On both sides of the train, at about 10 mm. distance from each other, stronger ants are to be seen, distinguishable from the others by their foxy colour and very thick heads with gigantic mandibles. These 'thickheads' play the same _rôle_ in the ant-state for which they are cast in cultured communities. They look after the order of the march, and allow none to turn either to the right or left. The least confusion in the regularity of the march makes them turn round and put things straight again. While the procession of the brown workers streams on unceasingly with a swarming motion, the 'officers,' as the natives call these thickheads, run constantly backwards and forwards, ready to take the command on meeting any difficulty. The crossing of streams by these creatures is the most interesting point. If the watercourse be narrow, the thickheads soon find trees, the branches of which meet on the bank on either side, and after a short halt the column set themselves in motion over these bridges, rearranging themselves in the narrow train with marvellous quickness on reaching the further side. But if no natural bridge be available for the passage, they travel along the bank of the river until they arrive at a flat sandy shore. Each ant now seizes a bit of dry wood, pulls it into the water, and mounts thereupon. The hinder rows push the front ones even further out, holding on to the wood with their feet and to their comrades with their jaws. In a short time the water is covered with ants, and when the raft has grown too large to be held together by the small creatures' strength, a part breaks itself off and begins the journey across, while the ants left on the bank busily pull their bits of wood into the water, and work at enlarging the ferry-boat until it again breaks. This is repeated as long as an ant remains on shore. I had often heard described this method of crossing rivers, but in the year 1859 I had the opportunity of seeing it for myself. It is remarkable that the military or driving ants of Africa exhibit precisely similar devices for the bridging of streams, namely, by forming a chain of individuals over which the others pass. By means of similar chains they also let themselves down from trees. It must be observed, however, that these and all the above observations, being independently made and separately recorded, serve to corroborate one another so strongly that we can entertain no reasonable doubt concerning the wonderful facts which they convey.