University of Virginia Library

This and the following figures (which do not, of course, include the omission of several hundred accidental variants among the manuscripts) are slightly understated to allow for disagreement in a few questionable readings. I am not here primarily concerned with Garrod's errors, which, amid a wealth of valuable detail, are relatively unimportant.

The Letters of John Keats, ed. Hyder E. Rollins (1958), II, 157, 162. This edition is hereafter cited parenthetically as Letters.

Again the evidence is ambiguous, for while E and w usually show the same revised readings, there are still several instances of disagreement between them in which 1820 agrees sometimes with E, sometimes with w, occasionally with neither. The readiest explanation would seem to be that Woodhouse entered the w readings at various times, perhaps some of them even after the poem was printed.

I am indebted (as always) to Miss Mabel A. E. Steele, Curator of the Harvard Keats Collection, who tells me that the shorthand notations seem to have been written hastily: "the second word is almost certainly 'March,' and, if we accept that reading, the first is probably 'before.' It breaks down into 'b,' the first part of either 'f' or 'x,' and what ought to be 'l,' because the stroke seems to go down. If it went up from the loop, stopping at the top, it would be 'r.'"

I quote the w version as the one more likely to represent the minutiae of Keats's lost copy accurately; the E transcript (reproduced in most respects faithfully by Garrod, p. 238 n.) shows nine variants in punctuation and spelling. In w the stanza is numbered "7"; with the "corrected Copy" before him Woodhouse struck through the original fourth stanza (see Garrod, p. 237 n.) and renumbered the next three stanzas "4," "5," and "6." In E, and presumably therefore in Keats's lost fair copy, all the stanzas were unnumbered.

So Ew; Garrod's "pleasure" (2nd edn. only) is a misprint.

Again I transcribe the w text; the E version (followed except in three marks of punctuation by Garrod, p. 252 n.) shows eleven variants in punctuation and spelling, and has (as Garrod indicates) "close" for "quick" in the second line of the new stanza. In both E and w the last line quoted here ends with a comma.

"The Hoodwinking of Madeline: Scepticism in 'The Eve of St. Agnes,'" SP, LVIII (1961), 533-555; see especially pp. 544-545, 548-549.

The intense evangelicalism of Keats's otherwise amiable and worthy publishers is well illustrated in their letters written to Severn at Rome while Keats lay dying (see Hyder E. Rollins, More Letters and Poems of the Keats Circle [1955], pp. 109-118).