WEATHER WOOLARD 237 



theory have not been successful in practice, although some of 

 thern are of great theoretical importance (Woolard, 1936). Among 

 the investigations with the greatest immediate practical promise, the 

 most noteworthy, perhaps, are those by Sverre Petterssen of the 

 Norwegian Meteorological Service, following earlier similar work 

 by Giao and Angervo. On the basis of the ordinary kinematical 

 theory of fluid motion, he has developed rigorous methods for a 

 kinematical analysis of the synoptic chart in conjunction with the 

 frontal and the thermodynamic analyses, and has derived procedures 

 for detecting indications of important developments, and for calculat- 

 ing the movements and transformations of pressure systems and of 

 fronts. His formulae (Petterssen, 1933, 1936), when used with 

 judgment by an experienced forecaster, are a valuable aid (Weight- 

 man 1936, 1936a), although they do not contribute to the dynamical 

 interpretation of the kinematic phenomena. Plate 3, figure 1 shows 

 some results of applying Petterssen's methods which may be com- 

 pared with the actual occurrences shown in plate 3, figure 2. 



Under the influence of these and other modern developments, 

 weather forecasting, at first exclusively empirical, is now progressing 

 along sound physical lines; and it has become practicable during 

 recent years to supplement, though not replace, the empirical methods 

 of weather forecasting, as developed in the nineteenth century on 

 the basis of experience alone, with practices based on an understand- 

 ing and an explicit application of the physical laws to which the 

 phenomena conform (Weightman 1936, 1936a). For a long time 

 to come, however, weather forecasting must continue to be a com- 

 bination of physical reasoning with methods based on accumulated 

 practical experience with synoptic charts. We cannot yet, and per- 

 haps may never, safely do without the empirical judgment of the 

 experienced forecaster. 



Meanwhile, it maj' reasonably be expected that all investigations 

 designed to elucidate the physical mechanism of weather phenomena 

 will not only contribute toward a better understanding of these 

 phenomena, but also lead, because of this increased understanding, to 

 an eventual increase in the precision, and an improvement in the 

 accuracy, of forecasts; and will moreover have the distinct advan- 

 tage of making weather forecasting less of an esoteric art, irrespective 

 of the extent to which they may effect actual improvements over 

 the results already obtainable by empirical practices. The problems 

 presented by weather phenomena and their prediction offer a fertile 

 field of research for the application of the classical physics. In addi- 

 tion, the weather provides many examples of physical phenomena, 

 on an impressive scale and often of a spectacular character (pis. 1, 4), 

 which afford excellent illustrations of important physical principles ; 



