A474 SUBTERRANEAN WATERS. 
than miraculous. The Abbé Paramelle—a French ecclesiastic 
who devoted himself for some years to this subject and was 
extensively employed as a well-finder—states, in his work 
on Fountains, that in the course of thirty-four years he had 
pointed out more than ten thousand subterranean springs, 
and though his geological speculations were often erroneous, 
high scientific authorities have testified to the great practical 
value of his methods, and the general accuracy of his predic- 
tions.* 
Hydrographical researches have demonstrated the existence 
of subterranean currents and reservoirs in many regions 
where superficial geology had not indicated their probable 
presence. Thus, a much larger proportion of the precipitation 
in the valley of the Tiber suddenly disappears than can be ac- 
counted for by evaporation and visible flow into the channel of 
the river. Castelli suspected that the excess was received by 
underground caverns, and slowly conducted by percolation to 
the bed of the Tiber. Lombardini—than whom there is no 
higher authority—concludes that the quantity of water grad- 
ually discharged into the river by subterranean conduits, is not 
less than three-quarters of the total delivery of its basin.t 
What is true of the hydrology of the Tiber is doubtless 
more or less true of that of other rivers, and the immense 
value of natural arrangements which diminish the danger of 
sudden floods by retaining a large proportion of the precipita- 
tion, and of an excessive reduction of river currents in the 
droughts of summer, by slowly conducting into their beds 
water accumulated and stored up in subterranean reservoirs in 
rainy seasons, is too obvions to require to be dwelt upon. 
The readiness with which water not obstructed by impermea- 
ble strata diffuses itself through the earth in all directions—and, 
consequently, the importance of keeping up the supply of sub- 
terranean reservoirs—find a familiar illustration in the effect of 
* PARAMELLE, Quellenkunde, mit einem Vorwort von B. Cotta. 1856, 
+ See LomBarpint, Importanza degli studi sulla Statistica da Fiumt, p, 27; 
also, same author, Sulle Inondazioni avvenute in Francia, etc., p. 29. 
