The fact is, that almost from the first I had taken a view similar to Mr. Watts’s as to the design of my book, and had determined to call the anthology by the title it now bears. On one occasion, however, I acted rather without judgment in sending Rossetti a synopsis of certain critical tests formulated by Mr. Watts in a letter of great power and value. In the letter in question Mr. Watts seemed to be setting himself to confute some extremely ill-considered remarks made in a certain quarter upon the structure of the sonnet, where (following Macaulay) the critic says that there exists no good reason for requiring that even the conventional limit as to length should be observed, and that the only use in art of the legitimate model is to “supply a poet with something to do when his invention fails.” I confess to having felt no little amazement that one so devoid of a perception of the true function of the sonnet should have been considered a proper person to introduce a great sonnet-writer; and Mr. Watts (who, however, made no mention of the writer) clearly demonstrated that the true sonnet has the foundation of its structure in a fixed metrical law, and hence, that as it is impossible (as Keats found out for himself) to improve upon the accepted form, that model--known as the Petrarchian--should, with little or no variation, be worked upon. Rossetti took fire, however, from a mistaken notion that Mr. Watts’s canons, as given in the letter in question, and merely reported by me, were much more inflexible than they really proved. Sonnets of mine _could not appear_ in any book which contained such rigid rules as to rhyme, as are contained in Watts’s letter. I neither follow them, nor agree with them as regards the English language. Every sonnet-writer should show full capability of conforming to them in many instances, but never to deviate from them in English must pinion both thought and diction, and, (mastery once proved) a series gains rather than loses by such varieties as do not lessen the only absolute aim--that of beauty. The English sonnet too much tampered with becomes a sort of bastard madrigal. Too much, invariably restricted, it degenerates into a Shibboleth. Dante’s sonnets (in reply to your question--not as part of the above point) vary in arrangement. I never for a moment thought of following in my book the rhymes of each individual sonnet. If sonnets of mine remain admissible, I should prefer printing the two _On Cassandra to The Monochord_ and _Wine of Circe_. I would not be too anxious, were I you, about anything in choice of sonnets except the brains and the music. Again he wrote: