178 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1951 



ARTIFICIAL NUCLEI 



Taken by itself the observational evidence, reviewed very briefly 

 above, does not constitute an adequate basis for a sound appraisal of 

 the potentialities of the cloud-seeding techniques. It is the purpose 

 of this paper to amplify and supplement the observational evidence 

 with the aid of existing knowledge of cloud physics and synoptic 

 meteorology. It is considered that both the laboratory and flight ex- 

 periments prove beyond any reasonable doubt that dry ice and silver 

 iodide will both convert a supercooled water cloud to an ice-crystal 

 cloud. The only necessary condition is that the cloud temperature be 

 a few degrees below the freezing point. For silver iodide, Vonnegut 

 (5) states that the maximum temperature is near —4° C. ; dry ice has 

 been shown to cause ice crystals to appear at —0.7° C. (6), but the 

 temperature should be low enough to allow an appreciable difference 

 between the vapor pressures over water and ice. There is no reason to 

 doubt that ice crystals will grow in a supercooled cloud. The size 

 attained by the crystals is dependent on their number, the initial liquid 

 water content of the cloud, the temperature and the depth of the super- 

 cooled portion of the cloud. It is an essential feature of the Bergeron- 

 Findeisen precipitation theory that the number of ice crystals be small 

 compared to the number of supercooled drops ; otherwise the terminal 

 size of the ice crystals would be of the same order as that of the cloud 

 drops and no precipitation would ensue. There is no satisfactory 

 information on the number of ice crystals produced by dry ice. The 

 estimate of Langmuir (7) that one pellet of dry ice forms 10^® ice 

 nuclei would imply concentrations of the order of 10^ nuclei/cm' even 

 if the nuclei produced by a few pounds of dry ice were distributed 

 through several cubic miles of cloud. Quantitative estimates of the 

 number of nuclei produced by silver-iodide generators have been made, 

 but it is nearly impossible to estimate their concentration in the cloud 

 after release. Atmospheric seeding experiments have resulted in the 

 formation of precipitation elements with a wide range of seeding 

 rates. The inference is either that a much smaller number of nuclei 

 is formed than has been estimated or that there is a marked selective 

 action such that only a small fraction of the nuclei become ice crystals. 

 If the latter is true, the situation would be analogous to natural con- 

 densation in which only a small fraction of the total number of con- 

 densation nuclei become cloud drops. It is not intended to imply that 

 overseeding is impossible, but it appears that this has seldom been an 

 important factor in the seeding experiments which have been reported. 

 When dry-ice pellets are dropped the initial seeding is in a vertical 

 plane from which the ice crystals appear to diffuse laterally. It 

 would be expected that the concentration of "seeds" would be a maxi- 

 mum at the center and would decrease rapidly on either side. If 



