is stressed in the verses for Sec. 4: Dochotar martrai lahechtrando ceneris hite trebaid cencuid cubais isinninis, "They suffered martyrdom through the unbelieving strangers who dwell in the island without a trace of confession." 31

The descriptions, says Zimmer, reflect an actual journey such as Irish clerics of the seventh century are known to have taken to the Orkneys, Shetland Islands, and the Faroes.32 The stream with the taste of new milk (Sec. 2) is the Gulf Stream flowing west of the Hebrides. From Dicuil's description of the birdinfested islands of the Faroes the author drew material for Sections 5 and 8, which also reflect the old story of Elijah under the tree of life in paradise preaching to the souls in bird shape, who beat their wings till blood flows. The latter conception, the author contaminated with the conception, appearing in Imram Curaig Hua Corra,3* of angels in bird shape surrounding God, the most brilliant bird being the archangel Michael.

In his later study, in which he stressed the knowledge of the Norse on the part of the Irish of the seventh century, Zimmer elaborated his view that an actual voyage of Ionan monks is reflected in the story, and that the monstrous inhabitants of some of the islands are the Norse. The "manes" of the dog-headed men represent the beards of the Northmen.35 From Adamnan Zimmer

31 Zimmer explains cubais as the equivalent of confessionis. Cf. Thurneysen's translation "ohnes alles Gewissen." Windisch, Irische Texte, I, 456, gives conscientia for cubus (from Grammatica Celtica).

32 Zimmer cites the passage from Dicuil, De Mensuris Terrarum (825 A. D.):

Sunt aliae insulae multae in septentrionali Britanniae oceano, quae a septentrionalibus Britanniae insulis duorum dierum ac noctium recta navigatione plenis velis assiduo feliciter vento adiri queunt. Aliquis presbyter mihi rettulit quod in duobus aestivis diebus et una intercedente nocte navigans in duorum navigula transtrorum in unam illarum introivit. Illae insulae sunt aliae parvulae. Fere cunctae simul angustis distantes fretis, in quibus in centum ferme annis heremitae ex nostra Scotia navigantes habitaverunt. Sed sicut a principio mundi desertae semper fuerunt, ita, nunc causa latronum Nortmannorum vacuae anachoritis plenae innumerabilibus ovibus ac diversis generibus multis nimis marinarum avium.

33 Dá Brón Flatha Nime, "The Two Sorrows of the Kingdom of Heaven," ed., G. Dottin, R. C., xxI (1900), 349-87.

34 R. C., XIV (1893), pp. 22-63. Cf. pp. 31, 33.

35 Zimmer notes the striking effect made upon the Irish by the bearded

quotes the striking parallels to the imrama found in the accounts of the sea journeys of such sixth-century Ionan monks as Cormac mac Liathain. Although he had previously accepted the personal participation of Colum cille as a part of the original form of the story, he here suggested that possibly not the great saint himself, but the Ionan establishment, was meant, or that perhaps the author of our text of the tale had substituted Colum cille for a possibly unnamed successor of Colum cille in his original. Zimmer accepted the statement of the text of Prose A that Domnall was the seventh-century Domnall mac Aed mac Ainmire.

Thurneysen was convinced that Poem formed the basis for the whole tradition, Prose A being derived, with changes and additions, from Poem, which in form is a complete poetic unit. The author of Prose A, wholly reliant on Poem for his material, was forced to supply an introduction from the very meagre intimations in Poem. Stanza 2 of Poem placed the voyage i tosach flatha Donncodha iar ndith Domnaill," in the beginning of the kingship of Donnchad after the death of Domnall." The author of Poem, in so dating the events of the tale, must have had in mind a wellknown King Donnchad-he must have meant the much mentioned king of Ireland, Donnchad mac Domnall, whose father, Domnall mac Murchado, had died in 763, Donnchad himself becoming king in 769. The fact that this Donnchad was named in the Félire Óengusso Céli Dé, "The Martyrology of Oengus the Culdee," may have suggested laying the story in his reign, which would fit well the poet's purpose of prophesying the Viking invasion as imminent. Whether the death of a brother of Donnchad, Fiacha, at the hands of the men of Ross, is historical or not is uncertain, though the statement, unmotivated by the story, that afterwards only two of the brothers of Donnchad were alive (St. 47) might point to a recollection of historical facts.

The author of Prose A, according to Thurneysen's hypothesis,

Norse and recalls that the Irish champion Cuchullin is presented in several descriptions as beardless.

36 Whitley Stokes, ed., Henry Bradshaw Society (London, 1905). The entry in question, Prologue 221, reads: Donnchad the wrathful, ruddy, chosen, or Victorious Bran of the Burrow, visiting tombs takes not from me the weariness of weakness." Thurneysen thought the diction of this document perhaps influenced the author of Poem.

misunderstood the reference in Poem to Snedgus and Mac Riagla as belonging to muntir Coluim Cille, " the family of Colum cille." The poet must have meant merely that they belonged to the monastic establishment of Colum cille; but the author of Prose A took the expression to mean that the two heroes were actual contemporaries of Colum cille. This misunderstanding produced in the mind of the author of Prose A a chronological difficulty, since Colum cille belonged to the late sixth century, Donnchad to the late eighth century. To solve this difficulty, he went back to an older king Domnall, whose rule ended in 642 and whose father Aed mac Ainmire was a contemporary of Colum cille, having died in 598. Although it is true that the reign of this king would come too late, this approximate synchronism was sufficient for a narrator and narrative of this sort. But this Domnall had no son Donnchad, and the author of Prose A invented a son Donnchad, and made him ruler over Tir Connell, as Fiacha was ruler over the men of Ross. Fiacha's tyranny was utilized for the motivating of his murder at the hands of the men of Ross and Colum cille was brought in to give the counsel that sixty couples of the men of Ross be set adrift on the sea as punishment. Having provided this introductory and motivating machinery, lacking in Poem, the author of Prose A went on with the story in Poem, following his source closely. He shortened occasionally, as in the Elias episode, which probably was not clear to him.

Prose B Thurneysen thought was based on Prose A, perhaps in the form we have it, with the verses of Poem interspersed among the prose sections. The author of Prose B was less learned than the author of Prose A. He not only followed Prose A in seeing in Domnall the seventh-century Domnall mac Aed mac Ainmire, but he blundered in having this Donnchad succeed Domnall as king of Ireland. He also introduced the fifth-century Maine mac Niall. He was poorly versed in geography, having Fiacha rule not only over the men of Ross but over Mugdorn Maigen as well, presumably because these districts at times had the same chiefs. Thurneysen rejected the theory that the presence of both similarities and differences in the introductory parts of Prose A and Prose B indicated a common source for the two forms. He thought the differences were to be explained partly as variations which even a medieval author allowed himself, and partly as a closer following of Poem by the author of Prose A.

The discussion which follows, based on some comparison of all the known versions, and a re-examination of the evidence afforded by the annals, the other imrama, and various hagiological documents, proposes to give some reasons for the following convictions:

1. Although Zimmer's interpretation of the tale can not be endorsed in all its details, this scholar was right in stressing the real experiences of Ionan clerics as the probable source of the tradition out of which our imram sprang and perhaps right also in attributing to the author some churchly and patriotic purpose. 2. Colum cille belonged to the original form of the tale. The Domnall mentioned is Domnall mac Aed mac Ainmire. 4. Donnchad, whether an historical person or not, was originally conceived as the son of Domnall mac Aed mac Ainmire, and was ruler over Tir Connell, not over Ireland.

3.

5. The time-setting is that heroic period of Irish church history which embraces parts of the sixth and seventh centuries-the era of Colum cille, of the famous sea-pilgrims, and of the missionary journeys to continental Europe.

6. The chronological error of making Colum cille and Domnall contemporaries is chargeable to the original author rather than to a reviser.

That Poem is the sole basis of the later forms of the imram, or that the author of Prose A relied exclusively upon it for his legendary materials, seems to me improbable. The presentation of the story in Poem is confessedly incomplete, and this fact, as well as a recognition of the character of the verse itself, indicates that the author was writing a poetic, half lyric paraphrase of a story which must have been known to him, and to his audience, in a more complete form. Whether the tradition were available to him in written form one can not be certain, although it would not seem improbable that it was. Kuno Meyer thought that Poem was probably written by the poet who composed the poem on Maelduin's voyage. If this conjecture be accepted, there is

37

97 ZCP, XI (1916-1917), 148. The poem itself is printed by R. I. Best, Anecdota from Irish Manuscripts (Dublin, 1907), 1, 50 ff. and by Meyer, loc. cit. Cf. Thurneysen, ZCP, v, 421; vI, 235; VIII, 80. Thurneysen argues against Meyer's view, and thinks the evidence suggests imitation by the author of the Maelduin poem rather than identity of authorship. (ZCP, XII, 1918, 278.)

strong analogical evidence that the author of Poem was basing his verses on a prose account, as he seems to have been doing in the case of the Maelduin poem.38 Fortunately, the old prose version of Maelduin's journey, Imram Curaig Maelduin, "The Voyage of Maelduin's Boat," is preserved. It perhaps dates from the eighth century or earlier.39

The following extracts from Poem, quoted from Professor Thurneysen's text and translation," will suggest the general character of the verse. I include the stanzas which give the hints for the background of the story:

1. Snedgus ocus Mac Riaghuil do munntir Coluim Chille, ros bui do brig: ro charsad rig flatha níme.

Snedgus und Mac Riaguill von Colum Kille's Genossenschaft-sie hatten solche Tüchtigkeit; sie liebten den König des Himmelsreichs. 2. Ar-roldatar fairggi fergaidh fedma comluind

i tosach flatha Domcodha iar ndith Domnaill,

Als sie auf die zornige See der Kampfesanstrengung gegangen waren im Anfang der Herrschaft Donnchads nach Domnalls Untergang,

3. Dos rala (for) mor di ruinib de-din nad gairdein nindsib inn aceoin uathaigh osin fairrgi.

trafen sie auf manches von den Geheimnissen Gottes-kein kurzer Schutz! auf den Inseln des schrecklichen Oceans, über der See.

The following group of stanzas is from the section dealing with the visit to the Isle of the Men of Ross. The chief has just welcomed the clerics.

...

45. Can bar cenel?" ni chelim ruin, Woher stammt ihr?" ... "Aus Irland, eine sicherer Fahrt

"A tir Erind-eirim nglindi—; di muintir duin (n) Coluim Chille."

39 Zimmer, ZDA, XXXII, 149; Thurneysen, ZCP, VIII, 80; x, 278; Meyer, ZCP, XI, 148.

3o The imram is preserved in LU, YBL, Harleian 5280, and Egerton 1782. YBL and Harl. include the verse paraphrase. As to the date of the prose version, see Zimmer, ZDA, XXXIII, 48; d'Arbois de Jubainville, Cours, Vol. 1, ch. 8. But J. Strachan, Trans. London Phil. Soc., 1891-94, criticizes some of the linguístic evidence on which Zimmer based his view. It must perhaps be admitted that the question of date for Imram Maelduin and its relations with Navigatio Brendani are not settled. Cf. the present writer's remarks in Manly Anniversary Studies, p. 283, n. 2. 40 Zwei Vers., pp. 9 ff., 21 ff.