CLOUD SEEDING — HOUGHTON 181 



sumption. Present knowledge of the concentration and properties of 

 natural nuclei of crystallization and sublimation is incomplete. Super- 

 cooled clouds are much more frequent than ice-crystal clouds down to 

 about —10° to —15° C. At lower temperature ice-crystal clouds be- 

 come more and more frequent although supercooled clouds have been 

 reported at temperatures at least as low as —35° C. Measurements 

 of the number of ice crystals appearing in an expansion chamber by 

 Findeisen and Schultz (8) have indicated that the first crystals appear 

 in the neighborhood of —7° C. The number of crystals formed was 

 found to increase slowly as the temperature was lowered. At a tem- 

 perature variously reported to be from —32° to —42° C. a very large 

 increase in the number of crystals occurs. Thus, this experimental 

 evidence is in general accordance with the observed temperature distri- 

 bution of supercooled clouds. The nuclei introduced by seeding are 

 active at temperatures of about — 5° C. and below. It is well to point 

 out here that it is not at all certain that ice crystals are a necessary pre- 

 requisite for natural precipitation. Moderate rain often falls from 

 tropical clouds that do not extend to the freezing level. The mech- 

 anism responsible for such precipitation presumably can operate also 

 in a supercooled cloud although the ice-crystal mechanism is more 

 effective if the ice crystals are present. 



The vertical development of convective clouds is often terminated 

 by an inversion. If the temperature at the top of such a cloud were 

 below —5° C. but still not low enough to activate the natural ice nuclei, 

 seeding might release rain that otherwise would not fall. From the 

 prior discussion it is evident that the precipitation could be substantial 

 only if the cloud were also deep and in active development. This is 

 the most interesting case considered so far. It is difficult to estimate 

 the frequency of occurrence of the requisite combination of conditions 

 from the data that are available. Certainly many convective clouds 

 are topped by inversions at a temperature level of between, say, —5° 

 and —20° C. but how many of them remain in active development after 

 they reach the inversion and in how many cases is there a lack of natu- 

 ral freezing nuclei ? Even though it might not be possible to secure 

 data on the freezing nuclei, a well-designed observational program 

 should help to answer these questions. 



It is of interest to consider whether continuous seeding would be 

 required to maintain precipitation in the case discussed above. It 

 has been suggested that when snow crystals are once formed they shed 

 tiny splinters as they fall, which then serve as verj' effective sublima- 

 tion nuclei. If this happens, the process would tend to be self-per- 

 petuating, after the initial seeding, as long as the circulation was such 

 as to carry the splinters into the updrafts. Information on splinter- 

 ing is very incomplete and no definite conclusions can be drawn. It 



