HOW THE SUN WARMS THE EARTH ABBOT 165 



6,000° absolute centigrade shown near it. Evidently they are much 

 alike in most parts, but the solar curve falls away from the other 

 in the ultraviolet. This weakness of ultraviolet rays is due mainh^ 

 to the powerful absorption of the sun's ultraviolet rays by the sun's 

 own gaseous envelop. But you will gather from this comparison of 

 the two curves, I am sure, an indication of one of the ways used to 

 estimate the sun's effective surface temperature as of the order of 

 6,000° absolute centigrade. 



THE SILVER-DISK PYRHELIOMETER FOR MEASURING TOTAL SOLAR 



RADIATION 



Plate 5 shows the silver-disk pyrheliometer. It is an instrument 

 which I invented at the Smithsonian Institution for measuring the 

 total heating power of the whole solar beam. It has been standard- 

 ized against the water-flow pyrheliometer, also my invention, and 

 over 70 copies of the silver-disk pyrheliometer have been made and 

 calibrated at the Smithsonian Institution and furnished at cost to 

 observers on all the continents of the world. At the front of the 

 pyrheliometer is a tubular vestibule with numerous diaphragms for 

 keeping out sky light and the wind, while admitting the sun's rays 

 to the box beneath. In this box is a blackened disk of silver, as 

 big as a dollar, and about five-sixteenths inch thick, on which the 

 sun's rays shine at right angles. In a radial hole, bored from one 

 side into the silver disk, is sunk the cylindrical bulb of a delicate 

 mercury thermometer. For convenience the stem of the thermometer 

 is bent at right angles and fastened alongside of the vestibule. The 

 observer, with watch in hand, notes the rate of rise of the ther- 

 mometer when he has admitted sun rays by opening the shutter 

 at the top of the vestibule tube. In this way the heat of the sun's 

 rays is measured. 



THE SOLAR CONSTANT OF RADIATION 



We cannot carry the silver-disk pyrheliometer to the moon to 

 measure the intensity of sun rays outside our atmosphere. It is 

 true that we did in 1914 send an automatic pyrheliometer into the 

 stratosphere (figs. 8 and 9) and got good observations with twenty- 

 four twenty-fifths of the atmosphere eliminated. Instead of actually 

 observing outside the atmosphere, w^e measure the area under one of 

 the bolometric spectral energy curves that I showed a few moments 

 ago, and also the area under the computed solar-energy curve outside 

 the atmosphere, which I also referred to. Dividing the latter area 

 by the former, we obtain the factor by which, if we multiply the 

 solar heat observed by the pyrheliometer, we may obtain the heat 

 which that instrument would have measured if we had taken station 

 on the moon and observed the sun there. 



