EXPERIMENTAL CLOUD FORMATION 
which are formed in the process of dissipation of clouds 
within which instability is set up, either by the joint 
effect of radiation and absorption, or by subsidence. 
TasieE I. Cuasses or CONVECTION CLouD 
Rate of varia- 
tion of wind Type of cloud International Cloud Atlas 
with height 
Zero Isolated cumulus N 10, 25, 55, 56 
Very small | Cloudlets arranged in N 16 
lines 
Small Short rolls, or long ir- N 11, 18, 29 
regular wavy rolls, 
nearly across wind 
Large Long rolls roughly N 387, 51, 57 
downwind A 14, 16; C 23 
The last column gives reference to pictures in the 
International Cloud Atlas. 
Since the rate of change of wind with height is 
closely proportional to the horizontal gradient of tem- 
perature, we may say that the rate of change of wind 
is large or small according as the isotherms are closely 
clustered together or far apart. It follows that con- 
‘vectional cloud will be aligned along the isotherms 
when these are clustered closely together, but at right 
angles to the isotherms when these are widely separated. 
Some attention was devoted by Deslandres [6] to the 
possibility of explaining the structure of the solar 
chromosphere as an aggregation of convection cells of 
the type we have called Bénard cells. This deserves 
passing mention as an early attempt to use the results 
of Bénard’s small-scale experiments to explain phe- 
nomena on a much larger scale. There is, in addition 
to this, a considerable literature in French dealing 
with attempts to explain lunar craters by the same 
mechanism. 
Some Further Notes On Convection 
Durst, in Part Il] of a memoir by Giblett and others 
[7], has interpreted observed variations of wind in the 
horizontal as due to convective motions such as are 
shown in Fig. 10, in which the simple convection cell 
is distorted by a small shear of wind. Durst’s results 
do not admit of brief summarising, and the reader is 
referred to the original memoir for details of the work. 
The effect of entrainment of air by an ascending 
current, in diminishing the local variation of wind with 
height has been discussed by Malkus [10] in a paper of 
considerable interest, in which the slope of an ascend- 
ing current to the vertical is derived theoretically, 
and in which it is shown that clouds may move with a 
speed different from that of their environment. The 
magnitude of this difference is shown to be a function 
of the vertical shear and the rate of entrainment of 
mass into the jet of rising air. If the entrainment of the 
ambient air is greatest on the upwind side, Dr. Malkus 
finds that the cloud will tend to grow on the upwind 
side, and to dissipate on the downwind side. 
Malkus, Bunker, and McCasland, in a paper on 
1261 
observational studies of convection [11], have discussed 
clouds formed by forced convection due to the islands 
in the neighbourhood of Woods Hole, and have elab- 
orated the discussion of the effects of wind shear previ- 
ously given by Malkus. 
The Theory of Bénard Cells by O. G. Sutton 
In a recent paper Sutton [15] has shown that all 
the critical temperature differences (AT), as observed 
by Chandra and by Dassanayake, should satisfy a 
relationship: 
AT _ ¥4 
where 7") is the temperature at the base of the chamber. 
In the course of the development of his theory, Sutton 
shows that, when the curve and line whose equations 
are 
y = 8.1 X 103(AT)* — 0.0117, 
y = AT/T) 
are plotted for any given 7,, their intersections will 
yield the critical temperature differences. The straight 
line intersects the curve in two points, is tangential to 
the curve, or has no intersections. The striking feature 
is the possible existence of two real roots, the higher 
root corresponding to the polygonal or Type I convec- 
tion, and the lower root to the Type II convection, 
of the type shown in the greater part of Fig. 5. 
While the values of AT varied from 3.5C to about 
90C, the base absolute temperatures varied from 295K 
to 380K, and the depths of the chamber varied from 
0.4 em to 1.6 em, Sutton’s plot of A7’/T> against (AT) 
gives an amazingly accurate linear relationship between 
the plotted variables, the equation for which is 
== = Lil $< OND) = OUT. 
Sutton’s paper is by far the most important theoreti- 
eal contribution to this subject since Rayleigh’s original 
paper. 
Desirable Extensions of the Investigations Mentioned 
Above 
The most desirable investigations would be the col- 
lection of details of variation of temperature and wind 
with height within the types of cloud discussed above, 
the depth of the cloud layer being carefully measured 
at the same time. 
The essential details of the laboratory phenomena 
have been collected, but the theoretical discussions re- 
ferred to fail to lead to such results as are summarised 
in Figs. 7 and 12, for convection of Type II. 
REFERENCES 
1. Bénarp, H., ‘Les tourbillons cellulaires dans une nappe 
liquide.’”’ Rev. gén. Sci. pur. appl., 11: 1261-1271, 1309- 
1328 (1900). 
2. —— “Sur les tourbillons cellulaires, les tourbillons en 
bandes, et la théorie de Rayleigh.”’ Bull. Soc. frang. phys., 
No. 266, pp. 112-115 (1928). 
3. Brunt, D., Physical and Dynamical Meteorology. Cam- 
bridge, University Press, 1939. (See pp. 219-220) 
