NOTES ON AUSTRALIAN ANTS. 
its walls continuous on all sides. It was always situated nearer the 
periphery of the stone or log than the centre, so that the outer wall of 
the cell was quite thin. In most cases a large gallery descended into 
the soil from the floor of the chamber perpendicularly or obliquely for 
a distance of about six inches and terminated in a second smaller and 
more irregular chamber. When the stone or log was turned over the 
queen lied precipitately into this second chamber, which is therefore 
used as a retreat in case of danger. It was sometimes difficult to 
capture the escaped queen because the gallery was often excavated 
between immovable stones or roots. Usually there was only one female 
under a stone, but on one occasion I found three, each in a separate 
cell and separated by nearly a foot of earth from the others. Some of 
the females were quite alone, but others had a number of eggs scattered 
on the lloor of the upper cell or a small cluster of young or nearly half- 
grown larvae. The latter were sometimes found feeding on fresh pieces 
of insects, such as caterpillars and the gasters of dealated ant females 
of the genera Camponotus and Orthocrema. This food, of course, 
must be obtained outside the nest, and since the superficial cell is closed 
off on all sides, we must assume that the regularis female does not 
remain rigidly confined like the females of the higher ants during her 
whole colony-founding period, but leaves her nest from time to time 
to secure insect food for her brood and also in all probability, nectar 
and sap for herself. This is indicated also by the following observations 
and inferences. First, I have taken a few regularis females wandering 
about in the open. They could not have been recently fecundated 
individuals because the regularis nests contain no young sexual forms 
during October and November and it was far too early for any nuptial 
flight of the species. Second, I found two incipient nests, each con- 
taining a cluster of sound and active larvae but no females. I could 
only suppose that they happened to be out foraging at the time when I 
uncovered their cells. Third, the outermost earthen wall of the cell 
in several instances looked as if it had been broken open and restored 
repeatedly. In the case of Mynnecia analis mentioned on p. 44 I 
actually found an opening in this wall ! And fourth, the cells inhabited 
by the females and larvae were always exquisitely clean, indicating 
that the former must carry all refuse insect food to the outside, as 
Clark has described (see p. 22) for the adult colonies. I believe, therefore, 
that the regularis female makes her cells soon after her nuptial flight 
and then leads the life of a recluse till October or November, occasionally 
breaking through the outer wall and foraging for food. With the return 
of spring in October the more abundant food-supply enables her to lay 
a number of eggs and to rear a few larvae with insect food which she 
captures on similar excursions. Additional evidence of this behaviour is 
given in connection with several of the following species of Myrmecia.” 
In the above account some peculiar statements are not 
quite explained. First, the latter half of the first paragraph 
on p. 26 would make it appear that I had stated that the queen 
fasts for a period of seven or eight months while maintaining 
herself on her body fat. I stated definitely that the larvae are 
fed on insects, supplied by the queen alone, and that it was six 
to seven months before the first workers left the cocoons. In 
this statement Wheeler absolutely supports my observations by 
the following statement on page 39 : — 
[ 6 ] 
