88 LIFE HISTORIES OF AUSTRALIAN ODONATA, iv., 



with a long-bladed draining-spade, and two large, sharp 

 knives for cutting the sedge. The swamp was unaltered in 

 appearance, and I soon found a large clump (marked A in 

 the diagram) near the spot which I had examined a year 

 previously. On cutting away the sedge level with the clump, 

 we fovmd three distinct, round holes. These were the 

 entrances to three nearly vertical channels, of which two 

 (a J and a^) are shown in the diagram. By slicing away the 

 matted roots surrounding them, we cut down to the water- 

 level, about six inches, and then followed the channels down 

 another foot or more, where they turned into a wider and 

 looser horizontal channel. The latter, however, was soon 

 lost, as it collapsed on examination. The outer channel, a^, 

 being in firmer mud, proceeded downwards another six 

 inches or more, and then turned into a horizontal channel, 

 loose, and easily destroyed like the other. 



We next began to dig out the clump A at the side border- 

 ing the small depression or water-hole C. On cutting out a 

 slab, the whole depth of the hole, we discovered a well-made 

 channel (a 3) opening out at ttie side of the clump A, heloiv 

 the water-level. This was not quite vertical, and on follow- 

 ing it down, about a foot, it turned into a softer, horizontal 

 channel, which I took to be the same one that connected a, 

 and ag. I then followed this horizontal channel under C, 

 where the mud was very loose and slimy. As long as there 

 wei'e roots enough to hold together at all, there was a dis- 

 tinct passage ; but it collapsed almost on touching. How- 

 ever, it seemed to be a distinct connection, of a very flimsy 

 kind, leading from the clump A, under C, to the clump B, 

 and was about the right width for the easy passage of 

 Fetalura larvae. Probably, in the softer ooze, the larvae just 

 push their way along, and leave an unstable passage-way 

 behind them, which collapses whenever a rain-storm increases 

 the pressure from above. 



So far (after working nearly an hour) we had found no 

 larvae. But on dredging out the soft mud along the hori- 



