518 E. H. DONKIN : I wish to say distinctly that I offer this analysis as a mere attempt to move in the right direction, though the goal may be distant. If my formulas were applied concretely to the eyes, nose, mouth and chin of some particular expressive face, they might seem like a pick- axe offered for use in watch- making. But they may none the less be a rough outline of the truth. The timbre of certain musical instruments and of certain voices seems to be a case curiously illustrative of this whole subject. Compare the timbre of the oboe and the flute. The latter we should probably call full and round ; the oboe's quality might be described as harsh, or as pinched and strained. Yet the aesthetic rank of the oboe in the orchestra would be generally considered as higher than that of the flute. The flute is sweet but common-place ; the oboe means pathos and depth. And to my mind there is an inkling of similarity between this case and that of the ex- pressive but irregular face, or the interesting but incomplete view, or the line that is the better for being too short. The oboe's timbre may perhaps be said to sound as though some of the tone had been withdrawn : as though the cord had had one of its strands taken away through all its length. (Can physicists say whether there is any foundation in fact for this impression ?) Thus the oboe has a pathetic suggestive- ness ; it is expressive. But is this to be accounted for on the same theory as that suggested for the expressive face ? Different as the instances seem, they are fundamentally similar. And here another very beautiful class of aesthetic material may be referred to as illustrating both. The oboe's timbre may be grouped together with the short fragments of lost ancient poetry. Have not the fragments of Sappho or Ennius a distinct aesthetic charm through their very fragmentariness? Or when Tennyson's " Morte d' Arthur" is presented as a frag- ment, has it not a glamour which it loses in the completed idyll ? And the reason seems to be that we are willing to credit the fragment with the splendour that would have belonged to the completed poem, and that imagination en- dows this non-existent poem with ideally supreme splendour: thus in our former phrase, the quality of ideally supreme splendour persists onwards from the one factor, the imagined poem, to the other, the fragment, in spite of the antagonistic difference, viz., fragmentariness, which so cruelly hampers the forlorn fragment : just as the quality of nine-syllabled- ness persists onwards into the five-syllabled line. And
