University of Virginia Library

Some Textual Problems in Yeats
by
Russell K. Alspach [*]

THIS PAPER IS BASED ON THE VARIORUM EDITION of Yeats's poems that Peter Allt, late professor of Anglo-Irish literature at the University of Groningen, and I have been preparing jointly for some ten years. Because of Peter Allt's untimely death in an accident in London two years ago the finishing of the work fell to me. His death was a severe personal blow, and it was a great loss to Yeats' scholarship.

In our work, we confined ourselves to versions of Yeats's poems published by him in magazine or book. We included no manuscript versions—they are quite numerous and widely scattered, and comparatively few are available for study. (And no manuscript version has the final polishing of proofreading by Yeats.) Nor did we include versions of his poems quoted by other writers, e.g. Katharine Tynan[1] and Dorothy Wellesley.[2] At best these are secondhand. As our standard text we used the two-volume limited and signed edition of The Poems of W. B. Yeats published by Macmillan of London in 1949. The publisher's brochure that accompanied this edition reads in part:

With this definitive edition at hand it is possible to trace the evolution of a Yeats poem from its first published version through all its revisions to its final form and to see how in the process Yeats's concept of what he wanted the poem to say developed.

Under revisions I include all changes, even those in punctuation. Perhaps this is dangerous ground; much has been said about Yeats's carelessness in, and lack of knowledge of, punctuation. But the testimony varies. We know, for instance, that he wrote to Robert Bridges in 1915 that 'I chiefly remember that you asked me about stops and commas. Do what you will. I do not understand stops. I write so completely for the ear that I feel helpless when I have to measure pauses by stops and commas.'[3] But Mrs. Yeats told me in the summer of 1954 that she believed shifts in punctuation were very much a problem in any textual study of her husband's work, that W. B. was careful about 'stops and commas,' and that several times he had become quite irate with a publisher who had taken it upon himself to change the poet's punctuation. It took me a long time to convince my colleague on the variorum that punctuation must be noted. His point of view originally, and the point of view of others, was that it should be noted only where the meaning or rhythm is affected. But who is to judge where the meaning or rhythm is affected? Certainly not the textual critic or scholar who is working in the realm of facts and not of aesthetics. And in the case of Yeats the conflicting testimony about his knowledge of punctuation must be kept in mind. I believe further that a close study of the punctuational changes in his verse will convince others, as it has me, that these changes were in the main deliberate and not accidental.

That Yeats was an inveterate revisor soon became apparent to discriminating readers of his poetry during the years in which it was being published. He continued to revise until the end; he was never content. He comments often about his revisions. The best-known of these comments is the quatrain he wrote for the second volume of The Collected Works of 1908:

Not so wellknown, but pertinent to this study, is his statement in the preface written in 1927 for the thirteenth reprinting and revision of Poems (1895): 'This volume contains what is, I hope, the final text of the poems of my youth; and yet it may not be, . . . . One is always cutting out the dead wood.'[5] We can, I think, profitably examine the last words of this statement, 'One is always cutting out the dead wood,' and the implications that arise from it.

In any textual study of Yeats's poems no edition, printing, impression, or reissue can be ignored. For example, the volume Later Poems was published in London in 1922 and reprinted there the same year. These printings are identical. In February 1924 it was reprinted again; this time there are a number of changes. In April 1924 it was published in New York, and this printing differed from the London printings of 1922 and February 1924. In 1926 and 1931 there were further reprintings in London: these printings differ from all previous printings and from each other. (A New York reprint of 1928 is identical to the first New York printing, that of 1924.) A second example is the trade edition of The Wild Swans at Coole, published in London and New York in 1919. These are identical. But a London reprint of 1920 has several changes. It is here, for instance, that we have for the first time the final form of 'An Irish Airman Foresees His Death.' A third example is the Selected Poems of 1929, published only in London. (Parenthetically, this is usually ignored by Yeats's textual critics, but textually it is very important. Many of the changes ascribed to Collected Poems (1933)[6] actually appeared first here.) Selected Poems was reprinted in 1930, 1932, 1936, 1938, 1951, and 1952—the last two in the Golden Treasury series. The printings of 1929, 1930, and 1932 are identical; but there are changes for the 1936 printing. From then on the printings are the same.

Nor can it be assumed that simultaneous periodical publication in Britain and America meant identity of form. In her admirable study of Yeats's revisions of his later poems, Professor Witt[7] has discussed a case in point: that of 'Among Schoolchildren' published first in both The Dial and The London Mercury for August 1927. Other examples are numerous. 'The Two Kings' appeared in Poetry and The British Review in October 1913. In the 252 lines of these first printings there are over fifty changes, great and small, between the two. In 'Upon a Dying Lady,' in The Little Review, August 1917, and The New Statesman, 11 August 1917, there are some twenty changes in the seventy-three lines. And between the versions of 'The Second Coming' in The

Dial, November 1920, and the London Nation, 6 November 1920, there is a startling change. Line 19 reads in The Dial, 'That thirty centuries of stony sleep,' and in The Nation' 'That twenty centuries of stony sleep.' Even with Yeats's handwriting it's a little difficult to make 'thirty' out of 'twenty.'

So far in these examples I have been pointing up the truth of Yeats's assertion about 'always cutting out the dead wood,' and assertion that would have lead one to believe that if he had the latest-dated Yeats volume he had the latest versions of its contents. But there are curious exceptions, exceptions where 'dead wood' persisted beside 'live wood.' One of these is The Wind Among the Reeds, published in London and New York in 1899, and reprinted in both places a number of times. The last printing was London, 1911, called the 'sixth edition.'[8] All printings are identical. But in the meantime The Wind Among the Reeds was revised for the first volume of the 1906 Poetical Works published only in America,[9] and again revised for the first volume of the 1908 Collected Works published only in England. In 1908, therefore, we have in print three versions of the poems in The Wind Among the Reeds: the original version, the version in The Poetical Works I (1906), and the version in The Collected Works I (1908). To add to the mixup, The Poetical Works I (1906) was reprinted nine times, with no changes, to 1922. The same anomaly happens with the 1931 printing of Later Poems that, while it has revisions, does not have all the revisions made for the 1929 Selected Poems. And the 1936 revised printing of Selected Poems does not have all the revisions of the 1933 Collected Poems. An accurate study of the development of a Yeats poem demands, it seems to me, the ability to work one's way through this maze in order to recognize the 'dead wood' and the 'live wood.'

Closely allied to the 'dead wood' and 'live wood' textual problem is the problem of the 'early' Yeats and the 'later' Yeats. Any one even faintly familiar with Yeats is aware of this much-discussed division of his work. The difficulty is the wide variance of opinion about where 'early' stops and 'later' begins. One critic dates the 'early' Yeats from 1889 to 1901, another from 1889 to 1910, and another from 1889 to 1914. Perhaps we cannot do better than take the poet's own dates in definition of 'early' and 'later.'

He first uses the words 'Early Poems' in 1906 as a subheading in The Poetical Works I. Under 'Early Poems' he includes the long narrative

poem 'The Wanderings of Oisin' and certain of the ballads and lyrics from his first two published volumes: The Wanderings of Oisin and Other Poems (London, 1889) and The Countess Kathleen and Various Legends and Lyrics (London; Boston and London, 1892). He uses the same subheading with almost the same inclusions in 1908 in The Collected Works I. In 1913 in the Tauchnitz edition of A Selection from the Poetry of W.B. Yeats (Leipzig) he modifies the subheading to 'Early Poems (1885-1892)' and includes under it thirteen lyrics from the 1889 and 1892 volumes; later in the same year in A Selection from the Love Poetry of William Butler Yeats (The Cuala Press, Dundrum, Ireland) the subheading is again modified, this time to 'Early Poems 1890-1892': the contents are four lyrics from the 1892 volume. In 1921 in Selected Poems (New York) the heading and the number of poems are the same as in the Tauchnitz edition but with some variation in the selection. In 1925, in the title of the volume Early Poems and Stories (London; New York), he uses the words for the last time; there is a slight shift in the inclusions but they are practically the same as in The Poetical Works I (1906). From here to the definitive edition of 1949 these early poems with the exception of 'The Wanderings of Oisin' are called 'Crossways' and 'The Rose,' descriptive titles Yeats had originally used when he took the shorter poems from the volumes of 1889 and 1892 and published them in Poems (1895). Apparently, then, Yeats thought of his early poetry as that written up to 1892. For he could, in 1906, have included The Wind Among the Reeds (1899) and In the Seven Woods (The Dun Emer Press, Dundrum, Ireland, 1903) among his 'early' poems. And although by 1925 critical opinion was already dividing his work at this date or that date, Yeats stuck to his own division.

The term 'later poems' he uses, and by using defines, just once: as the title of the 1922 Later Poems, that includes verse from the following volumes: The Wind Among the Reeds (1899), In the Seven Woods (1903), Poems 1899-1905 (London and Dublin, 1906), The Green Helmet and Other Poems (New York and London, 1912), Responsibilities (London; New York, 1916), The Wild Swans at Coole (London; New York, 1919), and Michael Robartes and the Dancer (Cuala Press, Dundrum, 1921). The line of demarcation seems clear.

But having decided the dates of 'early,' the textual critic must next determine just what an early Yeats poem is: i.e., is it an early poem through all its revisions, or is it an early poem through only some of its revisions, or is it an early poem in only its first printing? Of the forty early poems, including 'The Wanderings of Oisin,' in the definitive edition there is not one that was not revised in greater or less degree.

Eleven got their final polishing for the definitive edition of 1949, seventeen for Collected Poems (1933), four for Selected Poems (1929), three for the 1927 revision of Poems (1895), two for Early Poems and Stories (1925), one for the 1912 revision of Poems (1895), and two for the first edition of Poems (1895). Thirty-two of the forty were revised finally, then, from 1929 to 1939: the last ten years of Yeats's life. And, as a rule, the later the final polishing the heavier the revisions throughout. The problem becomes more complicated when we study an early poem like 'The Ballad of the Foxhunter,' published first in the magazine East and West for November 1889, then in United Ireland for 28 May 1892, then in The Countess Kathleen, etc. (September, 1892), then in Poems (1895), and so on. In discussing this poem, one critic speaks blithely of 'the early version' and then proceeds to quote the first four stanzas of the version in Poems (1895). But this is the fourth version. He ignores the three earlier versions. Here are lines 5-8 as they appear in each of the versions up to 1895.

In the first, that in East and West (November 1889) they read:

In the second version (United Ireland, 28 May 1892) these become The third version (The Countess Kathleen, etc., September 1892) is identical with the second; the fourth (Poems, 1895) changes from the third: It will be noted that in line 5 'stable' has become 'stables' and in line 8 'slowly' has become 'gently.' One wonders by what magic this last version, differing so markedly from the first version and in degree from the second and third versions, becomes 'the early version.'

The same difficulty is present in 'The Madness of King Goll' that likewise has several early versions: the first in The Leisure Hour for September 1887, the second in Poems and Ballads of Young Ireland

(Dublin, 1888), the third in The Wanderings of Oisin, etc. (1889), and the fourth, and except for minor changes the final version, in Poems (1895). Each of the first three is extensively revised, yet we have a critic saying that 'The Madness of King Goll' was revised so thoroughly that the 1895 version is practically a new poem, the implication of the context being that there was but one earlier version. Another critic of Yeats's revisions, in commenting on the line 'And every mumbling old man said' from the second version of the poem, remarks that the metre exacts 'an unnatural emphasis [on the word 'old']' and that Yeats improved the metre for Poems (1895) by changing the line to 'And every ancient Ollave said.' But in the first version this line has a metre similar to the final metre: 'And every whispering Druid said.'

There are other examples at hand, but these illustrate sufficiently something of the problem inherent in the term 'early poem' and likewise the possible errors inherent in a study of 'the early version' of a poem. To come back to my question, 'What is an "early Yeats poem"?', I should answer it by saying that it is a poem written originally before 1892. And immediately I should add that in any discussion of such a poem the details of its early publication must be given and the exact printing date of any quotation.

Revisions of the early poems continued, as I said, right up to the end. It is not until the definitive edition, for example, that we can clear up the vexing question of whether line 39 in 'The Song of the Happy Shepherd' should be 'Rewarding in melodious guile' or 'Rewording in melodious guile.' The textual history of the line is curious and provocative. In the first printing, in The Dublin University Review for October 1885, and in successive printings up to the 1901 revision of Poems (1895), it was 'Rewording.' In the 1904 revision of Poems (1895) it became 'Rewarding.' In The Poetical Works I (1906) and The Collected Works I (1908) it is 'Rewording.' But in the 1908 revision of Poems (1895) and in all subsequent printings through Collected Poems (1933) it is 'Rewarding.' In the definitive edition it becomes 'Rewording'—a reversion to the original word and concept. (I am aware that what sometimes seem to be reversions in an author's work are nothing more in fact than the careless use of an earlier edition as copy for a later edition, but the reversions discussed in this paper were, I believe, purposely made by Yeats and represent a rejection of his own revisions.) Another revision in the same edition is a punctuational change—again a reversion—in line 7 of 'Ephemera,' 'When the poor tired child, Passion, falls asleep.' In the first printing, The Wanderings of Oisin, etc. (1889), the terminal punctuation was a period; but in the next printing, Poems (1895), the period was replaced by a

colon that remained through Collected Poems (1933). The colon shifts the implication of the next two lines: 'How far away the stars seem, and how far / Is our first kiss, and ah, how old my heart!' In the definitive edition the period reappears.

He even revised, in the definitive edition, his spelling of certain Gaelic names. In 'The Wanderings of Oisin,' for instance, 'Aed' becomes 'Aedh,' and 'Blanid' becomes 'Blanaid' (although the trade editions of 1950 and 1951 do not have the latter change[10]). He makes several slight punctuational revisions in the same poem; an example is the deletion of the terminal commas of lines 405 and 406, Book I. Here are lines 404-407 as they appear in all printings from Poems (1895) through Collected Poems (1950-51) except the definitive edition:

And here they are in the definitive edition: Apparently Yeats felt that the elision of the commas after 'cold' and 'love' make the last three lines more clearly modify, rather than be in apposition with, the first line.

Collected Poems (1933), given frequently by textual critics as the source of numerous revisions, has actually not very many and they are not very extensive. Among them are the insertion of a comma after 'last' in line 33 of 'The Rose of Battle': 'And when at last, defeated in His wars'; the hyphenation of 'bean-rows' and 'honey-bee' in line 3 of 'The Lake Isle of Innisfree'; the final changing of 'will' to 'with' in 'When You Are Old': 'And loved your beauty with love false or true' [the printings oscillate between 'with' and 'will,' the first 'will,'[11] that displaces the original 'with,' being perhaps due to a misprint]; the substitution of 'fear' for 'fears' in line 6 of 'Who Goes with Fergus?': 'And brood on hopes and fear no more'—it had been 'fears' from the

first printing in 1892; and a change in the punctuation in line 13 of 'The Two Trees': 'There the Loves a circle go.' This line, that Yeats first used in Selected Poems (1929), he originally punctuated 'There the Loves—a circle—go.' In 1933 he eliminated the dashes. In line 18, Book I, of 'The Wanderings of Oisin,' 'Where passionate Maeve is stony-still;', that makes dubious sense without a hyphen between 'stony' and 'still,' he hyphenates the phrase—another reversion, for it had been hyphenated in the first printing; and, also in 'The Wanderings of Oisin,' he changes the spellings of some of the Gaelic names. One of those changed is 'Conhor' that becomes 'Conchubar' and necessitates a rewording of line 80, book III, from 'And the names of the demons whose hammers made armour for Conhor of old' to 'And the name of the demon whose hammer made Conchubar's sword-blade of old.'

Selected Poems (1929) includes thirteen of the shorter early poems and 'The Wanderings of Oisin.' Of the thirteen shorter poems, six are revised slightly. Two have major revisions, 'The Man Who Dreamed of Faeryland' and 'The Two Trees,' revisions often credited to Collected Poems (1933). One critic strides in thus: 'Yet in rewriting the poem ['The Man Who Dreamed of Faeryland'] for the 1933 Collected Poems . . . .' Our critic buttresses this surprising statement by boldly assigning the following versions of lines 41 and 42 to the 1901 revision of Poems (1895):

and then says they were revised in 1933 to He is almost right about line 41, but quite mixed up about line 42. In the original printing in The National Observer for 7 February 1891 the lines read: Line 41 stayed unchanged until Selected Poems (1929) when it was revised to 'Did not the worms that spired about his bones': its final form. Line 42, however, went through several changes. The first was for its second printing, The Book of the Rhymer's Club (London, 1892), where it had a terminal comma added. It kept this form until Early Poems and Stories (1925) where it became 'Proclaiming with a low and reedy cry,' that was changed finally in Selected Poems to 'Proclaim with that unwearied, reedy cry.' The remaining revisions in the poem are likewise incorrectly assigned to 1933. We have lost four years in a poet's development. (Incidentally, no comment is made on one of the most striking 1929 revisions: 'That if a dancer stayed his hungry foot / It seemed the sun and moon were in the fruit' that replace 'A Danaan fruitage makes a shower of moons, / And as it falls awakens leafy tunes.')

The extensive revisions in 'The Two Trees' for Selected Poems are exemplified by the changes in lines 13-20. The original version of these lines that except for two unimportant punctuational changes lasted from its first printing in The Countess Kathleen, etc. (1892) through the final reprinting in 1929 of Poems (1895), reads,

Here is the revision for Selected Poems: becoming, I think, a clearer statement of the desirability-of-innocence theme so superbly expressed ten years earlier in 'A Prayer for My Daughter.'

For 'The Wanderings of Oisin' there are fifty-two changes, forty-two of which are punctuational with seven of the forty-two changes being reversions. The remaining ten changes are word and phrase revisions, with five 'maidens' becoming 'ladies.' One wonders whether to ascribe this to the years or better diction.

Revisions for Early Poems and Stories (1925) and for the 1927 edition of Poems (1895) are considerable, among others being 'Fergus and the Druid,' 'A Cradle Song,' 'The Sorrow of Love,' 'The Lamentation of the Old Pensioner,' and 'To Ireland in the Coming Times' for the 1925 volume; 'The Ballad of the Foxhunter' and 'The Countess Cathleen in Paradise,' for the 1927 volume. 'The Dedication to a Book of Stories selected from the Irish Novelists' was revised originally for its publication in The Irish Statesman for 8 November 1924; it was

printed in Early Poems and Stories in its revised form. 'Cuchulain's Fight with the Sea' has numerous revisions for both volumes; we might glance at these revisions, in particular the revised ending and some criticisms of it.

For Early Poems and Stories Yeats changed the title from 'The Death of Cuchulain,' deleted seven lines, and redid forty-two of the remaining eighty-six lines. For the 1927 Poems he deleted one line and redid twenty-two—ten of which he had just finished revising in 1925. Probably the most striking of the changes is the new ending of the poem, done for Early Poems and Stories. Here is the ending[12] Yeats revised:

and here the revision: Criticism of the old vs. the new ending varies widely. We have this from one critic: 'The second [ending] is fine too, but has not the same sense of water flowing on and on that is heard in the other. "And the waves flowed above him and he died" hold the invulnerability of the sea, and the majesty . . . of death.' But another writes that the new ending 'transformed a mediocre poem into a work of quite extraordinary power.' A third apparently thought both endings were the same: 'The changes [in the revision of the entire poem] do not, however, concern the contents—apart from one small detail: the name Finmole has disappeared, . . .' A little textual reading, not studying, would have helped number three. Incidentally, I agree with the second critic—that the poem is transformed into a work of extraordinary power.

'The Lamentation of the Old Pensioner' was made over for the 1925 volume into an entirely new poem. This rewriting has been discussed again and again; so far as I know, however, no one has commented on the brilliantly bitter and savage satire in the mere use of the word 'transfigure.'[13]

Numerous other revisions of the early poems occur in the following volumes[14] —going back chronologically from 1925: the 1912 reprinting of Poems (1895), The Collected Works I (1908), the 1901 reprinting of Poems (1895), the 1899 reprinting of the same volume, the 1895 Poems itself, The Countess Kathleen, etc. (1892), and The Wanderings of Oisin, etc. (1889). The last two are listed because of the many changes Yeats made between initial periodical publication and publication in these first two volumes of his verse. There is, as I have implied, a tendency either to ignore the earliest changes or to be unaware of them and to assign the first revisions to Poems (1895).

I have emphasized by mention or discussion a number of the volumes for which the early poems were revised. A ranking of these volumes based on the number of revisions would show, in my opinion, Poems (1895) as easily the first, followed by The Countess Kathleen, etc. (1892), The Wanderings of Oisin, etc. (1889), the 1912 reprinting of Poems (1895), Early Poems and Stories (1925), the 1899 reprinting of Poems (1895), The Collected Works I (1908), the 1901 and the 1927 reprintings of Poems (1895) in that order, Collected Poems (1933), and The Poems of W. B. Yeats I (1949). A ranking based on the significance of the revisions would show Poems (1895) as again first, followed by the 1925, 1912, 1908, 1892, 1899, 1901, 1927, 1889, 1933, and 1949 volumes. Selected Poems (1929), despite its importance, is not listed because it includes less than half the early poems. I think my ranking of the order of significance would be authenticated by a detailed study, a study that I believe would likewise show that reversions in later printings of the early poems most often go back to the 1889 and 1908 volumes.

Before leaving the early poems I'd like to call attention to some of the lines that should have been revised. One example is lines 15 and 16 of 'The Sad Shepherd': 'But naught they heard, for they are always listening, / The dewdrops, for the sound of their own dropping.'—almost unchanged in all printings; another is lines 5-8 in 'The Falling of the Leaves,' with its tongue-twisting line 6:

the third is lines 32 and 33 in Book 1 of the 'Wanderings of Oisin,' '"Why do you wind no horn?" she said, / And every hero droop his head?'—a hackneyed rhythm 'reminiscent,' as Peter Allt once remarked to me, of '"Shoot if you must this old grey head, / But spare your country's flag," she said.'; and, as a last example, line 51 in the same book of the same poem, 'Through bitter tide on foam-wet feet?' Yeats had a deal of trouble with this line. Finn is asking Niamh why she is paying a visit to his country. In The Wanderings of Oisin, etc. (1889) part of line 50 and line 51 read: In Poems (1895) this becomes In the 1899 revision of Poems (1895) 50 and 51 are Fifty remained the same, but fifty-one received its final form in Early Poems and Stories (1925): Part of Niamh's answer—lines 57-59, in the final form they got in the 1912 revision of Poems (1895)—is perhaps even worse,

The later poems, using Yeats's own division, include everything from The Wind Among the Reeds (1899) to Last Poems and Plays (London; New York, 1940), a total of 346 poems. The observations made and the conclusions reached about any textual study of the early poems are likewise applicable to the later poems. This is borne out by the incidental references I have made so far to the texts of a few of the later poems.

Beginning in 1903 a new textual source, the Dun Emer press, after 1907 called the Cuala press, comes into the picture. The Dun Emer press was started by Yeats's sister Elizabeth. Yeats himself took an active part in its work and for some years chose and edited the books it issued, as well as having most of his own poems, beginning with In the Seven Woods (1903), printed by it. As a matter of statistics, there are 308 poems in the definitive edition that date from 1903 to January 1939; of these, 264 are in Dun Emer-Cuala books and of that number 94 are first printings.

Because of Yeats's close supervision of the Cuala press books, one

might be led to believe that here is a textual source more important than others, and that here if anywhere would be a finality not found in other printings. But not so; the Cuala press printing of a poem is not on that account likely to show more permanence than any other printing of the poem.

But apparently the magic of the words 'Cuala Press' indicated to at least one critic that if a Yeats poem was printed in a Cuala press edition the poem would never again be changed. So in discussing 'Broken Dreams' and 'The Second Coming' from 'the 1918 and the 1920 volumes, . . . The Wild Swans at Coole and Michael Robartes and the Dancer,' he gives these two Cuala press volumes as the sources for his quotations from the two poems. But in both cases he quotes 1933 versions. This is not too serious for 'Broken Dreams' where the changes are minor, but it is rather disastrous for 'The Second Coming.' The version of that poem in Michael Robartes and the Dancer has, as lines 13 and 17,

Whereas we are told they read Hence by an assumption of the inviolability of a Cuala press printing the chance of commenting on two of Yeats's most brilliant changes is lost: of pointing out, for instance, how 'somewhere in sands of the desert' brings into the poem a sense of the vastness and desolation of 'the lone and level sands' stretching 'far away'; and how 'Reel' for 'Wind' expresses far more effectively the bewildered terror of the desert birds. These revisions in 'The Second Coming' were actually made for Later Poems (1922) and were retained to the end, but because when he quotes the poem our critic italicizes 'Spiritus Mundi' in line 12 it is evident that he used the 1933 version where the italicization first appears. (Incidentally, there was no '1918' Wild Swans at Coole; the Cuala edition was 1917 and the trade edition 1919. The colophon in the Cuala edition of Michael Robartes and the Dancer says it was finished on All Souls' Day, 1920, the title page reads 1921.)

I have indicated that just as we are in trouble with the careless use of the phrase 'early poem,' so with the phrase 'later poem.' Here, for example, we have one of Yeats's biographers talking about the poems in In the Seven Woods (1903) and saying that 'The first, and perhaps the finest poem in the new "middle-aged" group, was "The Folly of Being Comforted."', and he then cites lines 7-14 of, presumably, the

1903 version of that poem, but actually of the 1933 version that except for one minor change made for the definitive edition is the final version. Now textually, he could hardly have picked a more difficult poem than "The Folly of Being Comforted.' From its publication in The Speaker for 11 January 1902, down through the definitive edition, there are 36 changes in the published versions and a complete shift of meaning in lines 8-11,—a shift that appears first in the Tauchnitz A Selection from the Poetry of W.B. Yeats (1913) and is completed in Collected Poems (1933). Lines 8-11, in the 1903 Dun Emer and trade editions of In the Seven Woods, were The terminal semi-colon at line 9 makes that line modify line 8; the comma after 'her' in line 10 makes the remainder of that line and line 11 modify the first part of line 10. In 1913, after some intermediate and relatively unimportant changes, the punctuation was changed significantly: Now line 8 stands by itself, line 9 modifies lines 10 and 11, and there is a specific meaning in lines 10 and 11. The next significant change was for Later Poems (1922) where a colon replaces the semi-colon in line 8; the last was for the 1933 printing where a terminal comma after 'stirs,' in line 10, finally gave Yeats, it would seem, the meaning or meanings he wanted. It is this 1933 version that we are presented with as the 1903 version: proof that when a 'later' as well as an 'early' Yeats poem is discussed and quoted the exact printing date of the quotation must be given and pertinent facts from the poem's printing history.

The more important revision volumes for the later poems are The Wind Among the Reeds (1899)—on the basis of the many revisions from initial magazine printings; Poems 1899-1905 (1906); The Green Helmet and Other Poems (1912); Responsibilities (1916); Later Poems (1922); The Tower (London; New York, 1928); The Winding Stair and Other Poems (London; New York, 1933); Collected Poems (1933); and A Full Moon in March (London, 1935). It would be misleading to list these volumes in an order based either on number or significance of revisions, for in only two of them, Later Poems and

Collected Poems, are more than a handful of the later poems printed or reprinted. As a matter of fact, the later poems in their entirety appear only in the definitive edition of 1949 and the trade editions of 1950 and 1951.

These, then, are some of the textual facts and problems the student of Yeats must deal with. There are others: for instance, the existence of one version of a poem in the prose works and another version in the poetry volumes, examples being 'The Moods,' 'Into the Twilight,' and 'The Happy Townland';[15] or the scores of title changes Yeats made;[16] or the changes made for, apparently, the sake of the rhythm only: 'They'd mauled and bitten the night through' to 'They mauled and bit the whole night through';[17] or 'They'll cough in the ink to the world's end,' to 'All shuffle there; all cough in ink.'[18]

Finally, two amusing revisions, the second probably a misprint.

In the first printing of the poem 'Lullaby' in The New Keepsake (London, 1931) lines 3-6 read

As revised for its next printing in Words for Music Perhaps and Other Poems (Cuala Press, Dublin, 1932) they become In the succeeding printing, The Winding Stair and Other Poems (1933), line six reverts to and remains so.

The second, the probable misprint, concerns line 36 of 'Adam's Curse': 'To love you in the old high way of love.' In three of the printings, The Gael (February 1903),[19] The Poetical Works I (1906), and Selected Poems (1921) the words 'high' and 'way' have been joined, and the line reads 'To love you in the old highway of love.'

I give these two revisions to the Freudians 'for free.'

Notes