The same is true of the _Impatiens fulva_. This is an American plant, closely akin to the balsam of our gardens, which has now thoroughly established itself on the banks of some of our rivers, as the Wey, and the tributary stream that runs through Abinger and Shere. It has attractive flowers hung on the daintiest flower-stalks. It has also little green flowers that never open and almost escape attention; and yet they, and not the large flowers, are the great source of seed vessels to the plant--the great security that the life of the race will be continued.[18] Again, ugliness has borne away the palm of utility from beauty. So, too, in America the same happens with the _Specularia perfoliata_: in shady situations all its flowers are said to be cleistogamous, and to be wonderfully productive and strong.[19] The conditions of the problem in these cases are such as to make them of the last importance in our inquiry into the utility of beauty; for in each case we are comparing a conspicuous and an inconspicuous flower in the very same plant. The conditions seem to exclude the possibility of error in the result. Two explanations have been suggested of the origin of these cleistogamous flowers: according to the one, they are the earliest form of the flowers; according to the other view, they are degraded forms of the more beautiful flowers.[20] For our purpose, it is immaterial whether of the two explanations is correct; for either the development of beauty has diminished the utility of the flower, or the loss of beauty has increased the utility: in either event, utility and beauty are dissociated the one from the other. Another experiment Nature presents us with, in which the conditions are nearly, if not quite, as rigorously exclusive of error. The vast majority of orchidaceous plants are, as already mentioned, dependent on insect agency, for fertilization, and present a marvellous variety of contrivances for effecting cross-fertilization through their activity. But one of our orchids (the Bee orchis) is self-fertilized. I hardly know anything in vegetable life more striking or beautiful than to see its delicate pollinaria at a certain stage of its inflorescence descending on to the stigmatic surface and so yielding their pollen grains to the fertilization of their own blossom; and yet the Bee orchis has been found by observers to be as free a seeder as any of its tribe. Here the beauty and conspicuousness of the blossom, which are very great, are, as far as can be seen, useless; the plant gains nothing by the attractiveness which it offers, and the colouring and ornamentation of the blossom are, on the theory of utility, residual phenomena.