yielded seeds.  Flowers which are self-fertile would naturally produce seeds under these circumstances; but I am greatly surprised that Delphinium consolida, as well as another species of Delphinium, and Viola tricolor, should have produced a fair supply of seeds when thus treated; but it does not appear that he compared the number of the seeds thus produced with those yielded by unmutilated flowers left to the free access of insects:  ‘Bedeutung der Nektarien’ 1833 pages 123-135.) On the other hand, in some large masses of Geranium phaeum which had escaped out of a garden, I observed the unusual fact of the flowers continuing to secrete an abundance of nectar after all the petals had fallen off; and the flowers in this state were still visited by humble-bees.  But the bees might have learnt that these flowers with all their petals lost were still worth visiting, by finding nectar in those with only one or two lost.  The colour alone of the corolla serves as an approximate guide:  thus I watched for some time humble-bees which were visiting exclusively plants of the white-flowered Spiranthes autumnalis, growing on short turf at a considerable distance apart; and these bees often flew within a few inches of several other plants with white flowers, and then without further examination passed onwards in search of the Spiranthes.  Again, many hive-bees which confined their visits to the common ling (Calluna vulgaris), repeatedly flew towards Erica tetralix, evidently attracted by the nearly similar tint of their flowers, and then instantly passed on in search of the Calluna.

That the colour of the flower is not the sole guide, is clearly shown by the six cases above given of bees which repeatedly passed in a direct line from one variety to another of the same species, although they bore very differently coloured flowers.  I observed also bees flying in a straight line from one clump of a yellow-flowered Oenothera to every other clump of the same plant in the garden, without turning an inch from their course to plants of Eschscholtzia and others with yellow flowers which lay only a foot or two on either side.  In these cases the bees knew the position of each plant in the garden perfectly well, as we may infer by the directness of their flight; so that they were guided by experience and memory.  But how did they discover at first that the above varieties with differently coloured flowers belonged to the same species?  Improbable as it may appear, they seem, at least sometimes, to recognise plants even from a distance by their general aspect, in the same manner as we should do.  On three occasions I observed humble-bees flying in a perfectly straight line from a tall larkspur (Delphinium) which was in full flower to another plant of the same species at the distance of fifteen yards which had not as yet a single flower open, and on which the buds showed only a faint tinge of blue.  Here neither odour nor the memory of former visits could have come into