80 



J. HOPKINSON — THE CHADWELL SPBING 



effective for percolation, the defect in the amount being almost 

 entirely in the winter months. And not only so, but the effective 

 or winter rainfall of the three years preceding 1890-91 was much 

 less than that of the three years preceding 1895-96. The proba- 

 bility is therefore very strong that before the very wet year 

 1896-97 the store of underground water in this valley was much 

 greater than it was before the very wet year 1891-92. But that 

 it was so is not proved by the flowing of the Bourne in 1897. It 

 is necessary to take into consideration the time of the year at 

 which the excessive amount of rain fell. Though the average 

 fall at Berkhamsted and Nash Mills during the twelve months 

 preceding the outburst of the Bourne in 1897 was half an inch 

 less than it was in the corresponding period of 1891-92, neaily 

 two inches and a balf more fell in the nine months preceding 

 its outburst in 1897 than in the corresponding nine months in 

 the earlier period. It therefore seems that the flowing of the 

 Bourne in 1897 and not in 1892 is just as likely to be due to the 

 rain having fallen at a more propitious time for percolation, that is, 

 later in the year 1896 than in the year 1891, as it is to the 

 presence of an excess of stored-up water from the rainfall of the 

 previous year or years. 



Table III. — Mean Rainfall at Berlchamsted and Nash Mills for 

 various periods during the years 1890-92 and 1895-97 {April 

 to March). 



This view is very strongly supported by the records of percola- 

 tion at Rothamsted. In eight of the last 25 years the percolation 

 through five feet of soil with the surface kept free from vegetation 

 has exceeded 20 per cent, of the average for the 27 years 1871-98 

 (April to March). In 1875-6 it was 26 per cent, above the 

 average; in 1876-7 it was 38 per cent, above it; in 1878-9, 

 39 per cent.; in 1880-1, 49 per cent. ; in 1882-3, 50 per cent.; 

 in 1885-6, 23 per cent. ; in 1891-2, 26 per cent. ; and in 1896-7, 

 70 per cent. In this last year the average was thus exceeded by 

 44 per cent, more than it was in 1891-2, the actual excess of the 

 one year over the other being six inches. Allowing for the fact 



