204 
during the night. In addition to this process there may 
be still other causes for the formation of the nocturnal 
thunderstorm maximum. 
It is much more difficult to try to explain in this 
manner the nocturnal shower maximum over the 
oceans. In this case, no thunderclouds that could be 
reactivated are left over from the daytime. On the 
contrary, synoptic investigations and weather-recon- 
naissance flights made durmg World War II revealed 
that, during the day, there were no clouds over the 
ocean, whereas at night the region was densely covered 
with severe thunderstorms. This condition recurred for 
several days m succession. The synoptic situation, that 
is, the presence of a cold air dome, was favorable for 
convective clouds in these cases, but the initiation of 
the cloud formation is hardly attributable to long- 
wave radiation. One is much more inclined to ascribe 
it to a remnant of the diurnal temperature variation 
which, over the thermally passive ocean surface, would 
produce a diurnal instability variation that is opposite 
to that observed over the thermally active surface of 
the continent. 
THERMAL PROCESSES ON CLOUD ELEMENTS 
The cloud elements are the actual vehicles of thermal 
changes. As the droplets or crystals fall in the gravita- 
tional field or mix with neighboring air parcels contain- 
ing a different concentration of cloud elements, they 
do not remain in the volume of air in which they 
formed. Thus, they are no longer in equilibrium with 
their environment; evaporation will occur if the new 
environment is drier, condensation if the environment is 
supersaturated; likewise, freezmg or melting of the 
cloud elements may take place. 
Evaporation of Cloud Droplets. Evaporation of cloud 
elements takes place on the surface of clouds. It has 
already been pointed out in the papers by von Bezold 
[4] and later by Robitzsch that the lower temperatures 
which are frequently observed when flying into cumulus 
clouds [16] may be caused by evaporation of cloud drop- 
lets into the drier environment. This process was studied 
quantitatively by Findeisen [12], who found that cloud 
droplets in an environment of 90 per cent relative hu- 
midity have a lifetime of only 232 seconds before they 
evaporate. The conditions for evaporation seem to be 
especially favorable at the edges of cumulus clouds. In 
the strong wind shear that exists between the ascending 
cloud and the descending environment, violent turbu- 
lence and mixing of cloud air with the environment takes 
place. It has been observed quite frequently from gliders 
and airplanes that the turbulence is much stronger in 
the peripheral than in the interior parts of the cloud. 
However, photographic measurements of the ascending 
velocity of cumulus clouds [35] seem to show that the 
ascending motion in the outer portions is only slightly 
retarded, and that the air directly adjacent to the cloud 
is also lifted. The sinking occurs only some distance 
away from the cloud. The model proposed by Christians 
[8] concerning motion in cumulus clouds and compari- 
son with the hydrodynamic jet stream according to 
Schmidt [34] support this assumption. Then, however, 
CLOUD PHYSICS 
the relative humidity in the immediate vicinity of the 
ascending cloud tower cannot be very low, and the 
evaporation as well as the resultant cooling will there- 
fore be insignificant. 
Still less significant is the evaporation over the dome 
of an ascending cumulus cloud, for here the surround- 
ing air is strongly lifted, and its relative humidity is 
increased. This can even lead to the well-known phe- 
nomenon of caps (pileus) which stay for a short while 
over the swelling cumulus head separated from it by 
a thin, cloud-free region, until fusion takes place. How- 
ever, there may be a positive temperature difference 
between the moist-adiabatically ascendmg cloud top 
and the dry-adiabatically lifted air above it, a differ- 
ence which favors evaporation. This, however, is op- 
posed to the results of Petterssen and his collaborators 
[29] concerning the heights of cumulus tops. These 
heights are given approximately by the height at which 
the moist-adiabatic lapse rate becomes equal to the 
vertical temperature gradient of the surrounding air. 
Therefore, no large temperature differences and, in turn, 
no significant evaporation can be expected. On a non- 
swelling cloud, these temperature differences would be 
equalized by austausch processes or even reversed by 
radiation. 
Condensation and Sublimation on Cloud Elements. 
The opposite process, the condensation of water vapor 
on cloud elements that fall from their original air vol- 
ume, occurs everywhere in the cloud. However, until 
now, this process has been given little attention in 
theoretical studies of the structure of clouds and the 
lapse rate within them. According to the computation 
by Findeisen [11], condensation contributes very little 
to the formation of precipitation, particularly in com- 
parison with the essentially much more effective ‘‘chain 
reaction” process of Langmuir’s theory [18]; neverthe- 
less, it cannot be neglected in the thermodynamics of 
clouds. Condensation can only occur when the falling 
cloud elements are colder than their new environment. 
This causes the saturated water vapor to become super- 
saturated with respect to the cooler droplets. The 
amount of this supersaturation and its dependence on 
the temperature difference has been given, for example, 
by Harrison [15]. The temperature difference itself, 
however, depends on the heat transfer between the air 
and the falling water droplets; this transfer is, in turn, 
influenced by the condensation process. It is probably 
for this reason that condensation is rather ineffective 
in the growth of cloud elements. 
The sublimation of water vapor on ice crystals or 
small graupel pellets is much more intense, since cloud 
air which is saturated with respect to water is con- 
siderably supersaturated with respect to the falling 
solid particle, depending on the temperature. In this 
case, the supersaturation is so great that the difference 
in temperature of the crystals with respect to their 
surroundings can be neglected. The thermal reactions 
have been discussed by Harrison [15]. The downward 
inerease of the fall velocity and diameter of the droplets 
and the increase in vapor content of the air cause more 
heat of condensation or sublimation to be released in the 
