226 M. Arago on Artesian Wells 



this current the whole of the debris was speedily washed away ; 

 an effect which evidently would not have been produced in a 

 pool which was stagnant. 



At Stains, near Saint Denis, M. Mulot also has met a sub- 

 terranean water-course, in which the borer suddenly sank to the 

 extent of a yard, at the depth of 197 feet below the surface. 



At Cormeilles also, in the department of the Seine-et-Oise, 

 the borer, when it was sunk 220 feet, according to M. Degousee, 

 oscillated like the balance of a pendulum, under the action of a 

 very rapid subterranean current. 



But there is another proof more convincing than any that has 

 been yet adduced of the existence of a subterranean river, under 

 the town of Tours. 



On the 30th of January 1831, the vertical pipe of the foun- 

 tain of the Place de la Cathedrale., having been shortened about 

 twelve feet, the issue of the water, as might be expected, was 

 immediately increased. The augmentation was about a third 

 part ; but the water, previously very clear, having received a 

 sudden increase of its velocity, was agitated. During many hours 

 it brought from the depth of 335 feet the debris of vegetables, 

 " Among which," says M. Dujardin, " might be recognised the 

 branches of thorns, of the length of an inch or two, blackened by 

 their lying long in water ; also the stems and roots, still white, of 

 the plants of morasses, also different kinds of grain, in such a state 

 of preservation, that they could not have been above three or four 

 months in the water. Among these seeds, that of the gallium 

 used for coagulating milk, which grows in marshes, was recog- 

 nised. Fresh-water shells, and land shells, were also found The 

 whole of the debris bore a resemblance to that left after a flood 

 on the banks of small streams and rivers." These facts incon- 

 testibly prove that the reservoir of the third subterranean sheet 

 of water at Tours, is not the result, wholly at least, of simple in- 

 filtration through a bed of sand ; for ere this water could con- 

 tain shells and bits of wood, it is necessary it should move freely 

 in vertical canals.* 



" Since the seeds were not decomposed on their arrival at Tours, M. Du- 

 jardin has fixed some time less than a year as the period they had been in the 

 water. Now, as they ripen in autumn, and made their appearance in January, 

 four montht is likely to have been the correct time. This will probably be al- 



