H(Jl;\ KXPEDITION — NAHItATIVE. 69 



})y deep gorges, in many oi which wen^ little rock-[)0(jls, whilst IIk; valley al oiii- 

 feet stretched away westwards for twenty miles to bi'oaden out into a plain lying 

 to the noitli of tlie George Gill Range and opening into th(5 desert sandhill 

 eountiy out to the west of the main mountain ranges. 



We spent nearly three days nt Tempe Downs, collecting amongst tlie valleys 

 and on the hill sides and photograpliing amongst tin; blacks, a good number of 

 whom were camped out in the l)(;d of the river close by. Everything was as di'y 

 as usual except for a fair sized water-hole in the shelter of the Walker CJorge, in 

 which we secured a new spi'cies (.)f fish {Niiuatocentris ta/ci), which we afterwards 

 found in other water-holes a.UKjngst the ranges. Mr. Thornton U)V\ us that after 

 a heavy fall of rain small lisli are often found in the pools amongst the raxincs 

 high \\\) on the hill sides. The only possible way in which they can get into such 

 positions is by being carried there perhaps in the form of eggs attached to the 

 feet of birds, which the moment the rain comes, appear as if by magic. 



In the same water-holcjs wei'c abundant spijcimens of a mollusc, Melania 

 balonitciisis, crawling upon the lU)or, while, as usual, the sandy edge was marked 

 with the ridges left by the mole cricket {Gryllotalpa coarctata) as it burrowed its 

 way along the damp sand. We could always secure this insect at the edge of 

 water-holes with a sandy margin. 



On the rising ground between the Walk(!r and the Station Range, in a scrub 

 of gums. Acacias and Cassias, were somewhat open patches covered with tussocks 

 of porcupine grass (Tiiodia, sp.), and here I spent some time watching the habits 

 of a curious litth; black ant which had been described by Mr. Ivirby under the 

 name of IfvpocHnca J/avipes, and may be called the Porcupine grass Ant. 



Various explorers have already noted the presence of curious little galleries 

 which run along the surface of the gi'ound, often foi' long distances, from one 

 tussock of porcupine grass to another. In some parts, and especially on hard 

 sandy S(jil where the tussocks of grass are not too clos(^ togetluir, these galleries 

 as they are called, though tunnels would be a better name, foi'm a regular net- 

 work. Each is from (jne (juarter to half an inch in width and, in section, is 

 roughly semicircular in sliajie. They are made of grains of sand fastened together 

 with the resinous material ol.)tained by the ants from the gelatinous leaf sheaths 

 of the Triodia, and they form runs which lead from one tussock to another, along 

 which the ants can travel sheltered from the light and UKjre especially peihaps 

 from the heat of the sun. In many cases they lead for long distances w\i the 

 trunks of gum trees. 



