I am back at twelve o'clock. None of the strays is at the nest; but, a few minutes later, I catch two. At two o'clock, the number has increased to nine. But now the sky clouds over, the wind freshens and the storm is approaching. We can no longer rely on any further arrivals. Total: nine out of forty, or twenty-two per cent.

The proportion is smaller than in the former cases, when it varied between thirty and forty per cent. Must we attribute this result to the difficulties to be overcome? Can the Mason-bees have lost their way in the maze of the forest? It is safer not to give an opinion: other causes intervened which may have decreased the number of those who returned. I marked the insects at the starting-place; I handled them; and I am not prepared to say that they were all in the best of condition on leaving my stung and smarting fingers. Besides, the sky has become overcast, a storm is imminent. In the month of May, so variable, so fickle, in my part of the world, we can hardly ever count on a whole day of fine weather. A splendid morning is swiftly followed by a fitful afternoon; and my experiments with Mason-bees have often suffered by these variations. All things considered, I am inclined to think that the homeward journey across the forest and the mountain is effected just as readily as across the corn-fields and the plain.

I have one last resource left whereby to try and put my Bees out of their latitude. I will first take them to a great distance; then, describing a wide curve, I will return by another road and release my captives when I am near enough to the village, say, about two miles. A conveyance is necessary, this time. My collaborator of the day in the woods offers me the use of his gig. The two of us set off, with fifteen Mason-bees, along the road to Orange, until we come to the viaduct. Here, on the right, is the straight ribbon of the old Roman road, the Via Domitia. We take it, driving north towards the Uchaux Mountains, the classic home of superb Turonian fossils. We next turn back towards Serignan, by the Piolenc Road. A halt is made by the stretch of country known as Font-Claire, the distance from which to the village is about one mile and five furlongs. The reader can easily follow my route on the ordnance-survey map; and he will see that the loop described measures not far short of five miles and a half.

At the same time, Favier came and joined me at Font-Claire, by the direct road, the one that runs through Piolenc. He brought with him fifteen Mason-bees, intended for purposes of comparison with mine. I am therefore in possession of two sets of insects. Fifteen, marked in pink, have taken the five-mile bend; fifteen, marked in blue, have come by the straight road, the shortest road for returning to the nest. The weather is warm, exceedingly bright and very calm; I could not hope for a better day for my experiment. The insects are given their freedom at mid-day.

At five o'clock, the arrivals number seven of the pink Mason-bees, whom I thought that I had bewildered by a long and circuitous drive, and six of the blue Mason-bees, who came to Font-Claire by the direct route. The two proportions, forty-six and forty per cent., are almost equal; and the slight excess in favour of the insects that went the roundabout way is evidently an accidental result which we need not take into consideration. The bend described cannot have helped them to find their way home; but it has also certainly not hampered them.

There is no need of further proof. The intricate movements of a rotation such as I have described; the obstacle of hills and woods; the pitfalls of a road which moves on, moves back and returns after making a wide circuit: none of these is able to disconcert the Chalicodomae or prevent them from going back to the nest.

I had written to Charles Darwin telling him of my first, negative results, those obtained by swinging the Bees in a box. He expected a success and was much surprised at the failure. Had he had time to experiment with his Pigeons, they would have behaved just like my Bees; the preliminary twirling would not have affected them. The problem called for another method; and what he proposed was this:

'To place the insect within an induction coil, so as to disturb any magnetic or diamagnetic sensibility which it seems just possible that they may possess.'

To treat an insect as you would a magnetic needle and to subject it to the current from an induction coil in order to disturb its magnetism or diamagnetism appeared to me, I must confess, a curious notion, worthy of an imagination in the last ditch. I have but little confidence in our physics, when they pretend to explain life; nevertheless, my respect for the great man would have made me resort to the induction-coils, if I had possessed the necessary apparatus. But my village boasts no scientific resources: if I want an electric spark, I am reduced to rubbing a sheet of paper on my knees. My physics cupboard contains a magnet; and that is about all. When this penury was realised, another method was suggested, simpler than the first and more certain in its results, as Darwin himself considered:

'To make a very thin needle into a magnet; then breaking it into very short pieces, which would still be magnetic, and fastening one of these pieces with some cement on the thorax of the insects to be experimented on. I believe that such a little magnet, from its close proximity to the nervous system of the insect, would affect it more than would the terrestrial currents.'