148 



THK STRUCTURE OF THE NUCLEUS. 



air l>elo\v ami the pure air aliove, was liable to condense as a whole, almost explo- 

 sively. Tiiis occurred at a definite pressure and after condensation had already 

 l)e"-un in the nucleated region. Suspecting that the discrepancy might be due to 

 the hY""ronietric state of tiie atniosiiliere, I made the following tests, which bear out 

 this surmise. The first column shows the pressure decrement on exhaustion, the 

 second the effect produced on the nucleated atmospheric air in the dry receiver. 

 The third column of the table are the corresponding ajjproximate dew points, the 

 fourth and fifth other hygrometric data. In the second and third parts of the 

 table tlie results of artificially moistening and of drying the air are at once apparent. 



lAlil.E 2— CONDENS.VriON IN ADIABATICALLY COOLED IONIZED (MOIST) 



ATMOSPHERIC AIR. 



I. Room air. 



II. Room air over damp sponge. 



III Room air dried over CaCU. 



Clear 



3^5 



■ ^6 



It seems to me probable that a method of hygrometry is here suggested which 

 is worth a trial,' and for which suitable apparatus could be devLsed. In other 

 words, artificially nucleated air is suddenly cooled by expansion until a fog just 

 appears. The dew point is computed from the pressure decrement thus deter- 

 mined. If / be the tem])erature of the air in degrees centigrade and p its pres- 

 sure, and if the air is cooled from 20° and 70 cm., we may wi-ite approximately, 



' Prof. Cleveland .Abbe informs me that Espy's nepelometer of 1835 and 1850 is virtually (he 

 same as this. See Abbe's Treatise on Meteorological Apparatus, 1S87, pp. 357-59. "Espy showed 

 that by repeated condensation he precipitated all the nuclei and got an apparently drier air ; /. e., it 

 required greater exhaustion to produce the fog." 



