Sir Henry Taylor writes to Lady Taunton, May 23, 1862 (Correspondence, edited by Edward Dowden), of Tennyson's reading: 'As to his reading, he is a very deep-mouthed hound, and the sound of it is very grand; but I rather need to know by heart what he is reading, for otherwise I find the sense to be lost in sounds from time to time; and, even when I do know what the words are, I think more of articulation is wanted to give the consonantal effects of the rhythm; for without these effects the melodious sinks into the mellifluous in any ordinary utterance; and even when intoned by such an organ as Alfred's, if the poetry be of a high order, the rhythm so sounded loses something of its musical and more of its intellectual significance. In the best verse, not every word only, but every letter, should speak. Nevertheless, his reading is very fine of its kind, and it is a very rare thing to hear fine reading of any kind.' In regard to inflections, or bends, of the voice, of every kind, direct upward, or downward, or combinations of both, which are called waves (upward waves being a combination of downward and upward inflections, or bends, and downward waves the reverse, and double waves being a combination of upward and downward waves, or the reverse), I would say, what I have said in my 'Primer of English Verse,' that a reader must have a sub-consciousness of a dead level, by which, or from which, to graduate all his departures; and it is only by avoiding all non-significant departures that he imparts to his hearers a sub-consciousness of his own standard. There should never be in reading a non-significant departure from a pure monotony. Significant vocal intervals lose their effectiveness when they are mixed up with non-significant ones. Great effects can be secured through very simple means by a reader who strictly observes this principle. Every little bend of the voice tells. But a wriggling voice, the general tenor of which is a violation of this principle, cannot secure such effects. The hearer is presented with a jumble of non-significant and would-be significant intervals, which is less effective than a pure monotony would be. Appreciative reading is shown as much, perhaps, in what I will call _time_ melody, as in almost any other feature of vocalization. A reader's sense of the relative values of successive and involved groups of thought, is largely indicated by his varied (melodious) rate of utterance. And much of the pleasure which an appreciative listener derives from reading, as reading, is this indication on the part of the reader of a nice estimate of relative values. He feels that the reader is a qualified interpreter. This estimate cannot always be determined by what a writer makes, syntactically, principal, and what subordinate, in the construction of his language. Of course, a mere variation of time is not, of itself, sufficient. There must be an appropriate variation of tone-color, etc.