I often found that if bees which were brought to honey did not return at once, still they would do so a day or two afterwards. For instance, on July 11, 1874, a hot thundery day, and when the bees were much out of humour, I brought twelve bees to some honey; only one came back, and that one only twice; but on the following day several of them returned. This latter observation is important, as proving that bees can remember for at least a whole day the locality where they have found honey only once before, and that they so far think about their past experiences as to return to that locality when foraging. As the association of ideas by contiguity is the principle which forms the basis of all psychology, it is desirable to consider still more attentively this the earliest manifestation that we have of it in the memory of the Hymenoptera. That it is not exercised with exclusive reference to _locality_ is proved by the following observation of Sir John Lubbock:-- I kept a specimen of _Polistes Gallica_ for no less than nine months.[49] . . . I had no difficulty in inducing her to feed on my hand; but at first she was shy and nervous. She kept her sting in constant readiness. . . . Gradually she became quite used to me, and when I took her on my hand apparently expected to be fed. She even allowed me to stroke her without any appearance of fear, and for some months I never saw her sting. One other observation which goes to prove that other things besides locality are noted and remembered by bees may here be quoted. Sir John placed a bee in a bell jar, the closed end of which he held towards a window. The bee buzzed about at that end trying to make for the open air. He then showed her the way out of the open end of the jar, and after having thus learnt it, she was able to find the way out herself. This seems to show that the bee, like the wasp on the closed window-pane, was able to appreciate and to remember the difference between the quality of glass as resisting and air as permeable, although to her sense of vision the difference must have been very slight. In other words, the bee must have remembered that by first flying _away_ from the window, round the edge of the jar, and then _towards_ the window, she could surmount the transparent obstacle; and this implies a somewhat different act of memory from that of associating a particular object--such as honey--with a particular locality. It is noteworthy that a fly under similar circumstances did not require to be taught to find its way out of the jar, but spontaneously found its own way out. This, however, may be explained by the fact that flies do not always direct their flight towards windows, and therefore the escape of this one was probably not due to any act of intelligence.