THE PHYSICS OF ICE CLOUDS AND MIXED CLOUDS 
change of wind with height, suggesting that frontal 
surfaces at these heights may be more sharply defined 
than is usually supposed. 
There is a noticeable tendency for the lowermost 
parts of Fallstreifen, even in the cirrus levels, to become 
mammillated, presumably owing to the chilling of the 
air in the trail by the evaporation of the ice particles. 
The Fallstreifen path probably marks a stream of de- 
scending air whose motion is initiated by the drag of 
the particles and is sustained by this chilling. The down- 
draughts im precipitation are very pronounced in 
cumulonimbus and they play an important part in the 
development of these clouds by spreading out near the 
eround and acting like scoops to help fresh cloud 
growth near the shower borders [8]. A similar process 
on a smaller scale, may occur near Fallstreifen trails and 
lead to lines of cloud arranged along the direction of the 
wind shear. 
THE RELATION OF ICE CLOUDS TO 
CONDENSATION MECHANISMS 
High-level clouds are produced by the large-scale 
lifting of air masses, especially at frontal surfaces, but 
it is unusual for the clouds to appear as uniform sheets. 
Thus the first clouds to appear ahead of a warm front 
are characteristically isolated cirrus containing uncinus 
forms, and the overcast of cirrostratus which follows 
often contains Fallstreifen or is of very irregular density. 
Schwerdtfeger [19] attempted to show that isolated 
cirrus are the result of cooling in the high troposphere 
with the production of a convective layer, whereas 
cirrostratus is produced by the lifting of stable air 
masses over frontal surfaces. It would be possible to 
examine this view more critically now that frequent 
upper-air soundings are available. Isolated cirrus clouds 
could arise in uniformly lifted air because of the irregular 
distribution of especially suitable ice nuclei or as the 
result of small disturbances in the general air flow. 
Waved or rippled detail and lenticular-like patches of 
cloud are often seen in cirrus systems, and it is likely 
that orographic disturbances frequently reach the cirrus 
levels and trigger the formation of clouds. Whereas 
lenticular clouds consisting largely of droplets remain 
stationary at the crests of standing waves, similar 
clouds containing many crystals could survive descent 
in the troughs and continue their growth in ice-super- 
saturated layers. 
The growth of individual ice clouds in deep super- 
saturated layers may extend over a few hours before 
general decay sets in, and during this time the clouds 
may be carried hundreds of miles from their birthplace. 
The air may remain supersaturated and the growth be 
maintained even though the mechanism which lifted 
the air has ceased. Thus, the mere presence of ice 
clouds, whose decay is correspondingly protracted, by 
no means indicates an active condensation mechanism, 
and in this respect they differ from droplet clouds, 
which evaporate within a few minutes of the cessation 
of the condensation mechanism. Only young ice clouds 
accompany active condensation mechanisms; they may 
always be recognised by the presence of cirrocumulus 
197 
(droplet cloud) or patches of granular and flocciform 
cloudlets. Beautifully arranged delicate Fallstreifen forms 
are also young ice clouds—as they age they degenerate 
into diffuse streaks or fibres of lifeless appearance, 
usually described as cirrus filosus. 
As far as the writer knows, no methods of forecasting 
the extent and thickness of ice clouds are in use other 
than the making of estimates based on recent measure- 
ments or the relating of the clouds to frontal systems 
in accordance with textbook models. It is unlikely 
that much more can be done at present, as forecasting 
high cloud formation is essentially a matter of forecast- 
ing the vertical displacement of air at high levels, about 
which very little is known. More accurate measurements 
of humidity than are now made at these heights would 
probably also be required. A further difficulty is that 
abundant cirrus clouds often occur far from fronts and 
appear to have their origin in disturbances existing 
entirely above the surface layers. Thus cirrus systems 
may be seen to move with shallow cold ‘‘pools” and 
troughs which are clearly shown on 300-mb charts (but 
not on 500-mb charts). The occurrence of cirrus well 
ahead of cold fronts is also unexplained, but its presence 
seems related to the movement of the front, and its 
disappearance is symptomatic of the retardation of the 
front. However, there are few cirrus systems which are 
not highly organised, perhaps containing great parallel 
bands of cloud or long sharply defined clearing-edges 
and with Fallstreifen indicating systematically changing 
wind shears, so that it is certain that the mechanisms 
producing the systems have coherent structures which 
may be forecast once they are understood. It is un- 
fortunate that the sharpness of the boundaries of cirrus 
systems is so often masked on synoptic charts by the 
reporting of clouds near the horizon which are at a 
great distance—these reports might profitably be made 
only by isolated stations. 
CONCLUSION 
Tt is hoped that many important problems of cloud 
physics will have been sufficiently indicated in the 
paragraphs above. It may be remarked that one class 
of problems concerns the microphysics of cloud par- 
ticles, and another the macrophysics of the occurrence 
and growth of clouds, and that usually the two are very 
intimately related, especially at temperatures below 
OC. Recently, substantial progress has been made with 
the microphysics, but careful attention must now also 
be directed to the study of the formation and growth of 
clouds in relation to air movements. Here valuable 
information may be expected from the researches on 
atmospheric dynamics which are being pursued so vigor- 
ously. 
REFERENCES 
1. AaneNsEN, C. J. M., “Unusual Condensation Trails.” 
Meteor. Mag., 77: 17-18 (1948). 
2. Austin, J. M., ‘‘A Note on Cumulus Growth in a Non- 
saturated Environment.” J. Meteor. 5: 103-107 (1948). 
3. Byers, H. R., and Branam, R. R., Jr., “Thunderstorm 
Structure and Circulation.” J. Meteor., 5: 71-86 (1948). 
4. Cwitona, B. M., ‘Observations on the Incidence of Super- 
