of vapor floating in the atmosphere, as Langley thinks. 

 It has been observed that immediately after rain this effect is particu- 

 larly noticeable, even when the sky appears perfectly clear to the eye. 

 There is another cause,' which doubtless exerts a large influence; 

 tliat is, the irregular radiation of different portions of the moon's 

 surface. It has been found* that, photographically, the mean bright- 

 ness of the dark region is to that of the bright region in the ratio of 

 55 to 100. We might be prepared to expect large differences in the 

 radiant energy of the respective regions, and the average radiation 

 may therefore be more closely attained from a series in which the in- 

 dividual readings vary considerably, than from a more perfectly con- 

 cordant set. 



* Annals of Harvard College Observatory, Vol. XVIII. No. IV. 



