I 



TRANSACTIONS AND PROCEEDINGS OF THE [Sbss. lxv. 



If we look at Table L, we find that the over-average 

 growth of deciduous trees in the growing season is during 

 the months of February, March, and December, accom- 

 panied by under-average minimum thermometer, and in 

 October, November, and January by an over-average 

 minimum thermometer. And referring to the monthly 

 tables of plus or minus averages for seven years, we see 

 that the years 1888, 1889, and 1891 are responsible 

 for the October over-average, the years 1888, 1891 for 

 the November over-average, and 1888, 1889 for the 

 January over-average minimum thermometer, as shown in 

 Tables IL, III., and IV. 



Table II. 



^ 12th 

 I September 

 j- to 12th 

 I October. 



Referring to the Meteorological Eecords, there was a 

 severe hurricane on the afternoon of 2nd September 1888, 

 3 '09 inches of rain falling in three hours. There was also 

 O'lo inch rain on 3rd September. May we not conclude 

 that the impetus to growth thus given nine days before 

 measuring day was not at that date exhausted, and that the 

 good growth of this period is partly due to the hurricane, 

 whose cooling effects on the air are not felt during this 

 period ending 12th October ? 



The period ending 12th October 1889 shows both 

 maximum and minimum thermometers so little over par, 

 that, not to be too lengthy, it may be well to pass over this 

 period, and examine the period 12th September to 12th 

 October 1891. The actual rainfall for this period was 

 but 0-67 inch; but in the last twenty-three days of the 

 preceding August there fell 8 '4 5 inches, 4*03 inches 



