72 ROYAL SOCIETY OF CANADA 



in the same direction at the higher station by an interval of time, some- 

 times as great as 24 hours. This interval, which seems to be very 

 variable, depends probably upon wind conditions. 



An examination of the diagrams, beginning with the 1st February, 

 shows a rapidly increasing minus difference, wihile at the lower station 

 the temperature was falling rapidly. In this case the change of tempera- 

 ture at the lower station was anticipated by about fivie hours. There 

 was, h.owever, a veay sudden drop of temperature, accompanied by 

 change in the wind direction. Continuing the examination of the 

 curves it will be seen that the air temperature was abnormally low 

 while the thennometer difference remained very large negatively. On 

 the 3rd and 4th of the mo.nth slight diminutions in the difference curve 

 corresponded to increased temperatures at the lower station. It will 

 be noted that the difference curve continued to give a large negative 

 value until about midnight of the 13th, when it decreased rapidly until 

 6.30 a.m. of that day, at which time it reached the zero line. i^at- 

 withstanding this change, the temperature at the lower station followed 

 the diurnal curve until noon of the 14th, when the first break in the 

 cold spell was experienced. During this interval (13th at 6 a.m. to the 

 14th ait noon) the difference curve remained small and so continued 

 during the subsequent warm period until lower air temperatures were 

 again established. 



The next marked change in the difference curve began at 1.40 a.m. 

 on the 18th, and will be seen to have l:)een the pnecursor of higher tem- 

 tperatures which set in at the lower station at 11 a.m. Xo notable depart- 

 ures from narmial temperature conditions oectirred after this until 

 12 p.m. February 20th, when the difference suddenly diminished, and as- 

 sumed the positive sign at 4 a.m. ZSTotwithstanding this marlœd change 

 in the difference curve, the temperature at the lower station continued 

 to fall slightly until 7 a.m. It will here be seen that the difference 

 thermometer predicted by about 12 hours the warm wave which follow- 

 ed. After the return to colder weather, the conditions were again normal 

 until the nights of the 26th and 27th, when there was a similar oc- 

 currence to the foregoing, the warm wave being anticipated oh the 

 difference record by about 12 hours. 



The following table shows the normal differences of the tempera- 

 tures at the two stations for each of the months covered by records 

 thus far obtained. 



