T. Russell — Prediction of Cold-waves. 4:71 



at the surface of the earth, and a rise in temperature high up in 

 the air. This latter, however, would not be maintained long on 

 account of the greater radiation in the upper air, which would 

 cause the temperature to diminish rapidly to the normal pecu- 

 liar to the altitude and the time of the year. A mixture of the 

 air throughout a height of five miles, computation shows, would 

 cause a fall in temperature of forty degrees, if the temperature 

 . at the surface of the earth is 60°. 



The change of temperature upward in the air is very slight 

 in a region covered by an area of high pressure. This is the 

 region where a cold-wave is prevailing and intermixture has 

 taken place. In fact, there is sometimes in such a region an 

 increase of temperature upward, which is, however, mostly clue 

 to local causes, a ridge or peak radiating strongly into space 

 during the night, and. the thin layer of air cooled by contact 

 . with it, flowing down over the lower portions of the land into 

 valleys, and giving rise to an abnormal contrast of temperature 

 at slightly differing altitudes. 



The fact that most of the severe cold-waves in the northwest 

 start in the afternoon, just after the occurrence of the maxi- 

 mum temperature, tends to strengthen this view. The higher 

 up in the air, the less the diurnal range of temperature. At 

 the time of maximum temperature at the surface of the earth 

 the rate of upward diminution of temperature must be greatest, 

 and consequently at that time the greatest tendency to an in- 

 terchange of the air above and below. The fact that the same 

 density of isotherms with same extent of high and low in dif- 

 ferent cold-waves, do not always produce the same extent of 

 temperature-fall proves conclusively that the fall is not due en- 

 tirely to progress of cold air from the northwest, or dependent 

 solely on temperature conditions at the surface of the ground, 

 but that part of it must come from above. 



Cases can be shown on the Signal Service weather-maps, 

 where the fall of temperature certainly cannot be due to pro- 

 gress of air from places of low temperature to those of high 

 unless it comes down from the upper air, because there is no 

 lower temperature to the northwest to be carried by the winds. 



In the high area, the main part of the cold- wave is due to the 

 convective action in high and low strata. The high is to some 

 extent merely the result of low temperature. The equation 

 will not admit of a term consisting of the extent of high area 

 multiplied by the factor dependent on the density of the iso- 

 therms. There are a few cases, however, where a slight term 

 of this kind would improve the residuals. An area of high 

 barometer no greater than 304 inches has very little power to 

 transmit or produce a cold-wave to the east or southeast of it. 

 Up to that limit, the high pressure is mainly the result of the 



