According to WMR, the initial forms of most of DGR's translations were written “from 1845 to 1849.” The research was done at the British Museum as well as at home, where he had access to his father's considerable library as well as his scholarly advice and assistance. He must have completed a substantial corpus by 1847, since that year he wrote his letter to Leigh Hunt requesting his opinion of the work.

It seems clear from DGR's correspondence that he did not translate Dante's until late in this initial process of composition—in fact, 1849 (see his letter to Allingham of 23 July 1854). This is a striking fact because it means that his initial work would have been with some extremely difficult materials—texts not merely untranslated but scarcely even established in an editorial sense.

DGR continued to augment and revise the translations for almost thirty years. He told Millais in 1854 that he did his translations in the evening and that he had completed “upward of fifty poems” from the work of the poets before Dante (see Fredeman, , 54. 66 ). This work principally occupied him from time to time throughout the 1850s, and especially between February 1858 and early 1861. An important letter to Mrs. Gaskell (18 July 1859: Fredeman, , 59. 27 ) reveals a great deal about the state of the project at that point in time. He enclosed a copy of the printed proofs he had set in type the previous year (see below, the Printing History commentary) and added: “The proofs contain the series of Cavalcanti's poems forming part of my book, the general title-page to which you will find at the end of them. I do not know whether you are acquainted with the of Dante—his autobiography of his youth—or with such facts as are known about Cavalcanti—but these last you have probably gathered from M. de Circourt's papers, of which I hope now, relying on your mercy, not on my tardy deservings, to earn a sight. I must tell you, on behalf of my self-conceit, that the most laborious part of what I send you is not on the surface—having consisted in the arranging and rendering as far as might be comprehensible, this set of poems which are scattered in various editions without attempt of any kind to make sense of them either in the way they are printed or in their getting together—so that much which is in fact commentary is embodied in the translations & headings, as I have tried as far as possible to dispense with the wearisome adjunct of notes. Short notices of Cavalcanti & some others among my Poets will be necessary, & these are the only portions of my work still left to do.”

DGR seems to have done little more work on the translations until the very end of 1860—the literary year having been spent attending to his original poetry and the idea of publishing a volume of his own verse. At that point he received Aurelio Saffi's comments on his translations and he was inspired to drive this fifteen year-old project to final completion. From January until Dec. of 1861 he was deeply involved in seeing his translations through the press. It was at this point that he completed the various notes and prose commentaries and had his brother translate Dante's prose “divisions” for the poems in the .

DGR made a number of revisions to the first edition just before it was published (see below “Printing History”). The single most important of these changes—one of the two last that DGR introduced—involved the cancel of pages 409-410 (signature DD5). The original text carried DGR's translation of Cecco's bawdy sonnet “in absence of Becchina”. In its place he inserted another sonnet under the same title. The alteration, probably made at the publisher's insistence, oddly anticipates the problem and controversy over that would erupt later in the context of the publication of the 1870 .

In the early 1870s DGR was again much occupied with the work as he prepared for the publication of the second (revised) edition of the translations, the volume titled (1874). There he completely reorganized the contents and general design of the book, and he made a significant number of other corrections, textual changes, and important additions (and deletions).

Besides his family, DGR sought the opinion of a number of people about his translations: principally, Charles Cayley, Leigh Hunt, Tennyson, Patmore, Ruskin, Allingham, William Bell Scott, Charles Eliot Norton, Elizabeth Gaskell, and Aurelio Saffi. Their various judgments led DGR to a long process of “revising, improving, and suppressing crudities or quaintnesses” in the work, as WMR put it ( I. 105). But certain changes made in the proofs for the 1861 edition might be judged unfortunate for they “bear the marks of a moral censor at work—perhaps Ruskin” (see Cox and Nowell-Smith's note in vol. 25 (1976), 547-548). However that may be, DGR clearly set a high value on the project and worked at it assiduously for many years.

A major stage in the revision took place between January and May 1861, when DGR wrote and/or revised much of the accompanying prose notes and commentaries (see DGR's letter to his brother of 25 January 1861: Fredeman, , 61. 6 ). It was also at this time that he asked WMR to go over the whole of the book again to “correct my translation throughout, removing inaccuracies and mannerisms” and “to translate [Dante's] analyses of the poems (which I omitted)” (letter of 18 January 1861, Fredeman, , 61.3 ).

Only a handful of manuscripts of DGR's translations survive. There is an interesting scrap carrying a revision of his translation of Cino's canzone to Dante.

Although the translations have been subjected to various criticisms, particularly in the twentieth century, this book is one of the most important and influential works of translation ever written in English. Its influence was enormous, not least of all on Pound, Eliot, and other modernists who were not always as candid, or generous, in acknowledhing DGR's pioneering work as they perhaps should have been. The extent of its influence is not reflected in the initial sales of the first (1861) edition, of which only 600 or so copies were bought in the first eight years. When DGR's volume of 1870 appeared, however, its success turned readers' attention to the book of translations, which DGR then revised for its successful 1874 reprinting. The latter, or some combination of the 1861 and the 1874 editions, has been reprinted frequently.

As early as the summer of 1851 DGR was enquiring after a possible publisher for his translations (see II. 92-93). Nothing came of the connections he tried to make at that time, but his desire to publish remained. “I am still hoping to get them out as soon as possible,” he wrote Allingham in July 1854 (see Fredeman, , 54. 55 ). Maclennan helped DGR rouse some interest with Macmillan. That prospect came up empty, however, and DGR kept circulating his translations in manuscript throughout 1854-1856. Allingham suggested periodical publication in January 1855 but DGR was unsure and hoped that Ruskin might get his publisher “Smith and Elder to shell out something for them in a lump, which arrangement, if possible, I should prefer to any other, especially as it would spur me on to a speedy completion of the book” (letter to Allingham, 23 January 1855: Fredeman, , 55. 4 .

Though still without a publisher, DGR began having the translations set in type in June 1858. The work was done at the Chiswick Press by Whittingham. At this time he opened a correspondence with Macmillan about publishing the work but again nothing materialized. He sent the newly printed proofs, which comprised a series of Cavalcanti translations, to Charles Eliot Norton (see his letter to Norton of July in 196-206). Another set apparently went to Allingham about the same time, and a year later—in July 1859—he sent a set of the same proofs to Mrs. Gaskell.

The printing process at that point hung fire until January 1861, when DGR wrote his brother that he was “pushing at last with my Italian Poets at the printer's” because he wants “to get my own poems out at the same time as the translations”. At the same time he engaged his sister Christina to prepare of fair printer's copy of the whole book (see Fredeman, , 61. 8 ). It seems likely that DGR spent some time during that intervening previous year and a half composing prose notes and commentaries, which were needed if the book was ever to make its way with the public. During this period Ruskin apparently volunteered to write “a preface or introductory essay. . .to enrich my book, & which will add incalculably to its value in every way”, as he wrote to Sir John Simeon in March 1859 (see Fredeman's essay in XI (Summer 1961), 196). The essay was not done, however, and DGR had to write all the introductory materials himself.

A culminant stage of the proof printing was finished around 10 May 1861 in a volume, as DGR told Allingham, of “nearly 500 pages” (see Fredeman, , 61. 23 ). He wanted proof copies of his book not only to distribute to friends for further comment, but to show to publishers to try to induce them to take it on. From May until the book was published (on 26 December 1861) DGR was occupied with some small number of final revisions (the latest of which appear on the errata slip in the published volume) and with a failed effort to have some graphic material included in the book—specifically, an ornamented title page that DGR designed and drew and that was engraved (not to DGR's satisfaction) by Linton (see commentary below).

The tradition descending from WMR speaks of six copies of this proof volume having been printed off for DGR. But certainly more were printed and they appear to represent two states. The located copies are as follows: 1. Fitzwilliam Museum copy, a presentation copy to William Bell Scott from DGR dated 1861 with a woodcut of DGR's illustrated title-page, not finally published with the book, tipped in; 2. The Wrenn Library copy (U. of Texas), the only known copy in the paper wrappers that WMR described for the book in his manuscript note on the flyleaf verso of this copy; 3. The Princeton/Troxell copy, an inscribed presentation copy from DGR to George Meredith (formerly in the collection of William Harris Arnold); 4. Mrs. M. C. N. Munro's copy (this copy, which remains in the Rossetti family, was given by DGR to the sculptor Alexander Munro); 5. A copy briefly described in 1976 by Charles Cox and Simon Nowell-Smith in XXV (1976), 547-548. This copy recently passed from the library of Simon Nowell-Smith to the library of the University of Virginia. It lacks signature Y but includes all of the cancelled pages. The copy in addition has bound at the end 16 pages that include all the corrected leaves as well as the errata leaf. (The copy is further described in Charles Cox Catalogue 45 no. 159.) This is the presentation copy to Harry Leigh Douglas Ward, assistant in the Dept. of MSS., British Museum. It was acquird and sold again by Ian Hodgkins & Co. Ltd. from Catalogue 86 (spring, 1996), where it is described in the catalogue.

As Charles Cox pointed out in his catalogue entry for the book, the uncancelled pages in the Nowell-Smith copy “preserve (as well as a few misprints) words, phrases, lines, and in one case an entire poem which Rossetti (or more likely his publisher) thought fit to alter or suppress” (Catalogue 45, no. 159 page 24). The poem—Cecco Angiolieri's sonnet “In Absence of Becchina”—was cancelled from page 409.

Additional proof copies can be found at the libraries of Brigham Young University, the University of Arizona (Tempe), and Yale University. Copies were sent to William Allingham, Coventry Patmore, Alexander Macmillan, and John Ruskin. A copy would have been sent to Chapman and Hall and to Smith, Elder and Co. Some of the copies in this list undoubtedly represent duplications, the same copy having been sent to more than one person. Neverthless, it seems clear that more than six copies existed, as Fredeman points out in his XI (summer 1961) article (see pp. 193-194).

The most important copy, however, DGR's own, is at present untraced. This copy was owned by Jerome Kern and was sold at the famous Kern sale in 1929 (lot 1004). It is there described as having in it “the suppressed frontispiece and title page”.

Cancels appear in all of these copies, though copy 5 above seems the only one carrying a full complement of the cancels made before the book went into its final print run. Proof copy cancels were made for B2 (pp. 3/4), B6 (pp. 11/12), E6 (pp. 59/60), F2 (pp. 67/68), H1 (pp. 97/98), H4 (pp. 103/104), DD5 (pp. 409/410). Two of these cancels were made for the first edition (B2 and DD5). To the proof volume were added, for the first edition, the prelims, the final gathering (GG, the index), and a final (unnumbered) page (the last leaf of Errata and advertisement). The errata list cites errors on pages 208, 270, 317, and 444. Some of the published copies show the misprint “252” on page 352.

DGR talked with Macmillan, Chapman & Hall, and Smith & Elder about publishing the book. Eventually the last of these three, Ruskin's publisher, undertook the job, as DGR told Alexander Gilchrist on 26 June 1861 (see Fredeman, , 61. 43 ). Ruskin advanced 100 pounds to cover the publisher's expenses, a sum that would not be repaid for eight years, when 600 copies of the book had sold. The agreement was to publish the book at 12s. in a single volume without the etchings DGR proposed. The book was finally published late in December 1861.

DGR's plan in early 1861 was to bring out two books at approximately the same time: this volume of translations as well as a volume of original poetry that was to be titled . This book, advertised on the final page of the book of translations, was never published. The death of DGR's wife halted his plan and led DGR to the notorious inhumation of his manuscript book of poetry early in 1862.

The elegant binding design was DGR's:—dark brown cloth-covered boards gold-stamped with a simple decorative rule, and with gold lettering on the spine. His original purpose, to issue the book with one or two engravings, had to be abandoned because of the cost. At least three copies exist with tipped-in examples of a copper engraving for a decorative title page designed by DGR: the Fitzwilliam copy, the Wrenn copy, and a copy of the first edition in the Tinker Collection, Beinecke Library (Yale U). (This picture is best known under the title ). T. J. Wise gives a facsimile of the engraving in his Ashley Library Catalogue description of his copy of the first edition of the book but it is marked “missing” in the annotated British Library copy of the catalogue (see Wise IV. 113-114). The original drawing for this engraving is now in the Boston Museum of Fine Arts. A zinc etching of the drawing was made later.

The Princeton/Troxell collection has a copy of the first edition of the book with six small wash drawings. These seem to have been made by DGR, though for whom he executed them is unknown. They may represent ex post facto examples of the kind of illustrations he wanted to make for the book.

DGR executed numerous pictures in various media relating to Dante, his works, and the works of the other writers that appear in this volume. The core of this body of pictorial material centers in subjects related to .

One other drawing should be noted: the comical sketch he made in 1858, apparently as a parodic title page for the book he had been trying for years to see published.

DGR gives his principal sources for the translation at the end of his Preface to the 1861 edition. The 1847 letter to Hunt shows that by that year his originals were almost excusively drawn from the two-volume edition of .

Cox and Nowell-Smith, , XXV (1976), 547-548. Fredeman, (1961), 193-198. Gitter, Hayward, McGann, , 25-38. Surtees, , 230-231. Todd, (1960), 329-331. Todd, , (1961), 447. Wise, , IV. 113-114.