172 
MAGAZINE OF SCIENCE AND ART. 
of specific gravity and vertical motions; but the effect 
of such changes on the forms of clouds have certainly 
not been duly considered. 
57. It has been shewn how the precipitation of va¬ 
pour, in the internal ascending current ot the cumulus, 
causes the evolution of large quantities of lament heat, 
which increase the sensible temperature of the air, 
already warmer than the surrounding air of the same 
level. The upper motion would, therefore, be acce¬ 
lerated, or “ perpetuated," were it not that some coun¬ 
teracting cause (probably the weight of the watery 
particles) increases the specific gravity of the inass, and 
arrests its motion. But when the watery particles fall 
away from tho air, a part at least of this counteracting 
influence is removed; the air hecamos permanently 
warmer and permanently lighter than it was before, 
anil will bave no tendency to descend a^ain in the man¬ 
ner which produced the fountain-Hke form of the 
cumulus. W * 
It is not likely that the air would be entirely freed 
'from watery particles. If so, it would ascend rapidly, 
on account of its great buoyancy, to a greater elevation, 
in which upward motion all the phonomena of the 
cumulus would be again produced. The air will not 
finally come to rest until it reaches an elevation at 
which its density is exactly the same as the density of 
the surrounding air, and greater than the density of the 
strata beneath, out above which elevation it cannot rise, 
because cumu lose precipitation would then increase its 
density, and cause it to re-descend, as in the cumulus. 
The cloud will, therefore, extend itself outwards at 
that point where it can rest at equilibrum, and the tall 
columnas cumulus will appear with those widely ex¬ 
tended stratiform projections which give it the name of 
cumulo stratus. 
58. I am confident that any one who closely observes 
the natural form and general phenomena of a cumulo- 
stratus in the sky, will allow that this theory very 
naturally accounts for them. The usual shape, in¬ 
deed, strongly suggests the rising and spreading out of 
,a body of air, wnieh has not the tendency to descend 
again, but the truth of this may be made more appa- 
L refit by a simple experimental illustration, 
Place in the section-glass a stratum of water, con¬ 
taining oulv a trace of common salt, and gradually 
introduce below it a second layer, containing the same 
•trace of common salt, together with two parts in 1000 
of sugar. The latter stratum, possessing the greatest 
’ specific gravity, will lie tranquilly below the other. 
Then project up slowly from the lower side of the glass, 
jet of water coutainiug a trace of silver, together with 
about one part in 1000 of sugar. 
59. The last mentioned liquid is of a specific gra¬ 
vity, intermediate to the other two, and will rise up¬ 
wards with a cu mu lose head, through the lower stratum, 
and up to a certain elevation in the higher stratum. 
There its superior gravity will overcome its momentum, 
and it will overflow and redescend, as in the cumulus, 
into the.lower stratum. But here again its downward 
motion will be checked for a similar cause, and, in 
fact, it cannot attain equilibrium, except by spreading 
in the form of a distinct stratum, or stratose extension, 
intermediate to the other two. A cloud of chloride of 
silver is tljua produced, of which some of the varieties 
of form are shewn in fig. vi. and vii. 
The resemblance of this miniature cloud to the real 
form, is often, indeed, very striking, but though I be¬ 
lieve that the analogy is always essentially true, I need 
scarcely explain that in this, as in the other experi¬ 
ments/the conditions are, in many respects, so different 
to those of the atmospheric, clouds, that a striking or 
perfect resemblance cannot often be expected. It is 
evident that in-this experiment we do not represent the 
separation of the aqueous particles as rain, but only the 
motions thence resulting. The jet of liquid projected 
upwards, through a stratum of greater density, repre¬ 
sents the ascending air warmed by pTecijutation yet 
freed from the weight of the precipitated particles, and 
after rising a short distance into the highest and lightest 
stratum which represents more elevated parts of the 
atmosphere, it sinks ajgain in a fountain-like form, and 
spreads out in projections, which tend to form horizontal 
strata. 
60. The cumulostratus may, therefore, bo defined as 
a cumulus, in which a separation of tho watery parti¬ 
cles from the air is taking place to some extent, so that 
the aerial part rising upwards, and retaining the latent 
heat of the condensed vapour, has no tendency to rede¬ 
scend, but collects and spreads out at some point where 
it is in equilibrium with the surrounding air ; strati¬ 
form. projections may thus be formed, either at the 
summit, or any lower point of the columnar cumulus. 
[The name cumulo-strut us might mislead persons 
into the idea that it had something to do with the 
stratus , or the precipitation hy mixture, explained in 
a.separate part of this paper (24, 31), but the terms are 
merely descriptive of the forms, anil there is no con¬ 
nection whatever between their producing causes.] 
The Nurmis Kain-Cloup, on Thunder-Cloud- 
61. It will probably be thought rather presuming to 
extend the present method of investigation, and the pre¬ 
sent mode of discussion, to the explanation of that 
wonderful phenomenon, the thunder-cloud , of whose 
nature hundreds of theories have, no doubt, been formed, 
though no one has ever yet proved satisfactory. I 
would not, indeed, enter lightly upon a subject of such 
acknowledged difficulty, and would uot, without duo 
consideration, run the risk of addiug another truthless 
theory to the already long list. But the same mode of 
enquiry which has lead me to offer an explanation 
on the simplest principles of the Cirrus, the Cumulus, 
and the Cumulo-stratus, leads me so directly and ineri- 
tablv to the Cirro-Cumnlo-srratus, or thunder-cloud, 
that to shrink from attempting the more complex and 
striking case, would only be to acknowledge a want of 
confidence in the explanations of the simple cases. My 
theory of the thundcr-clond is, indeed, nothing more 
than that of the cnmnlostratus combined with that of 
the cirrus, and the same combination of experimental 
conditions gives wliat I shall offer as a miniature repre¬ 
sentation of a thunder-cloud. 
62. Suppose rain to be falling in large quantities 
from a cumulose cloud, which we may suppose to be 
surrounded on even’ side, as well its above, by saturated 
moist air. Tho cumulostratose form will soon be as¬ 
sumed, and we may imagine that considerable bodies of 
cloud will collect towards-its summit, in stratiform 
projections, which lie in tranquil equilibrium, removed 
from the agitated currents in the interior of the cumu¬ 
lus. But if we now further suppose, the subsidence of 
the watery particles to continue throughout the whole 
l*»dv of these projections, their specific gravity will 
diminish bv degrees. Now, these stratose projections 
had assumed a determinate noaition and elevation in 
tho atmosphere, where alone they could lie in equili¬ 
brium, because their specific gravity was exactly equal 
to that of the surrounding air on the same level. When 
their specific gravity is decreased by subsidence cf wa¬ 
tery particles, they can no longer Temain in equilibrium, 
but must rise upwards to a higher level- Here, then, 
we meet the precise condition, which we proved (37.:, to 
produce the cirrus. The upper parts of the body of 
cloud, after once coming to rest, will again acquire a 
tendency to ascend, and filtering upwards, and mixing 
with the moist hut colder strata of tho atmosphere 
above, we shall undoubtedly obtain a cnmnlostratus, 
with a spreading crest of cirrus, which is nothing but 
the cumulo-cirro-stratus, or thunder-cloud. 
63. We mav now have recourse to experiment, to 
illustrate the ‘form of cloud thus produced, carefully 
reraembering, however, that nothing but the mere mo¬ 
tions and forms are imitated in the liquids. 
Pour into the section-glass a stratum of water, which 
contains about two parts in 1000 of sugar, with a trace 
of common salt, and the temperature of which has been 
raised to abou£ 100 decrees of Fahrenheit. Then care¬ 
fully introduce beneath it a second stratum, at the ordi- 
