The feeding of _Antennophorus_ has been closely observed by M. Janet in his artificial nests, and is thus described by him: “The ants had acquired a habit of placing themselves, crowded one against another, in one corner of the nest, and thither came such as had their crops well filled, after a meal of honey, and disgorged it before the mouths of their comrades who had none. While the fasting ant was eating the honey thus disgorged, _Antennophorus_, riding on its head, took its share. To do this, it pushed itself forward, and thrust its rostrum into the droplet, and generally, whilst holding itself in position by means of the two hinder pairs of legs, it attached itself by means of the forward pair (which in this case, however, would represent antennæ) to the head of the disgorging ant.”[13] Perhaps there is some little mistake here—possibly I have not copied the passage correctly. There has been no hint before as to the modified antenniform legs of the parasite performing any other office than that of feeling and smelling, whilst the word “attach” or “affix” is that always used to describe the working of the sticky, cup-footed ones. In the position described the antennæ might very well act as supports, but hardly, one would think, in such a way as that their owner could be described as attaching himself through their means. Possibly it is the first pair of true legs that act in this way, but the matter is of no great consequence—not more than a war, say, or the fall of a ministry, in the general run of things. Suffice it that we have our picture, the little parasite stretched, like a bridge, between the heads of the feeding and disgorging ants, and taking its share with the latter.