This for the colours; then I would describe the shapes of the cloud-masses from their likeness to something else. Sometimes they would look like trees; then I thought what kind of tree they resembled, or it might be a flying figure, with a distorted hunchback rushing after it. As I followed these fancies it was wonderful what a tragic story that sunset sometimes told me before I was done with it. Once I was staying with a gentleman who added phrenology to his other accomplishments. He asked me if I never tried to write poetry, and I said, 'I had not'; to which he replied, 'Then try it, for I think you have the gift.' I sat down that night and attempted to make rhyme, but as I did not know much about the rules, and had no subject, I cudgelled my brains for words and rhymes without considering what was my theme, and therefore I failed because I had nothing definite to write about. As far as I can now remember, I think that my first attempt was a love-poem; but as I had never been in love, and had no woman to stand before me as a model, and no experience to serve me for the emotional part, it was all vague, and the result was exactly what might have been expected--meaningless words. Had I contented myself with writing about what I saw and knew, I might have made something. And this was what I learnt afterwards, after many failures: never to take up my brush or pen unless I had something definite to do--that is, never to depend altogether upon inspirations; have the object first vividly before me, and then it is not difficult to describe it, so long as one does not try to improve upon the model, or go out of the way to write or paint too finely. I discovered, after a great many failures, that nature cannot be improved upon, or even approached very near, and that the utmost my imagination could do was to put into recognisable, if faulty, shape whatever stood before my eyes, or the feelings which I myself experienced--in fact, I learnt that what we call imagination is not the gift of creating things out of chaos, but rather the remembering of emotions and scenes and real personages, and that the more vividly I could remember, the better work I did. Then I knew that Shakespeare's mighty genius lay in his vast powers of observation and in his direct simplicity of expression, and that the great charm of his characters lay in their reality, for they were people whom he had met and studied.