The earliest texts of this poem are the draft and the fair copy manuscripts in the Fitzwilliam Museum, both made in October 1869. These texts led directly into the first printed text in the Exhumation Proofs, which were printed around 30 October 1869. In this state the poem has already been completely recast from its 1847 form. We know that the poem was part of the manuscript book that had been buried with DGR's wife in 1862 and exhumed in early October 1869, and that DGR received the exhumed volume on 20 October. Since DGR was reading the recovered poems to his family and friends by 26 October, and since they were printed out by 30 October, it seems likely that was already in some recast form when DGR received it on 20 October. The surviving manuscripts support this view, but not conclusively. It is possible that the poem he had on 20 October was the 1847 precursor text— i.e., the poem we now know as “On Mary's Portrait Which I Painted Six Years Ago”, and that DGR rewrote it completely in late October for the Exhumation Proofs.

DGR revised the poem in local ways on the Exhumation Proofs in early November, and when the text was set in type for the Second Trial Book (26 November) he made major revisions, including the addition of five new stanzas (received stanzas 2, 6-9). The Ashmolean MS, a corrected copy of stanzas 6-9, represents DGR's further revisions, which he completed when he was correcting proof for the Second Trial Book.

See Commentary (Reception History) for the 1870 .

First printed in the Exhumation Proofs (30 October 1869); reprinted and revised in the Second Trial Book (26 November 1869), and first published in the 1870 , and collected thereafter.

Besides Browning's invoked at the outset, the poem is much in debt to Poe's many representations of beautiful dead women. This influence becomes apparent when the poem is read in relation to DGR's illustrations for (see e.g., ). The network of subtle relationships is also usefully explicated when we remember that this poem traces its origins back to the Poe-influenced story . The image in lines 100-101—particularly in the Fitzwilliam draft manuscript version—directly recalls DGR's illustrations for Poe's (see the Victoria and Albert drawing in particular).

Like the sonnet of the same title, this poem recollects Fazio Degli Uberti's great , which DGR translated. Uberti's poem, once regularly attributed to Dante, was a seminal work for DGR. And of course Dante's vision of Beatrice is a foundational presence in this poem.

Several other of DGR's translations from the early Italian poets are also related to this poem—for example Jacopo da Lentino's and his ; and Giacomino Pugliesi's .

Since DGR always regarded as a kind of memorial to his dead wife (she was the model for the painting), and since—contrary to what was believed by many earlier scholars and students—DGR had in fact done considerable work on that painting before his wife's death, the autobiographical features of this poem get underscored because of its strong relationship with the painting. The very fact that this crucial painting had not been completed before Elizabeth's death invested it for DGR with a grave import.

We should also remember that this was one of the poems buried with his wife. That fact emphasizes the poem's (as it were) subsurface biographical aspects. Of course the poem carefully disguises these matters, and forces only an oblique awareness of them; but the same obliquity is pursued throughout , where the autobiographical aspects of the work are never in doubt.

Baum, 26-33. Boos, : 215-226 Gregory, II. 135-137 Howard, , 12-19 Keane, , 204 Masefield, , 50-51 Rees, , 25-33 Riede, , 247-249 Riede, , 111-113 Surtees, I. 93-97