MAGAZINE OF SCIENCE AND ART. 
173 
nary temperature of about CO degrees, and containing 
the same trace of common salt, together with 2% parts 
of sugar in 1000. Lastly, project upwards, from the 
lower edge of the glass, a jet of distilled water, at the 
ordinary temperature, containing merely a trace of 
nitrate of silver in solution. 
The jet of the last-named liquid forces its way ra¬ 
pidly up through the lowest cold and dense stratum, 
with the usual cumulose termination, by virtue both of 
its momentum and its- considerable buoyancy. The 
heated upper stratum, however, although containing 
sugar, is of less specific gravity than pure water at the 
temperature of GO degrees, and, therefore, on entering 
this stratum, the upper progress of the jet is checked, 
and flowing over in the fountain-like, or cumulos,. ^m, 
it redesceuds into the denser stratum. A cumulosua- 
tose cloud is thus formed for the first few moments, as 
in a former experiment (57), but the heat of the highest 
stratum is soon communicated to the higher parts, por¬ 
tions of the injected liquid, and decreases their specific 
gravity. Hence they will acquire the power (as ex¬ 
plained in 35) of filtering upwards in curose streamlets 
to the surface, and the peculiar form is produced which 
is shewn in fig. viii. 
[It may he noticed that the curose threads obtained 
in the section glass are much larger and coarser, or 
even somewhat different in appearance from those pro¬ 
duced in a round glass vessel, but the difference does 
not appear to be an essential one, and is doubtless 
caused entirely by the varied form of apparatus.] 
€4. It will, doubtless, be difficult for persons not accus¬ 
tomed to observe the clouds, and consider their apparent 
nature, to recognise in the results of this experiment any 
resemblance whatever to the varied and majestic forms of 
the thundercloud ; but we have, at all events, the three 
essential actions, viz. ; 1 st. A rising stream representing 
the cumulus : 2nd. A tendency to spread laterally between 
the higher and lower strata representing the stratus ; and 
3rd. A wide-spreading crest of cirrus ascending from the 
whole* And the resemblance would, I think, appear 
striking if a person with the idea of this explanation in 
his mind were to observe closely a distinct instance of 
thundercloud conveniently situated in the sky. 
It is scarcely necessary to notice how much the natural 
form of the cloud will be modified by the conflicting hori¬ 
zontal currents which nearly always prevail In the atmos¬ 
phere. Under an Infinite variety of conditions an infinite 
number of different thunderclouds will be formed, but the 
mind must exercise upon these a process of abstraction in 
order to arrive at the forms or motions which are essen¬ 
tial. Such, I believe, are represented as completely and 
truthfully as could be expeeted in the miniature experi¬ 
ment just explained, 
65. Rain occasionally falls from a simple small cumu¬ 
lus, which in that case generally exhibits, sooner or later, 
a cirrose crest. When a number of cumuli or cumulo- 
stratus cover the sky, 1 have distinctly observed that those, 
and those alone, spread upwards into cirrus at the top, 
from which showers of rain ate falling beneath. Also 
Howard, the meteorologist, who is the first authority in 
the subject of the clouds, remarks of the Nimbus, that 
“ the cirri vegetate* as it were, io proportion to the quan¬ 
tity of rain falling, and give the cloud a character by 
which it is easily known at great distances.” Nothing 
could express so agcinrately as the term vegetate the man¬ 
ner in which the upward increase of the cirrus will proceed 
simultaneously with and in proportion to the downward 
motion of the precipitated vapour as rain. Given a cer¬ 
tain depth of rain fallen, and we could calculate pretty 
accurately the upward momentum which must have been 
communicated to any given body of air during its pre¬ 
cipitation. 
66. The subsidence of cloudy, that is, watery particles 
suspended in the air is made to play a very important part 
in these theories; it may be well to explain and illustrate 
this subsiding action a little. The extreme case is when 
the particles aggregate into drops of greater or less size 
which then fall with great velocity as rain ; but persons 
who are accustomed to examine the daily varieties of 
clouds closely will often detect appearances which unmis¬ 
takably indicate subsidence without the actual formation 
of rain. 
A dense veil of cirrostratose cloud will often towards 
evening exhibit over its lower surface a very peculiar 
"botyoidal” appearance, which is evidently caused by 
portions of air sinking on account of the cloudy matter 
with which they are loaded. These are at first of a dis¬ 
tinctly-terminated bag-like form? but sooner or later 
evaporation gradually takes place, and a bundle of cloudy 
streaks alone remains, resembling descendent cirrus, and 
producing a furry appearance. 
The same subsidence seems to take place occasionally in- 
the interior of cumuli, which show no outward sign of it. 
And thus may be explained the remark of Howard,. 
(“ Climate of London," I. lxix) that u a quickly evapo¬ 
rating cumulus sometimes leaves a regular cirrus behind 
formed out of the remnants of the cloud, which, in the 
intermediate state and just when it begins to show the sky 
through it, exactly resembles the pores and fibres of 
sponge.’* 
From these and other instances itmay be concluded that 
the subsidence of watery particles and the aggregation of 
watery particles into drops of rain arc two distinct pro¬ 
cesses. I am inclined to think that simple gradual subsi¬ 
dence is sufficient to account for the cumulo-stratose form, 
which becomes converted into the nimbus by the addition 
of a crest of cirrus when the more violent process of ng - 
gregation into rain commences. 
But it would occupy a large-sized volume If I were to 
enter into every development of these theories, to mention 
and attempt to reconcile every slight difficulty that may 
be felt, or to adduce quotations, descriptions, instances, 
&c., as authoritative proofs of all my statements. I must 
now proceed with further general discussions. 
Effect of Kain on tjie Temperature op the 
Atmospueue. 
67. There is one effect of rain worth noticing, and 
which can now be easily understood, namely, that it must 
tend to increase the temperature of the higher strata of 
the atmosphere. The temperature falls as we ascend into 
the atmosphere, because air when expanded absorbs heat; 
and if this were the only cause of change, some constant 
ratio would be preserved in every part, and we should find 
that however a body of air were transferred from one ele¬ 
vation to another, it would always have the same tempera¬ 
ture as the surrounding air. 
But if aqueous vapour be mingled with the air, we have 
a second cause which will affect the higher strata of the 
atmosphere, namely, the latent heat <»jven out when 
vapour is condensed. If a body of moist air be elevated a 
certain distance, the fall of temperature will occasion a 
partial condensation of the aqueous vapour, which, how¬ 
ever, in becoming reduced to the liquid form will give out 
sufficient heat to retard sensibly the diminution of tem¬ 
perature. It is impossible, therefore, that the higher parts 
of the atmosphere should be so cold as Daniell’s theory, 
depending merely on the absorption of heat, requires. For 
moisture is always mingled Inextricably with the air in 
greater or less proportion, and the one cannot undergo 
the changes of density and pressure which that theory' sup¬ 
poses without the other being affected and exercising a 
perturbing influence. 
68. But, from what has gone before, we may easily an¬ 
ticipate two cases which occur when air is elevated or con¬ 
densation otherwise occasioned. 
1st. When the condensed vapour subsides through the 
air and falls as rain. 
2nd. When the watery particles remain suspended as 
cloud-matter. 
In the first case it is evident that the air has received a 
permanent accession of warmth, and In every position will 
afterwards exhibit a higher temperature than Daniell's 
theory requires, or than it would possess if it had never 
contained aqueous vapour. Now, in every locality it is 
well known that there is a regular and determinate fall of 
rain which represents the amount of aqueous vapour con¬ 
densed in but separated from the higher strata of the at¬ 
mosphere ; consequently a determinate quantity of latent 
