173 



But the above cases arc all exceptional. Ordinarily, as above 

 stated, there are four or five distinct changes occurring in a general 

 way — the first about the end of June or beginning of July, the 

 second about the middle or third week of August, the third during 

 the latter half of September, the fourth about the same time in 

 October. 



Which of these periods shall be wet and which diy is a matter 

 of much uncertainty, and would depend in some measure upon the 

 character of the preceding spring ; but the alternation itself of wet 

 and dry is tolerably regular. Taking a broad view of the question, 

 and judging from the average fall of rain for each month in 

 Cambridgeshire, compared with the same averages at Greenwich 

 and Oxford, I am inclined to consider Jime as more often dry than 

 wet, in which case July with some portion of August will be wet, 

 the remainder of August with more or less of September diy, and 

 the rest of September and a great portion of October wet. The 

 wet of October may continue through November, but not un- 

 frequently, when October has been very wet, and more especially 

 if a considerable portion of September has been also wet, a dry 

 period follows in November, as in the case of an entirely wet both 

 summer and autumn mentioned above, making a fifth marked 

 change of weather in the series. Of course a very wide margin 

 must be allowed for variations in diflPerent years. Nor will it 

 suffice, in estimating the wet and dry periods as above, to place too 

 much reliance on the average rainfall for any particular months, as 

 obtained from a few selected stations, without reference to the 

 yearly average at each of those stations. The question of the 

 connection between these two averages has been taken up and 

 ably gone into by Mr. Gaster, and he shows that, " taking the 

 average of a considerable number of years, the portion of the 

 annual fall which occurs in each month varies considerably at 

 diiferent stations."* Consequently the months of most and least 

 rain will be different in different places. With respect to the time 

 of the maximum fall, Mr. Gaster considers the law to be " that as 



* " On an Investigation of the Monthly Percentage of the Mean Annual 

 Rainfall at Stations situated in the British Islands." — British Rainfall, 1867, 

 p. 33. 



