The Underground Drainage 



and on tracing this on to our surface map we found 

 that the underground river ,> 



did not follow the dry valley 

 on the surface, but ran al- 

 most due north towards the 

 stream which we had pre- 

 viously observed from Brown 

 Hill. 



With a view to deter- 

 mining the relation of the 

 surface to the underground 

 drainage, we mixed two 

 pounds of fluorescein pow- 

 der with an equal weight 

 of potassium carbonate, the 

 object of the latter being to 

 render the fluorescein read- 

 ily soluble in water. For 

 ease of carriage we made 

 this mixture up in four tins 

 of one pound each, and with 

 these set out for the water 

 sink on the stream which 

 flows from the lake on the 

 Silurian outcrop. We now 

 found that the valley below 

 the sink was quite dry, and 

 that the whole of the surplus 

 waters of the lake passed 

 underground near the point 

 where the stream crossed 

 the fault and flowed on to the limestone. 



169 



