Date and circumstances of composition. The poem was begun (and possibly first conceived) by WBY during his stay at Coole Park in the summer of 1901. The poet arrived to stay with AG in late Jun., and it is likely that he was able to read parts of CM, which she was then in the process of completing. This either gave WBY the idea, or confirmed him in the intention, of composing a series of narrative Irish mythological poems. In a letter (probably from Jun.) George Russell, telling WBY that he was ‘delighted to hear that you are writing verse again,’ remarked that ‘The story of the Baile is a little fantastic but I am sure you will make a beautiful thing out of it’ (LTWBY 1, 97). Writing to JBY on 12 Jul., the poet informed his father of his developing plans (CL 3, 87):

I am writing narrative poems of the Irish heroic age, the first things of the kind I have done since I wrote ‘The Wanderings of Usheen’. They will make a series I have intended to write ever since I was 20. I have shrunk from beginning them until my blank verse seemed sufficiently varied.

It is extremely likely that composition of the poem had begun by this stage, though WBY's remark makes it possible that the earliest stages were in blank verse rather than the eventual rhymed tetrameter couplets. Just over a week later, WBY wrote from Coole to Robert Bridges, and it is clear by now that the poem was taking its more familiar shape (20 Jul. 1901, CL 3, 91):

I am writing a half lyrical half narrative poem on two old Irish lovers, Baile, Honey Mouth and one Alyinn – to write the names as they are spoken. I then go on to other stories of the same epoch. I have in fact begun what I have always meant to be the chief work of my life – The giving life not to a single story but to a whole world of little stories, some not indeed very little, to a romantic region, a sort of enchanted wood. The old Irish poets wove life into life thereby giving to the wildest and strangest romance, the solidity and vitality of the Comédie Humaine and all this romance was knitted into the scenery of the country. ‘Here at this very spot the faery woman gave so and so the cup of magic mead. Not there by the hillock but there by the Rock’ and so on. This work has not been possible to me hitherto, partly because my verse was not plastic enough and partly for lack of a good translation. But now my friend Lady Gregory has made the most lovely translation putting the old prose and verse not into the pedantic ‘hedge school master’ style of her predecessors, but into a musical caressing English, which never goes very far from the idioms of the country people she knows so well.

Progress was fast, and on 25 Jul. WBY wrote to ECY that ‘I am getting towards the end of a longish poem – longish for me, 60 or 70 lines on an Irish legendary subject’ (CL 3, 94). There is a suggestion here that the ‘narrative’ element of the poem might have been composed before what the poet later called the ‘lyric’ interpolations were supplied: though even then, the number of lines is smaller than those carrying the main story in the finished poem. By 9 Aug., WBY could tell George Russell that ‘I have written a large part of a longish poem on Baile, the sweet spoken, and Aillinn’ (CL 3, 104), but this was in fact the entire poem, since the Berg MS copy in WBY's hand (see Textual and publication history) carries the poet's signature with the date ‘August 9th | 1901’. On 11 Aug., WBY wrote to T. Sturge Moore of how ‘I am doing all the chief stories of the first heroic age in Ireland in a series of poems,’ and reported that ‘I have just finished a half narrative half lyrical poem of about 200 lines, which is I think good’ (CL 3, 105). A version sufficiently complete to be sent abroad was dispatched to Lafcadio Hearn, acknowledged in a letter from him to WBY of 24 Sep. 1901: the editors of CL believe this to have been sent by WBY in Aug. (CL 3, 101).