424 Royal Society. 



circumstances regulating this change, which have been observed and 

 experimented upon bj- the author during the last forty years, are 

 explained in tlie paper. In the year 18.'55, he adopted a method of 

 ascertaining the supply of subterranean water, by the use of a very 

 simple bat effective instrument, contrived and used by the late 

 Dr. Dalton for that purpose: and he has supplied a series of 

 monthly observations, extending over fourteen years, both of the 

 rail) falling on the surface, and of that which is found to percolate 

 through to a lower level for the supply of springs and livers. 

 From his observations — 



inches. 

 The annual average of rain in the north-western part of 



Herts is 25-92 



The average fall of rain in the first six months of the year is 11 '12 



The average from July to December inclusive 14:*80 



The average in six months, from April to September inclusive 12*17 

 The average percolation through the Dalton gauge, from 



April to September inclusive, is 0"62 



And from October to the following March inclusive 9*61 



Total average annual amount of percolation 10*23 



The rivers and springs supplied from the chalk are generally found 

 to be in fullest flow about June, and to be most reduced in Decem- 

 ber; and the cause of this variation is considered to be the time 

 that the descending rain requires to percolate through the crevices 

 and fissures of the lofty chalk, hills, and to spread laterally in the 

 reservoirs till it reaches the outlet springs. 



The variation in the quantity of rain is found to range from 21*10 

 inches to 32*10 inches. The variation in the amount of percolation 

 is found to be far greater, viz. from 3*10 inches to 19*28 inches. 

 Consequently there is a variation in the springs, and in the flow of 

 the rivers, much greater than in proportion to the fall of rain. The 

 author shows that the season of the rain falling is the main incident in 

 the supply of these perennial springs ; and that their outflow is pro- 

 portional to the percolation which takes place during winter into 

 the lower beds, the summer rains being evaporated or taken up by 

 vegetation. 



He gives it as his opinion, that it is possible to drain off, by arti- 

 ficial means, great part of any river flowing out of the chalk; such 

 rivers being truly the natural drain and outlet of the subterranean 

 reservoir therein. He shows, by precise measurements and care- 

 fully recorded observations, that the subterranean water has a move- 

 ment, with a declivity of 13 feet 6 inches to the mile, in the direc- 

 tion of the dip of the strata and of the fall of the streams ; and he 

 states that the crevices or water-channels in the chalk are larger in 

 the neighbourhood of a stream. He therefore assumes that if a 

 large and deep well were sunk in any such locality, and the water 

 in it, by being pumped off by steam-power, were brought down to 

 and kept at a lower level, a deep-seated artificial vent being thus 

 formefl, the water would be so draitied off from the reservoir that 

 the springs would be dried up, and the river be partially or entirely 

 deprived of its flow of water. 



