THE ENTOMOLOGIST. 95 
by always causing the marked ant to cross from the strip of 
paper (H) to the larve over a particular bridge of paper (k), 
and if, whenever a stranger came, the paper bridges (K and L) 
were reversed, it would be shown whether the other ants who 
came to the larve had bad the direction and _ position 
explained to them. In such a case they would go right, 
notwithstanding the interchange of the paper bridges; but if 
they found their way by tracking the footsteps of the first ant, 
they would pass over the paper bridge (Kk), and thus be led 
away from the larve to the strip of paper (1). The result was 
that out of seventy-nine strange ants, which came up to the 
point at which the paper bridges diverged, twenty-four went 
straight along the strip of paper, eleven took the right bridge 
to the larve, while forty-four were misled, and went over the 
paper bridge (k) away from the larve to the strip of paper (1). 
He then slightly altered the arrangement, transfixing one end 
of the two paper bridges by a pin, and so fastening them by 
one end to the strip of paper (H), the other ends free, that 
each of them could be turned either to the larve or to an 
empty glass tray. When the marked ant came he turned one 
paper bridge (k) to the larva, the other (L) to the empty tray ; 
while whenever any other ant came he turned the bridges, so 
- that K led to the empty tray, and L to the larvee. Under these 
circumstances, seventeen ants which came along the strip of 
paper (H), without a single exception, went over the bridge 
(K) to the empty tray. He then varied the experiment by 
leaving the paper bridge (K) loose, as at first; but instead of 
having a separate bridge (L) he cut the strip of paper (H) into 
two pieces (#’ and H”); then, when a strange ant was coming, 
he rubbed his finger two or three times over the bridge (k), so 
as to remove—or, at least, confuse—the scent. As soon as the 
ant had passed over the first part (H’) of the strip of paper (u), 
and had arrived on the part (”), he took up the piece (n’) 
and placed it where the paper bridge (L) had been in the 
previous experiments, 7.e., so as to connect the end of H with 
the empty glass tray. By this arrangement the bridge K was 
left in its place, and, on the other hand, there was a bridge 
which the marked ant had crossed and re-crossed as often 
as K, but which led away from the larve. Under these 
circumstances, out of forty-one ants which found their way 
to the end of the strip (H), and within two inches of the 
larvee, fourteen only passed over the bridge (K) to the larvae, 
