144 



A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 



coUeagtie, Prof. Winslow Upton) with the data for nucleation given in the 

 above charts. A hniited degree of similarity was in this way made out for 

 cold weather, both in the general march of temperature throughout Linger 

 intervals and in its detailed variations. Thus the mid-day minima are obvi- 

 ously counterparts of corresponding temperature maxima. There were, how- 

 ever, man}^ otttstanding discrepancies, and it seemed probable that much of 

 this would be referable to local differences, as the observatory is differently 

 situated and at a distance from the laboratory. Beginning with December 28. 

 I have, therefore, been making daily temperature observations simultaneously 

 with the coronal observations, using a long mercury thermometer placed be- 



J p'rS^ l&Ikt, Zi 



IS U \Z7 29 M 30 31 4^ e_ 



^ 



iffi^ 



mjkc. n 



side the influx pipe. The temperatures are given on the tops of charts 7 and 

 8, and positive temperatures in degrees Fahrenheit are laid off downwards to 

 instire an easier recognition of coincidences, remembering that nucleation 

 maxima and temperature minima correspond. 



In the fragmentary temperattu-e observations before December 28 (chart 

 7) some relation is seen; thus the maximum between December 23 and 25 has 

 a tempera tiu-e equivalent, the broad minimum on December 22 and the mid- 

 day minimum on December 24 correspond, etc. Turning to the cases under 

 detailed observation, the mid-day minimum on December 28, the fall on Decem- 

 ber 29, and, to a smaller extent, the data on December 30 and 31 (in the latter 

 case other factors are active) agree. The marked minimum on both cases of 

 January i and, to some extent, January 2 are especially striking. 



In chart 8 the tjeneral maxima marked "c.w. " mav be noted, and the 



