THE CLASSIFICATION OF CLOUD FORMS 
clusive and exclusive without overloading them unduly. 
To some extent this is a matter of degree, as to how 
lengthy and detailed a definition may be without being 
so clumsy as to defeat its purpose. But it is also one of 
purpose, whether the definition is framed to describe 
a typical form or to delimit a class. A very suggestive 
result is obtained if one strips from the international 
definitions of the genera all characteristics that do not 
explicitly serve to distinguish one genus from another. 
The remainders are: 
Cirrus: a detached high cloud that shows no shading. 
Cirrocumulus: a layer (or groups) of cloud masses in 
regular arrangement, the elements of which show no 
shading. 
Cirrostratus: a continuous sheet of high cloud not 
showing relief and not thick enough to blur the outlines 
of the sun or moon. 
Altocumulus: a layer (or groups) of cloud masses in 
regular arrangement, or a continuous layer showing re- 
lief on its lower side, the smallest elements of which 
have a diameter less than ten solar diameters, and the 
thickest elements of which show shading. 
Stratocumulus: a layer (or groups) of cloud masses in 
regular arrangement, or a continuous cloud sheet show- 
ing relief in its lower side, the smallest elements of which 
have a diameter of more than ten solar diameters. 
Stratus: a uniform layer of low cloud, from which, if 
precipitation falls, it falls as drizzle. 
Nimbostratus: a nearly uniform layer of low cloud, 
from which, if precipitation falls, it falls as continuous 
rain or snow. 
Cumulus: a cloud of vertical development whose 
upper portion is not fibrous in appearance. 
Cumulonimbus: a cloud of vertical development whose 
upper portion is fibrous in appearance. 
The altered character of these definitions is at once 
apparent. They are less vividly descriptive than the in- 
ternational definitions, but emphasize more clearly the 
elements that distinguish one genus from another. 
Genetical Classifications. The present situation of the 
genetical classifications somewhat resembles that of the 
classifications based on appearance shortly before the 
first efforts at mternational agreement. In the several 
classifications, major processes are universally recog- 
nized: the overturning of air by through-going thermal 
convection, the stirring that characterizes turbulent 
layers, and the widespread lifting that accompanies 
horizontal convergence of the lower air layers in the 
forward portion of a cyclone. The manner in which the 
processes are arranged and subdivided, the relative 
weight given to different manifestations of a process, 
the number of additional processes thought worthy of 
mention, and to some extent, the nomenclature of the 
resulting clouds are matters on which full agreement 
has not yet been reached. Some aspects do not come 
clearly to the fore in any of the classifications, notably 
the difference between forced turbulence and thermal 
circulation in sheets of air, the manifestations of shear- 
ing motion, and the part played by radiation. The de- 
scriptions of cloud forms resulting from the various 
processes are somewhat incomplete, especially with re- 
gard to the dissipative stages in the life of the clouds. 
1165 
Physical Classifications. Bergeron’s classification of 
clouds according to their constitution, whether com- 
posed of complete crystals, skeletal crystals, or water 
droplets, covers only a single aspect of cloud constitu- 
tion, namely colloidal instability. Before a comprehen- 
sive physical classification can begin to take shape, the 
collection and assimilation of a more adequate factual 
basis is necessary. A start im this direction has been 
made by Nakaya? with regard to the laws that govern 
the crystalline structure of ice crystals and snow, and by 
Diem [8] and others with regard to the drop sizes char- 
acteristic of the various cloud forms. Information about 
the drop-size distribution and its origin [12] should also 
find application here. 
A View into the Future 
It is of course easier to criticize than to improve. 
Nevertheless, it is possible to foresee and outline cer- 
tain steps that will contribute toward progress in the 
field of cloud classification. 
Classification According to Form. One of the most 
pressing needs is an approach to a more rigorous logical 
treatment of the international definitions, making each 
category of the classification (family, genus, species, 
etc.) as nearly as possible a geometrical partition. The 
problem presents serious difficulties if the definitions 
are not to become cumbersome. It can be simplified, 
however, if the definitions, particularly those of the 
genera, are stripped of all provisional statements (¢.e., 
those including “sometimes” or “usually,” etc.) and 
all statements that do not contribute to rigorous dis- 
tinctions between classes. A further simplification may 
be achieved by abandoning the arbitrary restriction of 
the number of genera to ten and defining additional 
genera. For example, if detached lenticular clouds were 
given status as a genus, the definition of altocumulus 
(or other genera) would not have to be stretched to 
include these clouds. 
The effect of eliminating nondefining elements from 
the definitions of the genera would be to decrease the 
emphasis now placed on the genus category and increase 
the importance of the species category. Such a change 
conforms with the increased emphasis placed on species 
in the formulation of the codes for states of the sky. 
Toward Uniform Practice. As long as the materials 
and practices used in the training of meteorologists, 
meteorological observers, and other users of meteoro- 
logical information remain as divergent as they are 
today, it will be very difficult to attain uniformity in 
the use of the international classification. The publica- 
tion of an authoritative atlas and its distribution to 
libraries and similar repositories is not adequate as a 
means of achieving even a reasonable degree of uni- 
formity. 
The situation would be helped by the republication of 
an official abridgement of the International Atlas in a 
form suitable for wide distribution, so that it would find 
its way more generally into the libraries of weather- 
2. Consult “The Formation of Ice Crystals” by U. Nakaya, 
pp. 207-220 in this Compendium. 
