86 
MC ADIE. 
have systematic measurements of the cloud height. This 
is now done at Blue Hill, Upsala, Storlein, and Berlin by 
means of theodolites and double observing stations. A 
more direct way and one which we think is entirely prac¬ 
ticable is to send apparatus up into cloudland by means of 
kites. This would give us the conditions prevailing at dif¬ 
ferent cloud levels, and the records would not be momentary. 
Two other ways of investigating water vapor conditions 
at some distance above the ground may be alluded to: First, 
by means of the spectroscope. No less than 928 absorption 
lines due to aqueous vapor have been mapped. Dr. L. 
Becker has given three groups; one of 678 lines, with wave¬ 
lengths 6020-5666 ten-millionths of a millimeter; the second 
containing 106 lines of wave-lengths, 5530 to 5386, and the 
third having 116 lines, with wave-lengths 5111 to 4981. 
The infra-red portion of the spectrum is probably particu¬ 
larly rich in water vapor lines, and with the determination 
of an intensity scale for these terrestrial lines, the distribu¬ 
tion of vapor in the air may become known. 
Second, by means of the electrometer. In noticing the 
formation of cumulo-nimbus clouds we alluded to the part 
played by electricity. We have measured with a sensitive 
quadrant electrometer the pull in volts experienced by the 
air between one of these clouds and the earth. We could 
tell of the approach of the cloud while yet far off, and by 
the changes in the potential could roughly map out the sky. 
This may therefore be a second way in which to determine 
the w r ater vapor at a distance. The electrometer has the 
advantage of the spectroscope as a cloud detector in two 
ways : it can be used at night, and whereas the spectroscope 
ceases to be available when the cloud is at all dense, the 
electrometer does not. 
