EXPLORATIONS ACROSS THE liUKAT BASIN OF UTAH. 



but sink within a short distance. 



in or< 



ler to improve them their origin 1 



mist be exan 



ined. If tliev can be traced to*a 



crevic 



e in the solid rocks, we must try 



to prevent a 



loss of water, and excavate and i 



;ccure 



large cisterns or tanks. This < 



'onld be dor 



frequently, at an expense small c< 



•inpar 



ed with the great benefit derive 



d from such 



the total qu 

 ' be limited, 



v shallow 

 Mnnll tank 



which would have formed springs unless prevented by the heavy cover of loose ab- 

 sorbent material, or it may be the temporary result merely of the surface drainage. 

 Should it be permanent, and iu considerable quantity, it might be made accessible by 

 excavations and secured like the springs; if only temporary, dams could be constructed 

 across the ravine, and thus a large supply of water, at least for a part of the year, 

 could be retained. For experiments of this kind always a narrow point of the ravine 

 should be selected, where the water was likely to be gathered in one stream. A con- 

 stant subterranean discharge of water may occasionally be reached by shallow exca- 

 vations or deeper wells at the junction of branch and main valleys, especially where 

 projecting spurs of hills or some beds of rock or clav obstruct and contract tin- oasse-v. 



