346 PKOCEEDINGS OP THE AMERICAN ACADEMY 



successive storms. Much valuable work has been done of late years 

 in thus showing the average distribution of the meteorological elements 

 in areas of low pressure. A similar work needs yet to be carried to 

 completion for thunder-storms. 



Three observations a day are sufficient to define the slow weather 

 changes of the large cyclonic storms ; but in thunder-storms obser- 

 vations should be taken every fifteen minutes at most, that is, at the 

 rate of ninety-six times a day, so rapid is the motion of these storms 

 in comparison to their size. A simple method of portraying a storm, 

 thus observed at numerous stations, would be the construction of syn- 

 chronous maps of all its elements : thus the storm as a whole is seen 

 passing over the country. But even then it is not easy to bring all 

 the details of many maps into a single mental picture ; and, moreover, 

 the unfortunate lack of observations must for some time yet make these 

 separate maps very imperfect. Some method of composite portraiture 

 is needed that shall throw all the observations into their proper posi- 

 tion with respect to some controlling line, such as the storm-front, and 

 at the same time allow the records of one station to supply the defi- 

 ciencies of another. I have attempted to accomplish this in the 

 following way. The attitude of the storm-front (rain-front) is first 

 determined by charting all the times of rain beginning, and drawing 

 lines to show the position of the front for every even quarter of an 

 hour ; the direction and velocity of advance are also thereby deter- 

 mined, and generally, in the best developed storms, a certain line may 

 be chosen to represent the middle path of the storm : the average line 

 of rain-front and the middle path are taken as axes of co-ordinates ; 

 time intervals before and after the rain beginning serve as abscissas, 

 while distances north and south of the middle path are ordinates ; and 

 the ratio of abscissas to ordinates is known as soon as the average 

 velocity of storm progress is determined. The axes thus chosen are 

 next drawn on a sheet of tracing-paper : now if this sheet be laid upon 

 a map of the region traversed by the storm, and moved along in the 

 direction of the storm's advance (the line of middle path being coinci- 

 dent on the two), it (the tracing-paper) may be taken as representing 

 the storm stratum on its way across the country: every station that 

 furnishes a record may be imagined to trace a line on the storm 

 stratum about parallel to the middle path, and intersecting the rain- 

 front to one side or the other of the middle path, at a distance from it 

 measured by a positive or negative ordinate ; and all observations can 

 be 2ilaced somewhere on the lines thus traced, their distance in front 

 or behind the rain-front being measured by positive or negative 



