AND THE HERTFORDSHIRE BOURNE. 71 



to September, being a little over six inches iu the winter aiul about 

 eight-tenths of an inch in the summer. There has frequently been 

 no percolation whatever in tlie summer months ; this has occurred 

 in sixteen years out of forty-two, and without, so far as we 

 know, very materially affecting the flow of the Chadwell Spring. 

 In the summer of 1875 there was no percolation and the flow of 

 the spring did not in any month average less than two and a half 

 million gallons of water a day ; in the summers of 1880 and 1881 

 there was no percolation and the flow of the spring was never less 

 than two million gallons a day ; and in the summer of 1883 there 

 was no percolation and the flow of the spring was in one month 

 only below two million gallons a day, and in that month, 

 September, only just below it. The fall from two and a half to 

 under two million gallons a day with no percolation in the summer 

 months may be due to the gradual decline in the volume of the 

 spring, which will be considered presently. It is not due to 

 a decreasing fall in the previous winter, for the winter of 1882-83 

 was the wettest on record. I mention this fact merely to show 

 that the total failure of the spring from at least the beginning of 

 September, and probably a month or two earlier, cannot be due to 

 the absence of percolation in the summer. (See Appendix.) 



AVhat is here assumed to be themean rainfall in Hertfordshire is 

 the mean of that recorded at ten stations which have a record for 

 at least a quarter of a century, the longest being the Nash Mills 

 record of 64 years. The winter rainfall is that of the six months 

 October to March, the summer rainfall is that of the six months 

 April to September, and the yearly rainfall is that of the twelve 

 months April to March, except when otherwise stated. 



The mean rainfall in Hertfordshire during the winter of 1897-98 

 was 7"82 inches. During the last sixty-four years only once was 

 there a less winter fall, and that was 6 '08 inches in the winter of 

 1879-80. During the calendar year 1880 the Chadwell Spring 

 did not yield so little as two million gallons a day in any month, 

 the least being 2,178,000 in June. The summer of that year was 

 a wet one, but at Nash Mills there was no percolation through the 

 soil gauge. The pi'evious summer, however, was excessively wet, 

 22-56 inches of rain having fallen, being about 70 per cent, above 

 the average. In fact it was much the wettest summer on record, 

 and the percolation at Nash Mills was 6-94 inches, which is more 

 than an average winter's percolation, and double that of any other 

 summer since the experiments were begun. In the summer of 

 1897, on the other hand, the rainfall was below the average, only 

 11 '15 inches having fallen. The rainfall of the year 1879-80 was 

 therefore 28-64 inches, and that of the year 1897-98 only 18-97 

 inches. This is the smallest recorded fall for a similar period, but 

 twice it has not been an inch more ; in the year 1870- 71 it was 

 19-70 inches, and in the year 1854-55 it was 19-33 inches. 



If we now take a period of eighteen months, including two 

 winters and one summer, it will be found that the rainfall during 

 the eighteen months ending 31st March, 1898, was but little 



