THE CLASSIFICATION OF CLOUD FORMS 
By WALLACE E. HOWELL 
Blue Hill Meteorological Observatory, Harvard University 
Introduction 
Tf clouds occurred in a naturally limited number of 
forms, like regular polyhedrons, the problem of classi- 
fying them would be simple. Such, however, is not the 
ease. Clouds occur in an infinite variety of forms, and 
the classifications that have been devised to describe 
them bear the stamp of human and fallible notions 
about their appearance, origin, and structure. Order in 
clouds is the product of man’s will, and man’s will can 
change it. 
In its present status, cloud classification comprises 
not only classifications now in use, but also notions and 
concepts abandoned in the past that are still valid and 
may be given new value in the future. In taking stock 
of our present position, therefore, valid notions from 
the past must not be overlooked. A review of the pres- 
ent status may properly take the form of an historical 
summary of selected developments. 
Assessment of our present position has two aspects, 
logical and pragmatic. How completely and logically 
has nature been described? How suitable are the 
classifications for the uses to which they are put, and 
wherein lie their shortcomings? Thoughtful answers to 
these questions will point the way toward future prog- 
ress. 
The encoding of cloud forms for purposes of transmis- 
sion or recording of observations must not be confused 
with classification. Encoding is a sequel to classifica- 
tion; once classes are established, a code may be devised 
by assigning to each class a number, or, if there are not 
enough numbers to go around, by grouping two or more 
classes under a single number. However, classification 
should never be artificially restricted to meet the de- 
mands of a numerical code. Natural phenomena, fingers 
and toes excepted, do not occur by tens, and the attempt 
to force them into such a mold will usually be wrong. 
In the present work the problems of encoding are not 
treated. 
The Present Position 
The classification of cloud forms has been developed 
from two main points of view. One, beginning in the 
early 19th century, concentrated on the appearance of 
clouds as seen from the ground; its latest stage is the 
international classification of cloud forms. The other 
viewpoint stressed the processes that form clouds. It 
began in the early 20th century, and is still in a rudi- 
mentary stage. No general agreement has been reached 
and no international codification attempted. In the 
following paragraphs an attempt will be made to point 
out the accomplishments from each viewpoint that ap- 
pear to have value now and for the future. 
A beginning has been made towards what may event- 
ually become a classification of clouds based on their 
physical constitution, 7.e., drop-size spectrum, liquid- 
water content, form and size of ice crystals, etc. The 
factual basis for this approach is still far from complete, 
and few classifications based on it have yet been formu- 
lated. 
The International Classification and Its Antecedents. 
The first classification that is still vital was that pro- 
posed by Luke Howard [11] in 1803. He recognized three 
simple modifications, which he named cirrus, cumulus, 
and stratus. The definitions he gave would pass as cur- 
rent today, except for minor details. He next defined 
two intermediate forms, cirro-cumulus and cirro-stratus, 
and two combined forms, cwmulo-stratus and cirro- 
cumulo-stratus or nimbus. His intermediate forms sur- 
vive today as the middle clouds and the cirrocumulus 
and cirrostratus, but his combined forms have been 
largely lost except for surviving traces of Howard’s 
theory that rain, both of frontal and thunderstorm 
origin, is caused by the electrical interaction of two 
cloud layers. His definition of the nimbus therefore em- 
braced both cumulonimbus and nimbostratus. It is 
worth noting that Howard did not call his system a 
classification nor claim that it embraced all clouds, but 
intended it only ‘to impose names on such of them as 
are worthy of notice.” 
Howard’s work became the standard reference; but 
the authors of general meteorological texts, which be- 
gan to appear in the middle 19th century, often para- 
phrased it carelessly and erroneously. It was through 
careless copying of Howard’s definitions by various 
authors that the stratus gradually came to be described 
as “‘a cloud above the ground,” whereas Howard clearly 
applied the term to fog lying on the ground as well as to 
low layer clouds. The only concrete contribution during 
that period was made by Kaemtz, who differentiated 
stratocumulus from stratus, recognizing only the cumu- 
lonimbus half of Howard’s nimbus and defining the 
former broadly enough to include most rain clouds. 
The initial success of the French meteorological bu- 
reau in the late 1850’s stimulated the study of clouds as 
well as other branches of meteorology. It opened a pe- 
riod when many new ideas were advanced, a fertile if 
somewhat febrile period marked by originality in the 
invention of new classifications more than by perfection 
and consolidation of existing ones. Noteworthy because 
of its later influence was the system of Poéy [16] who 
regarded cirrus and stratus as the only fundamental 
cloud forms, the one being of ice and the other of water. 
He anticipated Bergeron [3] by ascribing the origin of 
rain to the coexistence of these two clouds which, when 
coexisting, he called “pallium.”’ Clayton [7] suggested 
stratus and cumulus as the two fundamental forms, a 
1161 
