ARCHITECTURE OF ANTS. 58 
always in the interior of trees, and de- 
sirous of being screened from observation, 
precludes every hope, on our part, of fol- 
lowing them in their several occupations. 
I tried every expedient I could devise to 
surmount this difficulty; I endeavoured 
to accustom these ants to live and work 
under my inspection, but all my efforts 
were unsuccessful ; they even abandoned 
the most considerable portion of their 
nest to seek some new asylum, and spurn- 
ed the honey and sugar I offered them for 
nourishment. I was now, of necessity, 
limited to the inspection alone of these 
edifices; but by decomposing some of 
the fragments with care, I hoped to ac- 
quire some knowledge of their organiza- 
tion. Here, we perceive horizontal gal- 
leries, hidden in great part by their walls, 
which follow the circular direction of 
the ligneous layers; there, parallel gal- 
Jeries, separated by extremely thin 
partitions, having no communication 
except by a few oval apertures. Such 
is the nature of these works, remark- 
able for their delicacy and lightness. 
» $s 
