AMERICAN HYMENOPTERA, 2G1 



or doors arc habitually left in the work, through which the ants 

 are niovintj: back and forth, apparently working upon the inside to 

 strengthen the arch. As sections of the building are completed these 

 doors are closed, so that they are plainly but temporary arrangements 

 for the convenience of the masons. 



On other parts of the foundation similar structures were going up. 

 At 2, Fig. 1 was a section of a vertical column, one side of which had 

 been cut away. It was two inches high, one inch across. The. ants 

 were working upon this in the same manner as described above. They 

 built not only from the bottom up, but from the side across. The 

 central opening in the figure was finally closed, leaving, when the 

 work ended, the opening at the foot of the column. The circular 

 gallery thus enclosed was one-half inch in diameter, which is about 

 the usual dimensions. The work of construction was not confined to 

 the space which, as in the above cases, was the original site of the 

 cone. Having occasion to lift up a fragment half the size of one's 

 head, which had been thrown to one side, I saw that the section had 

 already been made the nucleus of a new mound. Columns, corridors 

 and halls, corresponding closely with those outlined upon the under 

 face of the fragment, had been erected, which were thus quite united 

 to the fragment. In one of the halls was a small collection of dead 

 ants. The greater portion of one day was spent in studying and 

 recording the work upon this one hill. Other drawings were made 

 from different positions, but the method and result were the same. 

 As the activity occasioned by the shower continued for the remainder 

 of our stay, I had full opportunity by subsequent observations to 

 verify my notes. 



Fig. 2 is another example of architecture drawn from the same 

 broken hill. The figure represents a double gallery which was built 

 up against the perpendicular side H of a hole cut by the spade in 

 removing the cone. The gallery a, a, a, was carried along the base 

 of the side 3 inches, and then upwards toward the surface. 'I he 

 gallery was widened at two points c, c, to l] inches, as though in- 

 tended to serve as store-rooms for cocoons. Galleries opening down- 

 ward communicated with these enlargements. At e ants were arrang- 

 ing pellets along a projection on the side, for what purpose was not 

 apparent. 



My attention was next directed to the large hill PI. II, which with 

 its surrounding hill-cluster was on my regular " list." I took this 

 plan of keeping several hills under regular, daily, and indeed for 



TRANS. AMEK. ENT. SOC. VI. (35) NOVEMBEU 1877. 



