THE NATURE AND ORIGIN OF GILGAI COUNTRY. 345 



The coefficient of expansion of gilgai soil on wetting is 

 certainly great as illustrated by the experiment described 

 in Appendix I. 



The crabholey and melon holey country of our tablelands 

 and coastal regions is often formed in the same way as 

 4 hog wallows.' 



3. The removal of soluble ingredients by downward filtra- 

 tion cannot have produced the gilgais, for the soil is almost 

 impervious, being exceedingly clayey, and if removal by 

 downward percolation were the cause of gilgai formation 

 the hollows should contain less manganese. Drained soils 

 are usually low in manganese; gilgai soils are high in that 

 constituent. There can be little or no downward drainage 

 into subjacent water-bearing strata. 



The gilgai holes hold water for many months after rain 

 and appear to lose it only by evaporation. 



4. Great inequalities of surface are often caused in arid 

 regions by wind action on loose detrital deposits. One 

 might suppose the gilgai surface to have originated in a 

 quaternary, very arid, cycle prior to any kind of vegetation 

 getting a hold on it, by the wind scooping out the hollows. 

 At the present time the country is too well wooded for the 

 wind to have any such effect. 



But hummocks raised by the wind should have definite 

 shape and alignment. The irregularity of the gilgai surface 

 cannot be explained on this hypothesis. 



5. In swampy country of our coastal districts and table- 

 lands and in lowlying country frequently flooded, we fre- 

 quently find a tussocky grass or grasstree (Xanthorrea) 

 grass growing in tufts. The roots of each tuft keep on 

 raising the spot on which the tuft grows by their decay 

 -and intervening spaces between the tussocks are lowered 

 by nutriment being drawn away from them. The tussocks 



