NO. 8 WATER-VAPOR TRANSPARENCY FOWLE 3 



to observe in the region between 2 jx and 20 fx is wholly invisible and 

 very feeble even in the emission from the hottest terrestrial sources 

 so that extremely sensitive measuring apparatus was necessary. 

 Few substances are transparent to it. Glass as a rule is non- 

 transparent for wave-lengths g"reater than 2 fi, quartz for those 

 greater than 3 ix, and fluorite for those greater than 10 /x. A plate of 

 rock salt i cm. thick absorbs 50 per cent at 17 fx. and one of silvite 

 50 per cent at about 21 /x. As silvite was not available it was neces- 

 sary to use rock salt of which the Observatory, through the courtesy 

 of the Russian Government, possesses a number of large and beau- 

 tiful prisms. 



Preparations had to be made for observing a great range of 

 intensities. The radiation from a Nernst lamp, such as was used in 

 this research as a source of energy at a temperature of about 

 2,200° K., is at 20 fi only 1/100,000 as intense as it is at 2 /x. The 

 difficulties inherent in finding proper means for observing accurately 

 over such a range of intensities will be discussed when the means are 

 described. 



No mirror perfectly reflects an incident beam in one direction. An 

 appreciable portion of the beam is scattered in all directions as may 

 be easily noted by looking from any direction at a silvered mirror 

 upon which a beam of sunlight falls. This scattered light becomes 

 more and more troublesome as that part of the spectrum is ap- 

 proached where the intensity is only 1/10,000 that in the brightest 

 part. Field Hght, as it is usually called, finally amounts to the 

 whole of the observed energy. Means had to be provided either for 

 eliminating or measuring it. 



In order that the air, the transparency of which was to be meas- 

 ured at normal atmospheric conditions as to temperature and pres- 

 sure, should contain sufficient water vapor, it was necessary to make 

 most of the observations during the hotter summer months. Air 

 of 50 per cent relative humidity at a summer temperature of 40° C. 

 contains 12 times as much water vapor as of the same relative 

 humidity and a winter temperature of 0° C. Even then, with the 

 length of path possible in these laboratory experiments, the air 

 column contained only as much vapor as may be found in the atmos- 

 phere during the winter months along the zenith path of the sun's 

 radiation to the earth. On a summer day there may be 10 times as 

 much vapor as this in the sun's path. 



In 1908 preliminary experiments were made at intervals between 

 the regular work of the Observatory. During 1909 and 191 1, in the 



