NO. 4 SOLAR RADIATION ABBOT, FOWLE, AND ALDRICH 35 



may enable you to suggest sources of error which may be removed before the 

 flights take place, or at least satisfactorily determined in advance by experi- 

 ments. 



This instrument is a modified form of our disk pyrheliometer. A blackened 

 aluminum disk, a (fig. 5), encloses a thermometer, d, whose stem is shown in 

 enlarged cross section at d 1 . The cavity for the bulb of the thermometer 

 within the disk, a, is filled up with mercury, and sealed at the mouth with 

 thread and wax as in our pyrheliometers. The disk is enclosed in an interiorly 

 blackened aluminum box, b. Two polished copper rings, k k 1 , limit the solar 

 beam to a cross section less than that of a. As the temperature of the disk a 

 changes, the mercury in the stem fluctuates, thus allowing the sun to print on 

 more or less of the length of the photographic drum, e, according to the tem- 

 perature. Thus when the paper (solio paper) is removed, there is a record 

 like this (see fig. 9) : 



A clock work / rotates the drum, and at the same time causes the shutter, 

 g h i t 1 , to be for four minutes in the position above the disk a as shown, then 

 four minutes opened (as partially shown dotted at the left), then again closed 

 as shown, and so on, rotating, at the end of each four minutes, 180° on g as 

 an axis. The shutter comprises three parts. Of these i and t 1 are polished 

 aluminum disks, and h a polished silver cone. The angle of the cone, h, is such 

 that all rays from a must go either directly or by reflection to the sky, none to 

 the earth. Hence when the shutter is closed the disk a observes the sky 

 directly, or by reflection, though not the zenith sky. When the shutter is open 

 the disk observes the sun plus the sky, at this time the zenith sky. Hence the 

 difference between the radiation exchange when the shutter is open, or closed, 

 is not entirely due to the sun, but in part to the difference between zenith and 

 horizon sky, and to the imperfect reflection of silver. These differences are, 

 however, not large, and they may be approximately determined. At high 

 levels the skylight will diminish, and the difference of radiation exchange to 

 surroundings (other than the sun) between shutter open and shutter closed 

 may become very small indeed, compared to solar radiation. The shutter is 

 made, when closed, to hide the sky to 30 zenith distance from all parts of the 

 disk a, when the apparatus hangs as if suspended from the balloons. The 

 apparatus is hung by a steel wire of nearly 25 meters length below the balloons. 



In order to prevent the mercury in the thermometer from freezing, the cup 

 b is wound outside and underneath with resistance wire, and batteries are 

 taken along to heat the wire. Their action is automatically controlled by a 

 curved strip of brass and invar c lying in a groove in the cup b and arranged 

 to open against platinum points and complete circuit when the temperature of 

 the curved strip goes below o° C. [This arrangement was not used in the 

 most successful flight, and is not shown in fig. 5.] The whole apparatus is 

 covered with a blanket of black silk and down, excepting the top of the disk a, 

 the shutter h, and the thermometer stem d. 



Each instrument is to be repeatedly calibrated against silver disk pyr- 

 heliometers before sending it up, and the flights are to be made on cloudless 

 days, and pyrheliometer readings taken on the ground during flight. A 

 correction to the aperture for zenith distance of the sun will be made. 



As stated above, similar experiments have already been made with consider- 

 able success in 1913. Records to 13,000 meters were obtained, but for lack of 

 the heating apparatus above mentioned the mercury froze, and prevented 



