in the appreciation of sounds explains, indeed, the failure of attempts to modernise early poetry, such as those of Dryden and Pope. In such cases we expect the old harmony between earlier sounds and ideas to be kept up by the moderniser, whose ideas and sounds are both more or less different, and consequently the harmony into which he transforms the old verse. Our expectation is, of course, disappointed; it overlooks at once the subtle progress we have observed, and the peculiar fitness of certain sounds for certain ideas—a fitness which the poet of any age, just in proportion as he is a poet, is sure to detect and to express for him who has the ears to hear.
If there ever lived a poet who was likely to clearly express these very subtle relations of sound, speech, and thought, and their effects on translation, that poet was Shelley; and, though it often happens that a man who himself knows how to produce an effect has not reflected upon his powers so as to rationally explain their operation, we may see from the following quotation that Shelley was not unconscious of the process by which his own exquisite harmonies of word and thought were produced, and the impossibility of transferring them from one language to another, which must needs be a different sound-instrument. "Sounds, as well as thoughts," says Shelley, in his Defence of Poetry, "have relations both between each other and towards that which they represent; and a perception of the order of these relations has always been found connected with a perception of the order of the relations of thought. Hence, the language of poets has ever affected a sort of uniform and harmonious recurrence of sound, without which it were not poetry, and which is scarcely less indispensable to the communication of its influence than the words them-
