EAINFALL AND FLOODS. 29 



October, and it also fell in each case within a shorter 

 interval of time. There was another circumstance attending 

 the fall in March, 1889, which no doubt contributed 

 materially to render the flood of that month the highest on 

 record. Out of the total quantity of 3 2 inches falling in 

 48 hours, nearly an inch and a half fell within 8 hours, 

 implying a rate of fall during that time nearly three times 

 as great as the average rate of the entire fall. There was 

 yet another feature about the rainfall of March, 1889, 

 calculated to increase the danger. A fall of snow had 

 occurred two days before the rain, and although this had 

 so far melted that the quantity remaining on the ground 

 when the rain commenced was too small to be in itself 

 of much account, there can be little doubt that the ground 

 was frozen underneath it, and therefore in an unfavourable 

 condition for absorbing water. 



On the other hand, in comparing the fall of last October 

 with these two other falls, it should be noted that the first- 

 named was the culmination of a long succession of rains 

 to which we find no parallel in the other cases. The rainfall 

 of March, 1889, was almost isolated. In October, 1882, 

 although the principal rain was preceded by smaller falls, 

 these were not nearly to the same amount as those which 

 preceded the special rainfall of last October. To show this 

 at a glance, the rains of the ten preceding days have in 

 each case been grouped into two divisions. 



Weighing these several considerations one against the 

 other, we are, I think, in a position to form some sort of 

 idea of what the maximum rate of discharge would have 

 been in October, 1882, and in March, 1889, if the arrange- 

 ments had then been such as to permit of the whole quantity 

 of water being discharged without overflow of the river. 

 In a matter of so much uncertainty one hesitates to name 



