Mr. W. S. Jevons on the Cirrous form of Cloud. 29 
2nd. The precipitate of chloride of silver has a tendency to 
subside, its gravity being much superior to that of water; but 
during the precipitation of aqueous vapour into cloud-matter a 
great amount of latent heat is given out, which, by expanding 
the air, diminishes its density, and gives it a tendency to rise 
above surrounding portions instead of subsiding, on account of 
the weight of the watery particles. Hence it is that cirrous 
fibres will be observed ascending much more frequently than 
descending; and that the ends of fibres have a general tendency 
to curve upwards, as we may observe in nature. 
3rd. Atmospheric strata being usually in horizontal motion in 
different directions, cloud-fibres produced by the interfiltration 
of portions of these cannot have a perpendicular direction, but 
will generally lie in a nearly horizontal position, except perhaps 
towards the extremities. They will also more generally partake 
of the motion of the upper than the under current. 
It is easily conceivable, on this theory, that by different com- 
binations of currents and other circumstances, the most variable 
appearances and shapes will be assumed by the cirrous fibres, 
even such a variety as we observe in their beautiful natural 
forms. 
To render this paper complete and convincing, it would be 
necessary, in addition to showing that such cirrous filtration 
may take place among gases, to prove by facts that the condi- 
tions of our theory, viz. a moist and dense stratum of air lying 
upon another portion of air, however small in quantity, also 
saturated with aqueous vapour, but specifically lighter, may or 
does actually occur in the atmosphere. But our present know- 
ledge of the motions of the currents of the atmosphere, and more 
especially of the distribution of aqueous vapour throughout them, 
is so imperfect as to render this extremely difficult. I must 
leave this task to more competent persons, content if they shall 
first of all pronounce that the essential pomt of the theory, a 
filtration action, is a reasonable explanation of the form and ap- 
pearances of the cirrus. 
I think, however, I may mention several instances in which 
such circumstances may very probably occur, and where this 
theory seems to me to afford a very probable explanation of 
remarkable phenomena, or else to lead directly to such an ex- 
planation. 
Thunder-clouds in Sydney are nearly always preceded by a 
sort of lofty spreading crest of cirrus oy cirrostratus, moving 
rapidly in a direction contrary to the wind at the surface of the 
earth. Beneath this there at last appear large masses of cumu- 
lous storm-cloud, either in huge towering columns, or in irregular 
torn portions exhibiting great agitation, from which the rain and 
