90 EXPERIMENTS WITH THE DISPLACEMENT INTERFEROMETER. 



in terms of the diameter d of the cloud particles, computed from the successive 

 exhaustions. Here one may average the diameter corresponding to vivid green 

 color in the order 



Disk, io*d=i.2 2.5 4.0 5.5 7.0? 8. 5? cm. 

 Ring, io 4 J= 1.8 3.2 4.7 6.5 7.7? cm. 



where above io 4 d = 6.$ cm., the .removal by exhaustion is initially too rapid 

 to catch the successive cases of maxima. If we also remember that in the first 

 exhaustion the coronas are filmy, it is thus a plausible result of observation 

 that the vivid green disk reappears whenever the diameter of particles in- 

 creases by a quantity less than i.5Xio~ 4 cm. and greater than i.2Xio~ 4 cm. 



lac 





A // 



_ 



QP 



4 5 

 FIG. 50. 



8 



If the prolonged occurrence of vivid color throughout several exhaustions at 

 first, and the unavoidable accumulation of errors in the successive exhaustions 

 toward the end of the series, be taken into consideration, it is not unreason- 

 able to adjust the maxima of disk color for diameters of fog-particles in- 

 creasing in multiples of i.3Xio- 4 cm. as follows: 



io 4 d=i.3 2.6 3.9 5.2 6.5 

 No. i 2345 



7.8 cm. 

 6 



