Dr. Malcolm Madaren — Desert-icater in W. Australia. 303 



abundant supply of water at night. They are regai'ded as mysterious 

 and indeed somewhat uncanny, since their nocturnal flow is always 

 preceded by Aveird hissings and by hollow sounds of rushing air. 

 Certainly no explanation of the undoubted fact that water appears 

 in them at night only, during the summer season at any rate, has, 

 so far as I know, yet been advanced. How many ' night wells ' 

 there are I have never been able to ascertain. The cattlemen speak 

 of several, but are hazy as to their geographical position. In 

 March, 1911, I had an opportunity, while driving across country 

 from Ravensthorpe to the railway station at Broomehill, in the 

 south-western portion of the State, of visiting for a few minutes the 

 only night well whose position I could definitely fix. It lies between 

 Magitup and Cowallellup (Carollillup) in lat. 34° 5' and long. 

 118° 25'. 



The ' Avell', or rather rock-hole, is situated on the top of a rounded 

 boss of platy granite-gneiss, standing from 6 to 8 feet above the 

 bed of a stream of the usual West Australian type, the course of 

 which is marked across the plain by a double line of large she-oak 

 and paper-bark trees. Along the sandy creek bed were several pools 

 of salt water, from which the salt was beginning to crystallize out 

 on the margins, for it was then long since rain' had fallen. I had 



Fig. 2. Cross-section of 'Night Well'. A, trough; B, internal cavity; 

 C, C, vertical joints. 



expected to find in the night well an ordinary gnamma hole, but 

 it differs materially from these. Across the boss there ran two 

 vertical joint-planes, 2 feet apart at the widest portion exposed, 

 but almost meeting at the other end of the bare rock surface, which 

 was some 60 by 30 feet in area. For a length of 15 feet the rock 

 between these joint-planes has been removed, yielding a trough 

 with a width of from 18 to 30 inches and a maximum depth of 

 2|- feet. Tln^oughout the summer this trough is empty by day, 

 but contains from 12 to 18 inches of water at night. JS'ear the 

 bottom of the trough, and in the rock wall of one side, was a 

 horizontal rift extending the whole length of the trough, and marking 

 the plane of separation of a thin surface plate of gneiss from the 

 main mass below, and it was through this rift that the water appeared 

 and disappeared. The platy granite-gneiss here found is by no 

 means common in Western Australia. This, indeed, is the only spot at 

 which I saw it during traverses covering many hundreds of miles. It 

 is, however, recorded by the Geological Survey from near K^orthampton, 

 and also in the Gascoyne and the Pilbara region in the north-west 

 of the State. AVhile a portion of the watei" contained in the cavity 

 that clearly exists beneath the surface-shell of gneiss is probably 



