Till.; ariiiu Ti'ijK oi- iiik NtrcLEUs. i;j3 



critical density is not readied, it will eseape |)reci])itnti.)n on exliaiistion. Tlieiv is 

 no direct evidence of snch particles in the abovt^ leseai'cli. 



The loaded nucleus dnrini^r preeipitation will also ten.l t.. re-evaporal.- i.. iis 

 original nucleai' size and densi(y, and thus escape removal. F<.r tlie stahle dianiet.-r 

 is conditioned solely l)y the amount oC s.-hite //////r^% entrappcl. Ii is fnmi this 

 point of view that the paitial piecipitations at the lower i)re.s8ure decrements are 

 to he explained. The aliove case of calcic chloi-ide I'oi' pressure decrements of 2 

 rm. may he cited. If it he supposed that all partich-s in evidence im- preripi- 

 fated, the siicressioii is as follows, for a, l-ota,l of .".(id particles: 



Kxiiaustion, 12345 



Precipitation, 149 88 56 7 none 



Kscape, 151 63 7 none 



Tiie xci'v slow suhsidciice of the small pai'ticles in these diffuse coronas o-ives 

 excrption.-d opportunity for evaporation; hut all strong coi'oiias, in spite of the 

 moi'e i-apid pi<'ci[iitation which a<-com|>auics them, rc(piire more than one exhaus- 

 tion to free the air of nuclei. 



In spite of this repeated partial i)recipitation, thei'e is never any irregularity 

 in the coronas, showing the chai'.-icteristic rapid diffusion of the aqueous nuclen.s. 

 It is common experience that the dilfiise coronas for small pi'essni'e decrements 

 form gradually. Pi'ohahly, as a result of tiie evaporation in (juestioii, the shrink- 

 ing coronal figure becomes more luminous. The strong, sharp coronas at hi^di 

 pi'essure decrements ai'e foi'ined at once. For some occult reason, early coronas 

 are liable to be lai'ger than later coronas, <-ait. par., a result ap[)earing in many of 

 th(^ tiiplets above. 



45. J-iJqitations for diliife volutions.— The endeavor may ite made to give these 

 results a more definite foiin. Not much will be gained by this, for the chief 

 desideratum is an ex[)ei'iinenlal investigation of the vapor pressure of the more 

 nearly concentrated solutions. Tin; following group of ecpiations will, therefore, 

 fall short of meeting the actual conditions, though the estimates in jj 48 leave much 

 of the question open. 



If p,. and ^;co ^^^ t'le vapoi' [)ressuies conx'sponding to radii /■ and 00, 7' the sur- 

 face tension, p and cf the density of the liquid solution and its vapor, Kelvin's 

 equation asserts, 



The [)resent licpiid is a solution and its vapor [)ressure, j^^^, is thus below the noiinal 

 value for the pure li(piid. 



For the I'easons stated in the last paragraph, it is conceivable that a state of 

 eqiiilibi'ium will be eventually reached in which di'ops large or small may all have 

 the same vapor pressure, namely that of the free surface of the li(piid, and that 

 evaporation must therefore cease and the droplets persist, however small. In fact, 

 after the lapse of time, ec^uation (1 ) for ;i given droplet will be replaced by 



