472 T. Russell — Prediction of Cold-waves. 



low temperature over the area, and there is no appreciable cold- 

 wave without a very considerable low to the east. In the low 

 area, the fall of temperature is due both to the intermixture of 

 upper and lower air and the presence of cold air brought from 

 the northwest by the action of the typical winds around it. 



The greatest fall of temperature. 



An examination of the weather-maps in 217 cases preceding 

 cold-waves shows, that the greatest falls in 134 cases occurred 

 inside of the lowest isobar of the low area of pressure or within 

 100 miles of the center of the low. In 62 cases, it was south 

 of the center of the low 200 miles or more. In 8 cases it was 

 north of the center ; in 4, west of it ; in 3, east of it ; and in 6 

 cases so remote from the center as to have no apparent relation 

 to it. In at least 80 per cent of all the cases of cold-waves, the 

 place of greatest temperature-fall can be located twenty-four' 

 hours beforehand somewhere on the map within a radius of one 

 hundred miles of its true place by taking it at the place of 

 highest temperature within 100 miles of the center of the low. 



The magnitude of greatest temperature-fall is conspicuously 

 dependent upon the temperature gradient on the weather-map 

 preceding it. The values of maximum temperature-fall, pres- 

 sure and temperature gradients given in Table II. might be 

 used for deriving the greatest fall. Taking the fall as propor- 

 tional to the product of the temperature gradient by the pres- 

 sure gradients in five hundred miles, the mean greatest falls of 

 53*6, 43-3, 33-5, 24*0 and 14-3 degrees, give values for the factor 

 respectively of 1*49, 1*66, l - 29, 1'54, 1-55, the mean of which 

 is 1'48. This factor multiplied by the product of the 500 mile 

 pressure and temperature gradient in any case, will give an 

 approximate value of the maximum fall in a coming cold-wave. 

 The value found in this way would be good if the areas between 

 the isotherms were more regular than is usually the case. 



A better method was found to be the following : 



On a line drawn from the point of greatest prospective tem- 

 perature-fall, and perpendicular to the isotherms, about where 

 they are closest together on the map, measure the distances be- 

 tween the isotherms. The temperature at the place of greatest 

 fall after the cold-wave has prevailed, will be the weighted 

 mean of the mean temperatures of the various sections of the 

 line between the isotherms, the weights being taken directly 

 as the lengths of the various sections, and inversely as the dis- 

 tances of their centers from the point of greatest prospective 

 fall. The mean of the temperature from the point of greatest 

 fall to the first isotherm to the northwest of it, or for at least a 

 distance of 200 miles from the point of greatest fall when 

 there is more than one isotherm in the distance, is taken 



