474 T. Russell — Prediction of Cold-waves. 



badly, those of Dec. 5, 1885, and Dec. 20, 1887, when there 

 were twenty-degree fall areas of 863,000 and 723,000 square 

 miles and the method gives no twenty-degree fall. There are 

 twenty-four cases where the observed area being 300,000 square 

 miles or more, the computed is less than half the observed and 

 may be classed as not good. There are four cases of over 300,000 

 square miles where the computed area is more than twice as 

 great as the observed area. In the other cases the agreement 

 is tolerably good. 



Shape and Position of Twenty-Degree Temperature-Fall Area. 



In the case of areas of twenty-degree fall greater than 

 200,000 square miles when the " high " is to the northwest of 

 the low area, the ratio of the axes of the area in the average of 

 cases is 2*5 to 1*0 For the case of the high southwest of the 

 low, the ratio is 4 - to 1'0. In double Y-shaped lows and 

 exceptionally long areas of low pressure, the ratio is about 5'0 

 to 1/0. Lows accompanied by highs no greater than 30*3 

 inches in pressure have the fall areas of same general shape as 

 the isobars of the low area. 



Where the isothermal lines are close together, which is 

 always the case in a great cold-wave, the long axis is pretty 

 certain to be at least two and a half times the length of the 

 short one. 



The position of the long axis is usually from southwest to 

 northeast. It is parallel to the long axis of the area of low 

 pressure or parallel to the general direction of the isotherms.. 

 It is always sure to extend in the direction of the open end of 

 the low area of pressure. The shape of the ten-degree tem- 

 perature-fall area pretty generally conforms to the shape of the 

 twenty-degree fall area. The ten-degree falls very generally 

 overlap on successive days. The twenty-degree fall areas rarely 

 overlap and when they do, only slightly, not more than through- 

 out a strip fifty miles in width in the case of the very greatest 

 cold-waves. The larger the twenty-degree fall areas are, the 

 more likely they are to be just tangent to each other or slightly 

 overlap. When a well-defined area of twenty-degree fall has 

 already occurred, a consideration of its distance from the center 

 of low pressure, or point of prospective greatest temperature- 

 fall, is of service in determining what the length of one axis 

 of the twenty-degree fall may be the next day. 



At the southern limit of a cold- wave, when there is doubt 

 about the position of the lower boundary of a temperature-fall 

 area, the wind direction is useful in locating its position. In 

 the lower Mississippi valley, when the winds having been 

 northwest, and there have been falls of temperature to the 

 north, and the winds turn to the northeast, no farther falls in 

 the vicinity need be anticipated. 



