Mr. W. S. Jevons on the Forms of Clouds. 251 



niical causes, namely, the momentum which we communicate in 

 the fii'st ph\ce to the injected liquid, and the motions which 

 afterwards arise from differences of specific gravity purposely 

 occasioned. Supposing the analogy of this experimental cloud 

 to the real atmospheric cloud to hold good, it follows that the 

 latter is a subject to be treated on the simplest dynamical 

 principles, while the conditions of our experiment actually 

 indicating the proximate causes of the atmospheric phseno- 

 menon, afford us a new and safe giiide to its ultioiate causes. 

 Leaving others, however, to subject the miniature thunder-cloud, 

 or rather (see 20) the idea which is evolved in this, to a com- 

 parison with the actual phsenomenon, and assuming on my own 

 part that it will be favourable, I proceed to offer a suggestion as 

 to the conversion of the experimental conditions into those of 

 the atmosphere. 



28. Having shown (in 20) how the subsidence or aggregation 

 and rapid fall of wateiy particles within a cloud may cause the 

 remaining air (still partially cloudy) to become permanently 

 warmer and specifically lightei', so as to seek a definite elevation 

 above the plane of precipitation, and there to spread out gra- 

 dually in horizontal stratiform projections, let us suppose this 

 action to have gone on to a certain extent within a cloud so as to 

 produce the cumulostratus. The latter will consist of two parts, 

 the active cumulose cvntre, and the tranquil bodies of cloud 

 which are derived from and surround it. But if we now suppose 

 subsidence and the formation of rain to continue throughout 

 the cloud, it is evident that the specific gravity of the stratose or 

 tranquil parts of the cloud will be still further decreased, but 

 gradually decreased. Having assumed a given elevation, because 

 there alone their specific gravity was equal to that of the sur- 

 rounding air, the equilibrium of these stratose parts can now no 

 longer be maintained. As they gradually decrease in sj^ecific gra- 

 vity, the uppermost layers of air, which are most free from watery 

 particles, must tend to ascend to greater elevations, and here we 

 meet precisely the conditions in which, as shown in my previous 

 paper, an upward cirrose filtration will take place. Small streams 

 and threads of warm and moist air will ascend from the upper 

 part of the cloud, and forcing their way into the superior strata, 

 also moist, but of less temperature, renewed precipitation will 

 undoubtedly take ])lacc according to the mixture theory of Dr. 

 Hutton, and the cliaractcristic crest of cirrus will be produced. 



29. It is not ))ossible within the narrow limits of this paper to 

 answer any probable or apparent objections to this theory, nor to 

 offer the many su])porting arguments and instances which may 

 occur to me. I cannot resist noticing, however, that the me- 

 teorologist Howard, iu speaking of the nimbus or cumulo-cin'O- 



