blindly came back to the same place. On the other hand, if during the night a hive is moved to a distance much greater than that I have mentioned—say, 50 yards—the bees in the morning will start work quite freely; none will go back to the old site, but all will find their way back to their new position. This also looks as if these insects had a sense of direction, and came back to the point from which they set out on any particular journey.

On another occasion I had reason to bring a hive into my laboratory to have them under close observation. There are four similar windows close together, but of these the bottom sash of only one opens. I opened that for the bees to come in and out by, left the other three closed, and put the hive with its door facing the open window. Many of the bees on leaving the hive struck the closed windows, stayed on the glass for a time, and I never saw one in such circumstances leave the glass and drop down to the hive, 2 ft. away, or go round to the open window, 12 m. away. They all—to the number of hundreds—stayed on the glass till they fell dead or dying on to the sill. After the first day I blocked the other windows, and left them so for a fortnight, by the end of which time the bees were thoroughly used to their new quarters, and were working and breeding vigorously. It is fair to presume that by now practically all the bees in the hive were familiar with the open window, and the way in and out. I raised the blinds, and again all those that struck the glass stayed there and died, though they surely knew of the open window just beside them. In this case, too, it hardly seemed that the bees could see for any distance sufficient to warrant the supposition that they found their way about by sight, or they surely would have flown back to their hives.

Such are the facts and considerations that induced me to undertake the experiments I am about to record. At the same time, I may state that I did not believe in a sense of direction. On one occasion I moved a hive 50 yards during the night, and shut it up till I was at liberty to deal with it in the morning. I then took out fifty bees and marked them, carried them 200 yards across the garden and liberated them. They all went back to the place where the hive had been on the previous night. This in itself disproved the idea of a sense of direction, and proved that the bees found their way by sight; and numerous other experiments, and chiefly those by Romanes, clearly prove the same facts. The apparent want of sight argued in my other experiences is easily explained by the consideration that the actions of bees are prompted not by reason, but by instinct, which I think in no other case is so well characterized by its true epithet of “blind instinct.”

However, notwithstanding my conviction that there was no sense of direction in bees, Dr. Farr's theory seemed well worth testing, and I tested it by the following experiments:—

1. I saw that in some cases it would be necessary to work with drones instead of with workers, owing to the difficulty of handling the latter. So I first proved that drones would come home rapidly and unfailingly from any distance that I was likely to experiment from. I took ten drones and rolled them in flour, and carried them away from the hive to the distance of one-third of a mile. All returned within five minutes.

2. I tried the same experiment with workers, but it failed entirely. The workers settled on the grass or a tree and cleaned themselves of the flour before they came home; many also, since they were out, stayed out for their load of honey or pollen, and so by the time of their return were unrecognisable.

After various other attempts to secure an easy way of watching the bees come home I was forced to adopt the arrangement that Romanes made in his experiments and have the experimental hive indoors. My first attempt to test the magnetic suggestion was to put the bees in a box of soft iron, of sufficient thickness of wall and of sufficiently contracted content to screen off all the earth's magnetism from the bees inside the box. This proved exceedingly inconvenient, and I then determined to use a powerful bar magnet, which would so distort the lines of force due to the earth's magnetism as to render them inappreciable to the bees carried close to the magnet.

(a.) I took twenty-seven bees from the hive and carried them 100 yards away round to the back of the building. Nine bees had returned within two minutes, twenty-four within fifteen minutes, twenty-six within forty-five minutes; and one did not return, and was probably injured during capture or transit. Those that came back between two and fifteen minutes from liberation were usually laden with pollen, having seized the opportunity for commencing work.

(b.) I took twenty-six bees to the same place, carrying them on their outward journey in a test-tube lying on a bar magnet with the north pole held pointing south. Twenty-two bees had returned within fifteen minutes, and the twenty-six within thirty minutes.

(c.) I took twenty-eight bees to the same place in the same manner, but this time the north pole, on which they were held, was rotated horizontally, so as to point to all points of the compass many times during the journey. Twenty-one bees had returned within three minutes, and the twenty-eight within fifteen minutes.

(d.) I took eighteen bees to the same place in the same manner, but held them between the north and south poles of a strong horse-shoe magnet, which I rotated horizontally and vertically during the journey. Four bees had returned within two minutes, thirteen within five minutes, and the eighteen within fifteen minutes.

The bees were coming home rather better in the latter experiments: this was due probably to the increasing temperature of the day.

These experiments give a sufficiently clear proof of at least one thing: viz., that if the bees do find their way home by a magnetic sense they do not do so in the way that we should. They do not take their bearings in their outward journey and then reduce their traverse and set home along the resultant line. If all the lines of force due to the earth's magnetism are exactly the same, the experiments prove that the bees do not find their way home by magnetic sense at all. It is exactly as if a man lived in a uniform plain crossed by a great number of parallel and exactly similar tracks. When he left his home he would say, “I have crossed thirteen tracks to the north, thirty-two to the east,” and so on, and then would be able to strike straight back home; but if he were led away from home blindfolded, so that he could not count the tracks, and then given his sight, he would not be able to tell in which direction his home lay. If the tracks were similar, if he were led away blindfold, and if he found his way back, we could say he did not find his way home by sight. This is exactly what I did with the bees. I blind-folded their magnetic sense, and they found their way home; so that I can say they did not do so by magnetic sense. But let us suppose all the tracks across the uniform plain were not similar. Suppose they increased in breadth in regular succession, one being 1 in. wide, the next 2 in., the next 3 in., &c., and the cross-tracks the same. Then a man would know “My house is at the intersection of the track 3 ft. wide and the track 9 in. wide.” Now lead him away from home blindfold and then set him free: no matter where he is he will return to his home. Let us apply this case to the bees. Suppose the lines of force due to the earth's magnetism are not all exactly similar, but increase in strength in a particular direction. Then the bee, no matter where it was taken, would, when liberated, know what line of force it was on, would know in what direction lay the line of force on which its hive was, and would fly direct for that. In this supposititious case my former experiment was completely inconclusive, because I blindfolded the magnetic sense of the bees and then gave them their sight in a locality with which they were perfectly acquainted. It was to meet

this possible view of the case that I made the following experiments.

3. I here wished to disturb or annul the magnetic field in which the bees usually lived, both on their outward and homeward journey: to follow up my original illustration, to obliterate all the tracks across the plain, and make only one broad and distinct track, that would go wherever the man went, and disappear immediately behind him. To do this I attached a small powerfully magnetized needle to the back of the thorax of the bee, carried him 100 yards away from his hive, and liberated him. The attached magnet was sufficiently strong to hold a small suspended magnet at right angles to the lines of force of the earth's magnetism. The weight of the attached magnet and the adhesive used was 40 milligrammes; the average weight of the bees used was 120 milligrammes; so that the load was one-third of the bee's own weight. I used drone bees in this experiment, as they were stronger, to carry the magnet, and as workers would have injured themselves by using their stings during the fixing of the magnet. I fastened the magnets to the bees in all directions—i.e., with the north pole pointing over their head, with the south pole pointing over their head, and with the poles pointing to left and right of the bees. About one-third of the bees were unable to carry their burden, and fell into the grass when liberated. These were not counted, only those that flew into the air from my hand being included in the following numbers: Bees liberated, 12; returned home, 8. Of the four that were lost two were evidently weak on starting, and the number that returned was large considering the disadvantages they were flying under.

This experiment answers the objection I supposed, and makes it still more improbable that the bees find their way home by magnetic sense, for this sense was in effect blindfolded, both on their outward and return journeys, and still they came home. The only other objection I can conceive is this: Suppose the attached magnet were not strong enough to completely obliterate the lines of force of the earth's magnetic field, but only to distort them, then, since in the last experiment the lines were distorted consistently equally on the outward and homeward journey, the distortion might not count for anything. A little consideration of the two former experiments will prove that this objection is invalid, but with a view to refuting it experimentally I made my last test.

4. In this case I took the bees away from home in their natural magnetic field. At 100 yards away I fastened to them the magnet in all the three positions before mentioned, and then liberated them: Bees liberated, 5; returned home, 3. Thus in the third case also it appeared that the bees' power

of finding their way home was not impaired by the disturbing of their magnetic field.

These experiments seem to me to prove clearly enough that bees do not find their way back to their homes by a magnetic sense of direction. As to how they do find their way home, I agree that they have an intimate sight acquaintance with the locality in which they are working, and that they fly from one large known object to another. I have never seen the muscular sense alluded to as a source of the information by which the bee finds its way home. Yet this sense appears to me very important as assisting the sense of sight; or it might even be that it is of equal importance with sight, by which it is checked and corrected. When a bee alights on the spot from which its home has been moved, on missing its home it flies back to some large known object, from which it seems to take its bearings, and then again flies straight to the old position of the hive. I had a hive on the east side of an open window, and facing it, so that a bee had to fly west to get out of the window. I turned the hive round to the west side of the window, but still facing it. The bees leaving the hive still tried to fly westward, and struck themselves against the front of the hive, and lost the opening of the window, just as a man will habitually turn to the right or left when he leaves his house-door. As for the bees that were coming home from work, when they got to the window they were hopelessly lost, and hovered about the old position, although the new position of the lighting-board was not 18 in. from the old. In their circling flight they often struck the hive or rested on the lighting-board, but that was of no assistance to them, as they immediately flew off again to look or feel for their home. When the hive was returned to its original position they rushed in at the opening with a contented hum. These facts seem to indicate that the bees are not dependent on the sense of smell for finding the hive, and that their sense of sight is not sufficient to guide them in new situations, and that they largely depend on the muscular sense.

To settle these points is, however, beyond my present purpose, which was merely to accurately compute the value of the suggestion that bees found their way home by a magnetic sense of direction.

If any further proof were needed that the bees do find their way home by sight the following experiments would furnish it. I took a hive from my laboratory after having it closed all night, and put it 300 yards away in the garden. I opened it at 11 o'clock on a fine sunny day. The bees came out, and, apparently tempted by the fine weather, set off working without taking sufficient notice of their surround-,