182 proceedings: philosophical society 



Accuracy and efficiency recommended the use of a interferometer of 

 the Fabry and Perot type for this work, since this apparatus can be 

 conveniently enclosed in a chamber in which the temperature and pres- 

 sure of the air can be regulated as desired, and it also permits simul- 

 taneous observations for a lai'ge number of different wave lengths. 

 Sections of the circular fiinges, produced bj^ various radiations from a 

 source of light illuminating the parallel plates of the interferometer, 

 were photographed either with a grating or a prism spectrograph, first 

 when the space between the plates was evacuated and then when dry 

 air at measured temperature and pressure was present. 



The index of refraction of air for a particular wave length was ob- 

 tained dii'ectly from measurements of the photographed interference 

 fringes, which allowed the ratio of lengths of this wave in vacuum and 

 in ail' to be calculated. Observations were made at spectrum intervals 

 of about 40 A from the extreme ultra-violet at 2200 A, through the 

 visible spectrum and into the infra-red to 9000 A . 



Complete sots of observations were made on dry air at atmospheric 

 pressure and at temperatures of 0°C., 15°C., and 30°C. These ai-e 

 quite well lepresented by the following dispersion formulae: 



13.412 0.3777 



(.-1)0X10^ = 2875.66 + ^^,^^, + ^-^^^^, ' 



12.288 0.3555 



(„ - 1). X 10' = 2720.43 + ^^^^, + ^;r^j^, 



12.259 0.2576 



(« - I),. X 10' = 2589.72 + ^-^^^ + ^-^^^^, 



The coefficient of index variation with temperature was found from 

 these ol)servations to be a function of the wave length. For long- 

 waves this optical temperature coefficient is identical with the density 



temperatuie coefficient, i.e., ,y^, but as the ultra-violet absorptio?i 



band is approached it increases rapidly, l)ecoming ^^y^ at 2500 A. 



Discussion: This paper was discussed by Messrs. Burgess, SwAnn, 

 CRi'rr]«]NDEN, and Meggers. 



The third paper, on Barometric ripples, was presented l)y W. J. Hum- 

 phreys. This paper was illustrated by lantern slides. Small pressure 

 changes, amplitude usually 0.1 to 0.3 mm. and period of 5 to 10 minutes, 

 and continuing for hours or even days together, are very common 

 during cold weather. 



As first demonsti'ated by Helmholtz, whenever layers of air that diffei' 

 in density at their interface flow over each othei-, long billows, anal- 

 ogous to giavity water waves, aie produced, which conform, af)proxi- 

 mat(;ly, to the equation 



