190 MisceUaneous. 



rounding woods. These roads or trails were distinctly marked upon 

 the surface of the ground, having in places a width of from two to 

 four inches which was stained a dark brown or black, probably by 

 the formic acid exuded from the insects ; the leaves and grass upon 

 which the trail was made were pressed down and smoothed by the 

 constant action of innumerable legs upon the surface. So well 

 marked were the trails that even without the presence of the columns 

 of insects that thronged back and forth upon them they were dis- 

 tinctly and easily traced. While following up one of these roads 

 the observer was impressed by the fact that it showed scarcely any 

 deviation from a straight line. In order to test this matter more care- 

 fully, he selected a large mound from which three roads radiated. 

 These were all traced to their termination at three several oak-trees, 

 up which the columns of ants ascended in search of food-supply 

 from numerous aphides which infested the branches of the trees. 

 The ant-roads were then carefully marked out by stakes stationed 

 at short intervals, a course which was made necessary by the fact 

 that they were carried for considerable distances beneath the tall 

 bracken, which had to be pushed aside in order to reveal them. 

 The result of his observations is as foUows : — 



Road no. 1 was twenty-one paces in length (about 65 feet) and 

 was carried in an almost perfectly straight line from the nest to the 

 terminal tree. No. 2 was twenty-three paces in length (about 70 

 feet). It varied less than 3 inches from a direct line measuring 

 from the nest to a point within 2 feet of the terminal tree. There 

 the column made a detour of about 6 inches from the straight line ; 

 but an abandoned path, continuous with the main road, which had 

 apparently been used at a recent date, was traced for a considerable 

 distance further without any deflection. No. 3 was the longest 

 road of the three, being thirty-four paces in length. It extended 

 for six paces in a straight line from the nest, at which point it 

 touched an old stump, which evidently deflected the path at a slight 

 angle. From this point it was again continued in a nearly straight 

 line as far as the beaten foot-path through the wood. Here the ant- 

 trail was obliterated by the friction of passing human feet, but the 

 ants themselves thronged over the pathway in a column much 

 broadened by continual interference and loss caused by foot pas- 

 sengers. The trail was, however, resumed at a point nearly oppo- 

 site that at which it touched the path, and was continued again in 

 a straight line six paces further to the tree, where it terminated. 

 When the entire trail was staked off it was found that its terminus 

 deviated less than 3 feet from a straight line drawn from the point 

 of departure at the ant-hill. The greater deviation in this case 

 seemed evidently to have been caused by the peculiar difficulties in 

 the chosen track. The three roads so radiated from the nest that 

 they were included within about one quadrant of a circle, of which 

 the two shorter trails might represent the radial boundaries of the 

 quadrant, while the longer trail was drawn nearly midway between 

 the two. 



Takinor the results of the three observations together it is mani- 



