A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 



141 



nucleation is not as high as before until the loth, when a sudden enormous in- 

 crease occurs, cotemporaneously with fall of temperature. The high maxi- 

 mum is succeeded by an equally low minimum, brought on by the rainy weather 

 of October 11 and 12. The nucleation then rises during the succeeding fair 

 days, to fall to a fourth minimum during the rain of October 14. No doubt 



wm. 



2im, 



Charts 1-48. — Daily Observations of the Atmospheric Nucleation of Providence. Number 

 OP Nuclei (») per Cubic Centimeter, Laid off Vertically. In Charts 1-8, n is given in Full 

 In Charts 9-48, n is given in Thousands per Cubic Centimeter. Charts 1-12, are Reduced from 

 an Older Arbitrary Scale. Local Winds, Rain (R), Temperature, etc., are Given in the Charts, 

 Arrows Showing the Direction in which the Winds Blow. Clear , Partly Cloudy a , and Cloudy • 

 Weather, etc., are Indicated with the Usual Signs. 



some variation is due to the variation of the temperature of the apparatus, for 

 which no coirectioni .^^^s made, biit the rain and fair weather effects as a 

 whole are unmistakable. The maxima on the 4th and loth correspond to anti- 

 cy clonal conditions. 



The apparatus was now modified in a way which need not here be instanced. 

 What is noteworthy in the next data (chart 2) is the occurrence of sharp minima 

 on October 16, 17, 21, 23, cotemporaneous with the passage of dense cloud 

 masses over the sky. In three of these cases the curve rises as soon as the 

 sky clears; on October 17 it does not do so, but the curve runs into the over- 

 cast condition of October 18. The pronounced minimum on October 19 during 

 clear weather shows that sunshine can not be a reason for an abundance of 

 nuclei. Similarly there is high nucleation on October 24 and 27, simultaneously 

 with an overcast sky. A number of night observations on October 18 give no 

 evidence of unexpected behavior. 



On October 27 the apparatus was again modified, by substituting a long 

 cylinder for the aspirator flask thus far used as a condensation chamber. The 

 data (chart 3) begin with high nucleation for an overcast sk>' and fall off to 

 the rain storm on October 28. The high nucleations are very fluctuating, 



1 Correction for temperature was applied at a later date, but thereafter again abandoned 

 as untrustworthy. 



