Bluegrass Messengers

The "Clerk Colvill" Mermaid- Harbison Parker 1947
The "Clerk Colvill" Mermaid
by Harbison Parker
The Journal of American Folklore, Vol. 60, No. 237 (Jul. - Sep., 1947), pp. 265-285

THE "CLERK COLVILL" MERMAID
By HARBISON PARKER

In "The 'Johnny Collins' Version of 'Lady Alice' " (Journal of American Folklore, Vol. 58, No. 228, April-June, 1945), Samuel P. Bayard demonstrates the identity of this variant of "Lady Alice" (Child No. 85) with the story told in "Clerk Colvill" (Child No. 42). He then endeavors to show that the replacement of the elf-woman who plays the role of the femme fatale in the Scandinavian versions of this ballad story by a mermaid in the "Colvill" variants and a maiden "A-washing a marble-white stone" in the "Johnny Collins" group is probably to be laid to the intrusion of ideas connected with the figure of the banshee in Gaelic folklore, who "has in all probability lent some of her nature and activities to the elfin women of 'Clerk Colvill' and 'Johnny Collins'." [1] He then makes a further conjecture:

And supposing that "Johnny Collins" is actually a version of "Clerk Colvill," as everything seems to indicate, we may perhaps make out one or two more fragments of its history. Noting its thoroughgoing difference in phraseology from the Scottish ballad, we may infer that it is the product of a recomposition of the older international piece. The recomposing may have been done in Ireland, although this cannot be proved; but undoubtedly there is more true Gaelic "bansheeism" about its fairy woman than appears in the texts of "Clerk Colvill." At any rate, it seems practically certain that the American variants of this version have been traditional in Ireland. [2]

Professor Bayard presents careful and detailed arguments in support of his two hypotheses, that folklore about the banshee has modified the Scandinavian elf-woman into a mermaid or a maiden washing a stone, and that the "Johnny Collins" versions, at least, have been current in Irish tradition. There are, however, difficulties in the way of these hypotheses; and there is, it seems to me, as strong support for the hypothesis that the metamorphosis of the elf-woman into a mermaid came about in the Shetland and Orkney Islands, the ballad having more likely passed through this milieu than through Irish ballad tradition. The difficulties besetting Professor Bayard's hypothesis will first be presented, and then the support for the Shetland- Orkney hypothesis will be adduced.

I
Professor Bayard bases most of his arguments in support of the hypothesis that elements of banshee lore have modified the motifs in "Johnny Collins" and "Clerk Colvill" upon an analysis of the first three stanzas of "Johnny Collins," especially of those in a text of his own collection, which he published in his article, and which I take the liberty of quoting here:

1. As Collins was walking the fields one day
All dressed in white linen so fine,
He spied a maiden, a pretty fair maid,
A-washing a marble-white stone.

2. She wrung her hands and tore her hair,
She waved with a lily-white hand,
Saying, Collins, dear Collins, come quickly here-
Your life is soon to an end!

3. She threw both arms around his neck,
She kissed both his cheeks and his chin,
Till the stars from heaven come twinkling down
On the banks where Collins jumped in. [3]

Professor Bayard raises several questions about this part of the ballad story:

. . .why should the maid who appears in the opening stanza be washing a marblewhite stone? Why, also, should she express such grief at seeing Collins (grief portrayed, as usual, by some of the common formulae of British balladry. . . ? How is she able to prophesy his speedy death? And why, after this prophecy, should she make love to him, or-as it happens in most of the texts-should he make love to her (a still more puzzling action)? [4]

He finds a partial answer in the relationship which he shows exists between the "Johnny Collins" and the "Clerk Colvill" ballads:

In both ballads the same ancient, floating motifs are joined together in the same order- forming a union which is truly "temporal and distinctive, . . . [5]

He seeks also to penetrate behind this relationship he so ably elucidates and to demonstrate that the complete answer to the more crucial of his questions, those concerning the maiden's washing, her ability to prophesy Collins's death, and her sorrowing in so doing, is to be found in the Gaelic folklore about the banshee. It is my opinion, however, that a more satisfactory answer to these questions can be found in the Scandinavian versions of this ballad story (to which Professor Bayard himself often looks for explanations of obscure details).

In answer to the question as to why the maiden should be "washing a marble-white stone," Professor Bayard points out that if we compare this line with the line in Version C of "Clerk Colvill," which depicts the mermaiden as "washin silk upon a stane,"[6] the obscurity is cleared up, for ... if we assume that the maiden was washing something on the stone, using it as washers in bygone times did to beat and fold garments on, we have a perfectly clear picture. [7]

This is the crux of his main argument, both for identifying the woman washing a marble-white stone in the "Johnny Collins" ballads with the mermaid, and for perceiving in both these figures traces of the banshee. Having pointed out that the idea of the maiden "washing a marble-whitestone" makes no sense until equated with the mermaid "washin silk upon a stane," he argues:

... if we concede what seems to me to be pretty clearly brought out in the texts: namely, that the washing done by these faery beings is apparently not purely casual, but is bound up with the tragedy in each piece, and has a sinister symbolism especially marked in "Clerk Colvill" by reason of the mermaid's speech-then, I think, it is hard to believe that the characters of the elf-woman in both ballads have not been modified by beliefs concerning the Gaelic "banshee." [8]
 
She is, of course, widely known as a mournful foreteller of death and disaster; but in this role we find her almost continually associated with water, and performing a ceremony of washing which has a grisly significance. Her station, we are told, is generally at fords in the river; the stone on which she folds the shirts of the doomed is in the middle of the water; [9] at times she is seen seated by the pool or stream washing the linen of those soon to die, and folding and beating it with her hands on a stone in the middle of the water-. . . and her being seen is a sure sign that death is near. Do not these details throw light on "Johnny Collins," with its washing on a stone, its lamenting woman, its death-prophecy, and its hero's leap into the water? [10]
. . . . . . . . . . . . . . . . . . . . . . . . . .

It seems that unless we assume that ideas about the banshee have intruded themselves into these ballad stories we shall be at a loss to account for the fairy woman's lamertation in "Johnny Collins" and the washing trait in both pieces. [11] The keystone of Professor Bayard's argument is his rejection of a previous suggestion that the sark which the mermaid is washing has the significance of a betrothal gift. Barbara M. Cra'ster, in a note in the Journal of the Folk- Song Society (London), in 1910, in which she was the first to notice the connection between the "Johnny Collins" and the "Colvill" ballads, says:

The description in "George Collins' [12] of the supernatural maiden "washing her marble stone" is most closely paralleled by the lines in "Clerk Colin" (Prof. Child's Version C), which run:

"An there he saw the mermaiden
Washin silk upon a stane."

In the other versions of "Clerk Colvill" she is found washing "a sark o' silk"; and this is explained by Prof. Child, in a note on "The Elfin Knight" (V, 284), as probably having reference to the old custom by which a shirt given to a man by a maiden signifies betrothal. [13]

Professor Bayard flatly rejects this suggestion:
... her belief that the shirt in "Clerk Colvill" signifies betrothal cannot be correct, since there obviously is no question of betrothal either between the mermaid and Colvill, or between Collins and the washing woman: both these men are either married or betrothed already, according to plain ballad statement or unavoidable inference. The mermaid, furthermore, is bent on slaying the clerk, while Collins' washing woman knows already that he is soon to die. [14]

Without the link of the laundered sark, no connection between the mermaid and the banshee can well be shown. Professor Bayard therefore must of necessity reject Miss Cra'ster's interpretation of the sark motif. He thinks that such a construction cannot be put upon it or upon the mermaid's "it's a' for you" in reply to Colvill's mention of the garment, since Colvill is already either married or betrothed. But does he believe that such a creature would have any compunction about trying to break up an engagement or a home? There seems to be no precedent in folklore for attributing such punctiliousness to a fairy woman. Indeed, the elf-women in the Scandinavian analogues seem not a whit disturbed by Sir Olav's insistence that he is already betrothed and is to be married on the morrow, but persist in their attempts to win him over; and in one version the Elf-King's daughter offers him:

A shirt of the silk so white and fine,
My mother has bleach'd in the moonbeam's shine. [15]

The offer of a silken shirt is made in five of the Danish variants [16] and two of the Norwegian [17] of this ballad story. In the two Norwegian and two of the Danish [18] the offer is not made till after Olav has stated that he cannot consort with the elves because he is to be married next day. In the remaining three Danish variants, he does not so state before the offer is made; but in reply to it he asserts that his "dearest" has already provided him with the article, thus reinforcing the betrothal significance, since otherwise there would be no point in his being explicit that it was his beloved who had furnished it. This offer of a shirt is considered to be an invitation to become betrothed, by Liestol and Moe in a note to the version of this ballad which they print in their collection, Norske Folkevisor:

First the elf-maiden asks Olav to dance without promising him anything. Then she promises him gold. Finally she promises to be his, for according to old custom a shirt was a love-gift from a betrothed maiden to her fiance. That a woman should sew a shirt for a man was a token that she was willing to give him her love. [19] Professor Bayard himself says:

It appears permissible to use foreign analogues of "Clerk Colvill" as well as British versions in the comparison. For since these texts are all evidently forms, not only of the same story, but of the same ballad, we may hope that a detail left unclear in one version may be more fully explained in another. [20] And surely it should be more satisfactory to explain a motif in the British versions on the basis of details from foreign forms of the ballad, rather than to elucidate it on the basis of hypotheses erected on folklore not indigenous to the known locale of these ballad versions.

That singers in Scotland likewise were familiar with the tradition of a shirt or sark as a betrothal gift is shown with explicit clarity in the ballad of "Allison Gross" (Child No. 35), wherein, as in the "Colvill" ballad and its Scandinavian analogues, the temptress is a being with supernatural powers, "the ugliest witch in the north country." [21] She offers many gifts to the man who supposedly relates the account, in an attempt to persuade him to become her "lemman sae true"; and he tells that

She neist brought a sark o the saftest silk
Well wrought wi pearles about the ban;
Says, Gin you will be my ain true love,
This goodly gift you sal comman. [22]

Furthermore, this ballad was supplied by Mrs. Brown of Falkland, from whom came also Version A of the "Colvill" ballad. [23] It therefore seems probable that in the stanza of Version A which reads:

"Ye wash, ye wash, ye bonny may,
And ay's ye wash your sark o silk:"
"It's a' for you, ye gentle knight,
My skin is whiter than the milk"

Mrs. Brown, for one, thought of the mermaid's reply as an allusion to the betrothal tradition of such garments, rather than as "a sinister piece of dramatic irony" [24] which Professor Bayard wishes to see in it.

Since this motif of the offer of a sark (always a silk one) as an inducement to a man to become betrothed to a supernatural being appears so often among the Scandinavian versions of this ballad story as to show it to be part of the tradition, and since the offer of the same type of garment (also silk) for the same purpose, by the same sort of creature, appears in another ballad, known, not merely in Scotland, but by the same person who knew one of the variants of the "Colvill" ballad, it seems to me that Professor Bayard is wrong in rejecting Miss Cra'ster's suggestion that the sark in the "Colvill" ballad is intended as a betrothal gift. And if he is wrong, then his hypothesis that the mermaid and the maiden washing the stone are figures evolved from the Scandinavian ballad story by the influence of lore about the banshee loses its main support; for if the sark in these ballads can be accounted for on the basis of details patently in the tradition of these ballads, it is unnecessary, and probably incorrect, to turn for explanation to Gaelic folklore, which cannot be clearly shown to have impinged on either branch of this ballad in the British Isles.

The repudiation of the banshee involves, of course, the sacrifice of a striking and simple explanation of the fact that the sark is being washed. Yet we are not left entirely at a loss to account for it. Professor Bayard remarks:

Among the occupations ascribed to fairies and other uncanny beings all over northwestern Europe is that of washing clothes. Supernatural people of the hills and "wild women" are described by tradition as washing their garments and hanging them out to dry just like mortals; and the attachment of such an occupation to elfin creatures anywhere need hardly be a matter for surprise nor would it be out of harmony with traditional beliefs. We have seen that the fairy mistress of "Clerk Colvill" is simply a hill- or wood-sprite in most versions of the ballad but becomes a mermaid, or some sort of water creature in the English and in one Italian form. And it seems reasonable to think that making a watersprite wash on a stone is simply the result of adding another familiar trait to a supernatural being. [25]

In the folklore of Scottish and Scandinavian regions is a tradition which may have reinforced (or conceivably have introduced) this motif of the laundered sark. In the Faeroes:

If a girl wishes to know the name of her future truelove, she must go out some very dark night quite naked, taking her with her sark, dip it into running water, . . . wring it out, and put it, rolled up, under her pillow. Then she will dream of him. [26]

In the Shetlands she had to . . . go before bedtime to a burn where three lairds' lands met, and dip a chemise in the water. When she retired, the "sark" was hung in front of the fire in her room, and while she kept anxious vigil, the man she was destined to wed would come into the place and turn the garment. [27]

This superstition, found also in the folklore of the northeastern part of Scotland[28] and the Borders,[29] show that the washing of a sark in a stream was understood to have connection with betrothal, and makes it seem probable that the singers of the "Colvill" ballad understood the sark to have amatory rather than mortuary significance.

The second question which Professor Bayard asks about the puzzling maiden at the stone is (to repeat):

Why should she express such grief at seeing Collins (grief portrayed, as usual, by some of the common formulae of British balladry), . . . ?

He deals with these "formulae" explicitly in two footnotes:

My own texts which preserve this detail usually have the maid wring her hands and tear her hair. Davis A and B . . . have "she screamed, she cried," and Cox A, E, . . . have a similar formula. [30] Only the British variant does not make her grief clear-it has her hail Collins, but not weep, wring her hands, and so forth. [31]

As a matter of fact, it is only the version which Professor Bayard prints (and perhaps others of his own collection) which makes her grief seem clear.

The British variants give:
She whooped, she holloed, she highered her voice. [32]

Cox's Version E gives:
She screamed, she cried, she changed her voice. [33]

His Version A has it that
She screamed, she cried, she changed her mind. [34]

Davis's Versions A and B both have merely:
She screamed, she cried. [35]

And Professor Bayard's is the only one which has
She wrung her hands and tore her hair.

This close similarity among the lines of the other versions renders the line in Professor Bayard's version suspect of being a corruption. It is, as Professor Bayard remarks, a common formula of balladry. It appears quite regularly in the American ballad, "Sweet William." Typical are the lines in a version collected in 1917 in Hardy County, West Virginia, from Mr. J.
Harrison Miller, who also supplied Cox's Version A of "Johnny Collins." His "Sweet William" has the lines:

She wrung her hands into her hair
Just like one who is in despair. [36]

But his "Johnny Collins" has:
She screamed, she cried, she changed her mind. [37]

It is evident that this singer, who knew both songs, did not confuse them at this point; it is equally clear that the opportunity for the line to drift from "Sweet William" to "Johnny Collins" was present. Another version of "Sweet William" was collected in 19I7 in Gilmer County, West Virginia, within the area whence came some of the "Johnny Collins" versions, and contains the line:

She [w]rang her hands, she tore her hair,
Just like a lady in despair. [38]

Two versions of the same ballad collected in North Carolina in 1916 have closely similar lines;[39] and a version of "Young Beichan" (Child No. 53) collected in Virginia in 1914 gives the line exactly the same as in Professor Bayard's version of "Johnny Collins."[40] The common occurrence of this line in other ballads in the American tradition and the limited occurrence of it in the "Johnny Collins" tradition, together with the fact that the "Johnny Collins" versions which contain it were all collected in the I930's or later,[41] whereas the line appears in other ballads in 1914, 1916, and 1917, make it seem very likely that the line is a corruption in Professor Bayard's version.

If so, his assumption that the washing maiden was in anguish when she prophesied Collins's death, and was thus exhibiting a trait of the banshee, is based on a line which does not belong in the ballad. [42] This likelihood that the only line which makes the maiden's grief unambiguous has drifted in from other ballads puts Professor Bayard on unsure ground when he says:

It seems that unless we assume that ideas about the banshee have intruded themselves into these ballad stories we shall be at a loss to account for the fairy woman's lamentation in "Johnny Collins" . . . [43]

If we abandon the notion that the fairy woman was uttering a lamentation, I think we can see in her whooping and holloing a reminiscence of the salutation to the hero uttered by the elfwomen in many of the Scandinavian analogues. In fact, the whole stanza which relates the woman's first address to Collins can, I think, be pretty closely equated with the greeting to Olav by the elf-women in the Scandinavian versions. The Hampshire version, which is the earliest collected of the "Johnny Collins" group, makes the connection most readily apparent:

She whooped, she holloed, she highered her voice,
And held up her lily-white hand,
"Come hither to me, George Collins," said she,
"And thy life shall not last thee long." [44]

Compare this with Grundtvig's Danish version (B), which is typical of several of the Scandinavian analogues:

The elf-king's daughter stretched forth her hand.
"Welcome, Sir Oluf, leave off your haste.
Bide a little, and tread in the dance with me." [45]

Keeping in mind Professor Bayard's remark that The drastic condensation which the story has undergone in "Johnny Collins" has no doubt helped to make it obscure, and thus facilitated misunderstanding and corruption as the song was handed down to later singers; facilitated likewise a loss of content. . . . [46] I think we can see a definite connection between the maiden's whooping and holloing and the elf-woman's "welcome, Sir Oluf"; between the maiden's holding up her lily-white hand and the elf-woman's stretching forth her hand; between the maiden's "Come hither" and the elf-woman's "Bide a little, and tread in the dance with me." (The dance-motif has, of course, dropped out.) As for the fourth line in the Hampshire version, which appears as "Your life is soon to an end" in Professor Bayard's version, the other American variants having lines of the same import, it can readily be equated with the lines in the Danish, Version B:

And wilt thou not dance with me,
Disease and sickness shall follow thee [47]

or, as the Danish Version D says, more menacingly,

Thy young life that will cost thee. [48]

Professor Bayard asks concerning the maiden in "Johnny Collins," "How is she able to prophesy his speedy death?" For answer, he turns to the banshee, who is "widely known as a mournful foreteller of death and disaster"; [49] but I think that her prophecy required no supernatural prognostic power, but was, rather, like the death-warning of the mermaid and the elf-women, who needed no more than mortal ability to make their predictions, since they themselves intended to carry them out. The mermaid tells Clerk Colvill he is going to die, but she has already made his death certain by the strip of sark. The elf-women in many of the Scandinavian versions tell Sir Olav that disaster is to be visited upon him, and thereupon they strike him, stab him, poison him, or force him to dance himself to death. I suggest that the maiden's startling "And thy life shall not last thee long" is a remnant of this threat made by the elf-women rather than a mere prophecy of death, such as the banshee is supposed to utter. There is, of course, a "drastic condensation," for the whole temptation scene between Olav and the elf-women, in which he consistently refuses their enticements, is lacking, and the beginning and end are represented in the two consecutive lines in "Johnny Collins":

"Come hither to me, George Collins," said she,
"And thy life shall not last thee long."

Another question which Professor Bayard raises about the maiden at the stone is ". . . why, after this prophecy, should she make love to him, or-as it happens in most of the texts-should he make love to her (a still more puzzling action)?" He regards this puzzling action as ... evidence that Collins had known her before the encounter narrated in the ballad.

It lies in the fact that he embraces her after she has foretold his death. Every variant of "Johnny Collins" that is at all complete or well ordered has the events in this order; and it is hardly likely that a man meeting a fairy for the first time would take her in his arms when she had just told him that his life was "soon to an end." [50] Yet this is not a completely satisfactory answer to the question, for no matter how familiar a man might be with a supernatural creature (or a mortal woman, for that matter), it would seem that a death-prophecy would give him pause. He who could proceed forthwith to embrace the prophetess and kiss her red, rosy cheeks, as Collins did (in the Hampshire version) commands remarkable insouciance indeed. The only satisfactory answer to this puzzle is to be found, I think, in the Scandinavian analogues. Professor Bayard has noticed that in the Faeroe Island versions the elf-maid brings the hero "a poisoned draught, makes him kiss her, and sends him home a dying man." [51]

In the Faeroe A and B versions, he first makes his choice of dying on the morrow rather than lying sick several years-a choice which the elf-women tenderheartedly offer the hero in many of the Scandinavian versions-drinks a poisoned draught offered him by one of the elf-women, and then, with the the sentence of death upon him, is commanded to kiss his murderess. He complies with this ironic request after she has not only told him he was going to die, but had given him the potion which was to kill him. Keeping in mind the "drastic condensation" which Professor Bayard says the story has undergone, and the "loss of content" which he says was thus facilitated, we can find in this sequence of action in the Faeroe versions a key to the puzzling activity in "Johnny Collins," in which the command to the hero to kiss the prophetess of his doom is omitted, and his amorous advances follow immediately upon the prophecy of his death in an astoundingly impetuous fashion.

We have now been able to answer the questions Professor Bayard has raised by comparing details in "Johnny Collins" with those in the Scandinavian analogues-a process which he himself states (as quoted above) is permissible. Furthermore, I suggest that a consideration of these details points to a gradual deterioration of the ballad story through garbling and condensation, rather than, as Professor Bayard suggests, "to a deliberate [italics his] condensation on the part of some traditional re-creator." [52]

We come now to Professor Bayard's hypothesis that "it seems practically certain that the American variants of this version have been traditional in Ireland." [53] This theory rests on grounds even weaker, it seems to me, than those for the banshee hypothesis. Indeed, the Irish-tradition hypothesis and the banshee hypothesis derive their main support from each other. If the banshee, then Irish tradition; if Irish tradition, then the banshee. Professor Bayard does not point to any Gaelic or Irish touches in the diction or style of the American versions; nor does he present any variants gathered in Ireland. His case for Irish tradition rests (except for the banshee) on the fact that many of the American versions were collected in a region settled largely by Irish, and on the presence of the place-name "Dublin" instead of the "London" of the English versions. To let him present his own case:

The only "Johnny Collins" variant recovered in Great Britain comes from Hampshire; all the others were written down in the United States. In the Hampshire text we find these lines beginning the final stanza:

Those news was carried to London Town,
And wrote on London gate.

In the American texts the name of London is regularly replaced by some form of Dublin. Ballad-singers do not localize their versions in this way without reason; and the only explanation that offers itself for this particular and persistent localization is that these variants of "Johnny Collins" had been traditional in Ireland, and were imported thence into this country. It is futile, of course, to speculate about whether an individual or one of many families could first have carried the version to this country. The variants agree closely and may have been known originally to only a few immigrants. But their distribution is curiously limited. My own texts came from a small area of southwestern Pennsylvania and northwestern West Virginia, a region which in the latter eighteenth century had a considerable infusion of Scotch-Irish and Gaelic Irish settlers. More southerly variants came from Doddridge County, West Virginia (only a few miles south of the region where my texts were gathered), from Hardy and Pocohontas Counties, West Virginia, and from Highland County, Virginia. Highland, Hardy, and Pocohontas Counties are all close together and lie in the Appalachian ranges which form the western boundary of the great "Valley of Virginia," first settled by Scotch-Irish and Germans from Pennsylvania. What evidence we have, then, about the travels of this version also indicates quite strongly the likelihood of Irish tradition. It may be noted likewise that aside from having London instead of Dublin, the isolated Hampshire variant of "Johnny Collins" differs from the American texts in no noteworthy respect. [54]

In saying that "in the American texts the name of London is regularly replaced by some form of Dublin," Professor Bayard gives the impression that no examples of this ballad gathered in this country have the name "London." This is not, however, the case. At least two versions gathered outside the area delimited by him give "London" as the place-name. One collected in 1938 was published by Asher E. Treat in "Kentucky Folksong in Northern Wisconsin." [55] It is so fragmentary that it cannot be definitely included in the "Johnny Collins" group which Professor Bayard segregates from the "Giles Collins" and "George Collins" variants of the "Lady Alice," [56] but the particular stanza in question is unmistakably the same. It runs:

They sent the news to London Town
And printed on London's gate.
There six fair ladies died all in one night,
All for young Collins's sake.

The second is even more fragmentary. It was collected in North Carolina in 1918. Its last verse runs:

Haul me over to fair London town
All in fair Ellender's bown. [57]

Fragmentary as this is, it is evidently the remnant of the stanza in question, and the "Ellender" corresponds exactly to the name of the lady in Professor Bayard's printed version.

What the situation in regard to the place-name amounts to, then, is this: in "a region which in the latter eighteenth century had a considerable infusion of Scotch-Irish and Gaelic Irish settlers," the name is given as "Dublin," or a corruption of it, while in two other areas in the United States and in England, the name is "London." The only evidence of Irish tradition which can logically be argued from these facts, it seems to me, is that the ballad circulated among Irish in this country, and that among them the ballad place-name was altered from "London" to the (apparently) more acceptable name of "Dublin." If "Dublin" were the original name, then it must have been changed to "London" not only in England but also in Kentucky (or Wisconsin) and in North Carolina-an alteration which does not seem reasonable, since in those states such a change would probably have involved a local town name, rather than the alien "London." Moreover, "London" appears in versions collected earlier than those which give "Dublin"-the Hampshire one twenty-four years, the North Carolina one twelve years before. This evidence is not, of course, conclusive, but what weight it has favors "London" as the original name.

Professor Bayard seems to think that the ballad might have come over with the original Irish settlers of this Pennsylvania-West Virginia region; but there is the strong possibility that it entered that locale from English, rather than Irish, tradition, for the sources he quotes in regard to the settlement of this region state that there were English there, though fewer in number than the Irish. It seems, therefore, equally probable that if the ballad was in circulation at the time of the settling of these regions, it was introduced by the English, the change to "Dublin" being made by Irish singers there, rather than in Ireland itself. Thus the last of the contentions in support of the banshee hypothesis is found to be hardly strongly enough supported to make the case for the banshee convincing.

II
Yet if the "Colvill" mermaid was not evolved under the influence of the banshee, what did bring her into the ballad? She becomes a puzzling figure when it is realized that the only other versions of the entire international body of ballads telling this story which cast a watersprite in the role of the fatal seductress are two very fragmentary Italian versions, in which the part is played by a "laundress" (lavandera). [58] The Scandinavian ballads, to which "Clerk Colvill" seems most closely related, have, for the most part, elves or trolls, but never a mermaid or watersprite. Indeed, elves so abound in the Scandinavian versions of this ballad story that Svend Grundtvig, in his Danmarks gamle Folkeviser, has given the group the title "Elveskud" (elf-shot, or elf-blow), referring to the lethal slap, blow, or knife wound dealt the hero in most versions.

The appearance of the mermaid is all the more puzzling when we consider that the elf is a figure quite at home in the folklore of nearly all parts of the British Isles. Thomas Keightley, in his Fairy Mythology, says:

In the English language, Elf, Elves, and their derivatives are to be found in every period, from its first formation down to this present time. [59]

Keightley goes on to trace the references to elves down through English literature; but we need look no further than other traditional ballads to find ample evidence, not only that elves were familiar to the singers of the ballads, but that they were, like those in the Scandinavian versions of this particular ballad story, figures of human stature. "The Elfin Knight" (Child No. 2), "Lady Isabel and the Elf-Knight" (Child No. 4), and "Thomas Rymer" (Child No. 37) show that the ballad elves, both male and female, were not unfamiliar as beings indistinguishable from ordinary mortals. "Thomas Rymer" is especially noteworthy because the Child version of it (A) and the Child version of "Clerk Colvill" (A) both came ultimately from Mrs. Brown, of Falkland. [60] This circumstance shows that at least one person who knew the "Colvill" ballad with its mermaid was also familiar with an elf-woman comparable to those in the Scandinavian versions of this ballad.

Nor is there such a popularity of the mermaid motif in the folklore of Great Britain as to make the metamorphosis likely. Mermaids were not, of course, unknown or unsung; but, to judge from the evidence offered by collections of folklore of England, Scotland, Wales, and Ireland, the interest in elves and allied land sprites is greater and more widespread than that in mermaids. There is only one locale in the British Isles where the nature of the folklore seems such as to make it a likely place for the elf-woman to become a mermaid. That is the Shetland and Orkney Islands, a milieu which also furnishes some interesting possibilities for accounting for other details in the "Colvill" ballad.

In an article in the Journal of American Folklore by J. A. Teit on "Water- Beings in Shetlandic Folk-Lore, as Remembered by Shetlanders in British Columbia," this information is given:

It was believed that a light and beautiful land underneath the watery tastes of the ocean was inhabited by mar-men or mar-folk.... These beings were like people when in their homes; but when they were travelling through the sea, they became half-man and half-fish, their upper parts remaining manlike, while their nether parts became fishlike, or were enveloped in a fishlike covering. The women had fine features, light skins, and very long yellow hair .... They sometimes came ashore in fine moonlight nights and sat on the rocks, combing their hair .... Stories are told of mermaids being married to Shetlanders, and these narratives do not differ materially from similar ones referring to seal-folk. Sometimes the same narrative, in fact, refers to both kinds, the woman being a mermaid in one version, and a seal-woman in another. There seems to be some confusion in Shetland folk-lore between these sea-people, or mar folk and the selkie-folk [seal folk]. ... It was the current belief that seals, under certain circumstances (or at will?) could assume human form.... Many stories were told of seals coming ashore, divesting themselves of their skins, and then dancing, gamboling, and enjoying themselves in human form. On the approach of man, they rushed for their skins and ran to the sea. There are narratives also about naked sealwomen captured by men, who, unobserved, had obtained possession of their skins; without these, the women were unable to return to the sea, and were doomed to remain on land until they could recover them. As seal-folk were very comely and wellproportioned, whoever saw them in human form was almost invariably enamored of them. In certain accounts, seal men are described as having had children by daughters of men, and men are said to have married seal women. Several of these stories differ only in minor details, and relate how a man hid a seal-woman's skin and compelled her to marry him. After having had a number of children, one day she discovered the skin when her husband was away (or one of her children told her where it was concealed), whereupon she deserted her home and children and returned to the sea. [61]

Versions of the foregoing story were "at one time rife in every Orkney island," according to W. Traill Dennison in an article on "Orkney Folklore. Sea Myths." He says further:

Writers on the subject, trusting to incorrect versions of old stories, have often confounded mermaids and seals together, and have treated the two as identical. [62]

From this confusion, by the Shetlanders themselves, as well as by writers (according to Teit's article), the "Colvill" mermaid may have received her metamorphic ability, which seems to have been possessed by no other members of her species save those of the Shetlands and Orkneys. She may even originally have been a selkie (such creatures are not unknown in the Child canon) and have been changed to a mermaid in Scotland by someone to whom selkies were unfamiliar. [63]

The circumstance which makes the Shetlands and Orkneys seem a place peculiarly apt for the elf-women of the Faeroe (and other Scandinavian) versions to become the mermaid of "Clerk Colvill" is that elves are very rare in the folklore of this milieu. The collection of folklore of the Orkney and Shetland Islands in the County Folk-Lore series published by the Folk- Lore Society of London[64] contains no mention of elves, nor does Jessie M. E. Saxby's Shetland Traditional Lore (1932).[65] The widespread belief in the potency of "elf-shot," the primitive arrow-heads found occasionally by the Islanders, forms the chief use of the term "elf"; and these elf-shot were not utilized by elves but by fairies and, more especially, by "trows." Fairies were known in the lore of that region; but the most popular figure among the land sprites was the "trow," related to, and probably derived from, the Norwegian "troll."

The special feature of the Shetland-Orkney folk beliefs which is pertinent to this discussion is that trows (and fairies) are not depicted as interbreeding with the human race.[66] Matches between men and female trows, with or without benefit of clergy, are unknown-or at least unmentioned-in the recorded folklore which I have been able to examine. (The same is true of the fairies of this region.) There being no fit paramours among the land sprites (according to the folklore), the men of these islands, when inclined to traffic with supernatural creatures, found wives among the sea creatures: the selkies or Finns, or the mermaids. It seems likely, then, that when the ballad story of Olafur and the elf-women came to the Shetlands and Orkneys, brought, perhaps, by Shetland fishermen, themselvs full of lore of the sea and the seashore, [67] it found no female land sprites of fit nature to take the place of the elf-women (figures unfamiliar to the lore of this milieu), and so substituted the seal-wife or the mermaid as the fatal charmer.

The mermaid's sark may even contain a reminiscence of the selkie's sealskin; and it might be that association between the betrothal-shirt of the Scandinavian versions and the selkie's garment-like skin, the loss of which made it necessary for her to accept a human husband (and which thus, like the sark, had nuptial connotations) formed a "hook and eye" which helped fasten the mermaid upon the ballad story. It is even possible that in the lines

"It's a' for you, ye gentle knight,
My skin is whiter than the milk"

in the Version A of "Clerk Colvill," the mermaid's reference to her skin is a reminiscence of the selkie- or mermaid-skin of Shetland-Orkney lore, though this cannot, obviously, be insisted upon.

As to the "gare" or strip of sark used to dispatch the clerk, this ingenious method is unparalleled both among mermaids and among the supernatural women in other versions of this international ballad story. There is, however, a curious parallel of a sort in "The Suffolk Miracle" (Child No. 272), wherein a dead man returns to carry his unsuspecting fiancee to the grave. He sets her behind him on his horse, and they ride off "as swift as any wind." [68]

But as they did this great haste make,
He did complain his head did ake;
Her handkerchief she then took out,
And tyed the same his head about. [69]

And when subsequently the grave was exhumed and the handkerchief discovered about the head of the mouldering corpse,
She was thereat so terrified
And grievd, she quickly after dyed.[70]

Professor L. C. Wimberly, in Folklore in the English and Scottish Ballads, remarks:

... "The Suffolk Miracle" ... illustrates possibly the same species of magic which we found to be exemplified in "Clerk Colvill" when the mermaid's garment, or a "gare" of it, seals the lover's fate. The ... piece ... tells of a maiden who binds her handkerchief about the aching head of her dead lover, and afterward dies.... Moreover, as the story goes, the maiden's death seems to be motivated by terror and grief, but it is probable that this is an instance of rationalization, and that originally the maiden, by virtue of a kind of sympathetic magic, owes her death to the fact that the ghost has gained possession of her handkerchief.... For one to leave an article of his clothing... in the possession of the dead is commonly held to be very dangerous. [71]

In another article, referring to this same likeness between the two ballads, Professor Wimberly admits:

It is true that the situation in one song is somewhat the reverse of that in the other, but the general idea underlying the two incidents is doubtless the same. [72]

That may be so; but it is noteworthy that in "The Suffolk Miracle" the mortal lets part of her clothing come into possession of a supernatural creature; whereas in "Clerk Colvill" it is part of the garment of a supernatural creature which does the deadly work.

Another possible source of the motif of the lethal power of the strip of sark may lie in the superstitions concerning the curative powers of "holy wells," superstitions rife enough in Scotland around 1600 to be recorded in witchcraft trials. One such superstition, involved in one of the trials recorded in Robert Pitcairn's Ancient Criminal Trials of Scotland, was that a wet sark put upon a sick person would cure disease. [73] Another superstition is recorded in another trial in the same work, wherein, in 1623, one Issobell Haldane confessed that

. . . scho went, silent, to the Well of Ruthven, and returneit silent, bringing watter from thence, to wasch Johne Gowis bairne: Quhen scho tuik the watter from the Well, scho left an pairt of the bairnes sark at it, quhilk scho tuikwith hir for that effect; . . .[74]

And a note to this confession explains:

Some shred of dress belonging to the diseased person was left on the brink of the well, in token of the disease having passed away with it. [75]

Regarding this superstition, Rev. Walter Gregor, in Notes on the Folk-Lore of the North-East of Scotland, adds:

No one would have been foolhardy enough to have even touched what had been left, far less to have carried it off .... Whoever carried off one of such relics contracted the disease of the one who left it. [76]

It seems but a short step from this tenet of witchcraft-that a strip of garment (often a sark) left at a magic well could communicate disease (usually mortal)- to the ballad lore that a strip of the sark of the mermaid at the well could cause the painful death of the clerk. The connection, of course, cannot be proved; but the superstition is known to have existed in the locale where the ballad was collected.

It may be objected that in positing the existence of the "Elveskud" ballad in the Shetland-Orkney locale, where no trace of it has been recovered, I am putting myself in the same position as that of Professor Bayard, when he says that "it seems practically certain that the American variants of ["Johnny Collins"] have been traditional in Ireland," without pointing to any gathered there. Fortunately, however, there is the precedent of Svend Grundtvig, who remarks, concerning "Clerk Colvill":

It is not improbable that this Scottish ballad-form has wandered in from the northern islands (Shetland or Orkney Islands). Like the Faeroese and Icelandic forms, it alone has the tale's first part, and together with them the Scottish ballad has also the killing kiss, which is found in Lewis's text, as in the Faeroese and Icelandic; the ballad's preamble agrees with the Faeroese, and it is even quite possible that we here find again the common Norse name for the hero, just as in the Faeroese-Icelandic form, in English disguise. Just as Herr Raddengaardin No. 13 is in English form changed to Sir Aldingar, in this manner Olavur could well become [Clerk] Colvill. [77]

Child, too, notes that" . . the Faroe versions are nearest to the English."[78] And if it is granted that the strongest relationship between the Scottish and the Scandinavian versions is shown by the Faeroe ballads, then it is justifiable to follow Grundtvig's suggestion and look to the Shetland-Orkney locale for a milieu wherein the mermaid might have usurped the place of the elf-woman. Indeed, this locale where the Scottish and Scandinavian tongues existed side by side for several centuries would seem the logical place for the ballad to leap the language barrier into Scottish tradition. Hugh Marwick, in the Introduction to The Orkney Norn states that " . . throughout the seventeenth century Orcadians must have been in large measure bilingual,"[79] and that " . . . after weighing all the evidence, it would appear certain that up to the end of the sixteenth century, at least, the Norn speech was still the usual language used by most, if not all, native Orcadians."[80] The tenacity of this Scandinavian language was even more marked in the remoter Shetlands.

Jakob Jakobsen, in his Etymological Dictionary of the Norn Language in Shetland, says that " . . . Lowland Scottish cannot have been spoken in Shetland generally, instead of Norn, so early as I700."[81] He continues: Even in 1600 the knowledge of English (Lowland Scottish) seems to have been very meager in Shetland .... That the Shetland Norn was still a living language in the middle of the 18th century, one may conclude from what is said by the Scottish writers George Low and Samuel Hibbert about the Shetland dance songs. Even rather late in the 18th century, Norn songs and ballads survived in the mouths of the common people, and were sung as the music to the native dance, which was the same as, or somewhat similar to, the chain-dance in a circle, still popular in the Faroe Isles. . . .[82]

However (to quote Jakobsen further): As Norn expired before it was ever scientifically examined and studied, the continuous specimens of it which have been hitherto recorded for us are very few... The fragments of Norn phrases and sentences which have been preserved in the Shetlandic dialect are small fragments of verse, nursery rhymes, fairy rhymes, riddles, a few proverbs, and some fragments of conversation. [83]

J. A. Teit, in the article on "Water-Beings in Shetlandic Folk-Lore . .", quoted above, makes the fact of the existence and subsequent loss of balladmaterial in the Shetlands even more emphatic. He says:

It has often been pointed out that two or three generations ago the Shetland Islands would have yielded a rich harvest to the folk-lorist. This would have proved almost equally true only a generation ago.... Many writers have in recent years collected
and published much of the folklore of the Islands, but no systematic and adequate efforts have been made in this direction. Thus, no doubt, much has either been lost or remains unrecorded. This is particularly true, I believe, of the traditional narratives and of practices of various kinds.... Increasing contact with Scotland since about the beginning of the seventeenth century has brought in new elements from the south. When the Scottish or English language became generally understood, this increment seems to have consisted almost entirely of ballads, folk-songs, and certain sayings and proverbs.... As the ancient Norse language gradually fell into disuse, during the eighteenth century, the popular old Norse sagas, ballads, and songs were superseded almost entirely by those of Lowland Scottish and English origin. Only fragments in partly corrupted forms of the Norse language have been handed down to the present day, while some others with native Norse themes appear in English garb. [84]

Might not "Clerk Colvill" be one of these ballads with native Norse themes appearing in English garb, lost in the Shetland-Orkney tradition, but preserved in the Scottish? The "Colvill" ballad had been committed to writing by a date when ballads in Norn can, according to the above-quoted authorities, be supposed to have been still current in the Shetlands, if not in the Orkneys. It seems quite possible that the "Elveskud" ballad could have been among the Visecks which shortly before 1774 "formed the accompaniment to dances that would amuse a festival party during a long winter's evening." [85]

These dances, according to Jakobsen, as quoted above, were similar to the Faeroe dances, which were likewise accompanied by the singing of ballads. And the Faeroe version (A) of the "Elveskud" group gives strong indication of being a dance-ballad in its refrain, which says, in part, "Get out on the floor; dance merrily." [86] It seems possible, then, that this ballad might have been part of the lost Norn dance-ballad repertory, and that somewhere in the bilingual milieu of the Shetland-Orkney Islands, the story was transferred into Scottish and thus survived in Scottish tradition the demise of the Norn analogues.

It is interesting that from this same locale can be supplied a possible historical source of the name "Clerk Colvill," in the person of Harry Colvile, Parson of Orphir (in Pomona, largest of the Orkneys), who was murdered in the Shetlands in 1596. He was a "Colvill"; he was a "clerk" in the double sense of being both a parson and a secretary to the Earl of Orkney. He was somewhat involved with the supernatural world-at least there was opportunity for such implication to arise in popular fancy-by virtue of his being the prosecutor of a (supposed) witch in a witchcraft trial. And he died a violent death, possibly as an outcome of his activities in this trial.

It is, admittedly, entirely possible that Harry Colvile had no connection with "Clerk Colvill" in the ballad. Grundtvig points out, as quoted above, the possibility that the name "Colvill" was derived from the "Olafur" of the Faeroe and Icelandic versibns, the "k" being elided with "Olvill" corrupted from "Olafur." It would then be necessary to account for the title "clerk," in place of the "sir" common to most of the Scandinavian versions. The title "herr" had dropped out of the Icelandic and Faeroe versions, leaving room for the substitution of a new one. The "clerk" might have come in through mere confusion, as, for example, in "Lord Thomas and Lady Margaret" (Child No. 260), in Version B of which the hero is called "Clerk Tammas."

(In this connection it should be noted that in Version A of the "Colvill" ballad the mermaid addresses Clerk Colvill as "ye gentle knight," an indication of confusion.) These considerations do not, however, obviate the possibility that the career of Harry Colvile influenced the form of the name and the title "clerk" in the Scottish version.

It is difficult to fit into an intelligible pattern the circumstances which appear to be reflected in the various legal documents connected with the murder of Harry Colvile which are recorded in Pitcairn's Ancient Criminal Trials of Scotland. Various details suggest, however, that Parson Colvile, a minion for Patrick, Earl of Orkney, tried on the Earl's behalf to get evidence against John Patrick, Master of Orkney, the Earl's brother, implicating John in an attempt on the Earl's life by witchcraft; it was evidently Colvile who interrogated one Alisoun Balfour under torture and wrung from her a confession implicating John. [87] John Stewart was evidently to have been indicted for "airt and pairt" of the murder of Colvile, though the prosecution of him, according to Pitcairn's records, seems not to have proceeded further than declaring him an outlaw for failing to appear for trial. [88] The man who paid the penalty for the murder was one Gylbert Pacok, who had been one of the witnesses of Alisoun Balfour's death at the stake, at which time she repudiated the confession wrung from her by torture. [89] According to the "Dittay" or indictment against William Bannatyne and James Lokie, who were acquitted, thirty persons were in the band which sailed from Scotland to the Orkneys and thence to the main island of Shetland, where they hunted Harry down. [90]

A century later, the Reverend John Brand found current in the Shetlands an account of the murder which named the killers as four brothers named Sinclar, and which attributed the crime to anger on the part of the populace of the Orkneys at Colvile's aiding and abetting the Earl in his program of oppression and cruelty. [91] The difference between this account and that which can be pieced together from Pitcairn's records suggests that among the uneducated folk of the islands-Brand presumably had his converse with ministers, since he was part of a commission sent to the presbyteries of Shetland, Orkney and Caithness-the account might have been romanticized to a greater extent, and perhaps to such degree that Parson Colvile's violent death and his connection with the witchcraft trial associated his name with the hero of the ballad. This is, of course, pure surmise; but it is not outside the bounds of possibility.

One final, minor item: from the place-names of this Shetland-Orkney locale can be furnished a possible explanation of the "Wall of Stream" which the mermaid inhabits in Version A of "Clerk Colvill." There is a "Loch o' Strom" on Mainland, the largest of the Shetland Islands; and, according to Jakobsen's Etymological Dictionary, quoted above, the word "strom" means "stream." [92] The exchange of "well" for "loch" would not be difficult, since water-sprites were thought to inhabit both, and the substitution might easily have taken place in Scotland if it had not already been effected in the Shetland-Orkney locale.

For the mermaid herself, and for a likely place in which the ballad story might make the change from Scandinavian to Scottish tradition, and perhaps even for the name "Clerk Colvill," I think we probably need look only to this Scandinavian-Scottish milieu. At any rate, I believe there is no need to look to Gaelic or Irish tradition for explanation of details either in "Clerk Colvill" or "Johnny Collins."

University of California,
Berkeley, California.
_____________

Footnotes:

1. p. 99.
2 p. 101.
3 P. 75.
4 P. 78.
5 P. 100.
6 Francis J. Child, English and Scottish Popular Ballads, No. 42, Version C (Vol. I, p. 389).
7 Bayard, op. cit., p. 88.
8 Ibid., p. 98.
9 Italics are Professor Bayard's.
10 Bayard, op. cit., p. 99.
11 Ibid., p. I00.
12 The earliest printed specimen of this group which Professor Bayard has dubbed the "Johnny Collins" variants (see p. 75 of his article). It appeared in the Journal of the Folk-Song Society (London), 3 (1909), 298, having been collected by the contributor, Dr. G. P. Gardiner, in I906.
13 Journal of the Folk-Song Society (London), 4 (I9I0), p. Io6.
14 Bayard, op. cit., p. 89.
15 R. C. Alexander Prior, Ancient Danish Ballads Translated from the Originals (London:
Williams and Norgate, I860), 3 vols. II, 306. (A translation of the Danish Version B No. 47,
in Svend Grundtvig's Danmarks gamle Folkeviser [Kjobenhavn: Thieles Bogtrykkeri, i856],
II, 114.)
16 Danish B, I, T, X, Y of No. 47 in Grundtvig, op. cit., II, II2 and IV, 843-846.
17 Landstad No. 40 and variant. (M. B. Landstad, Norske Folkeviser [Christiania: Chr. Tonsbergs Forlag, 1853], pp. 356, 843).
18. B. T.
19 Knut Liestol and Moltke Moe, Norshe Folkevisor (Kristiania: Jacob Dybwads Forlag, 1924), 3 vols. III, 178.
20 Op. cit., p. 87.
21 Stanza i.
22 Stanza 5.
23 Child: op. cit., I, 313. "Allison Gross" is from the Jamieson Brown MS, while Version A
of "Clerk Colvill" is from William Tytler's Brown MS.
24 Op. cit., p. 97.
25 Ibid., pp. 95-96.
26 Elizabeth Taylor, a note communicated from the Faeroe Islands, January 29, 1902. Folk-Lore, 13 (I902), I85.
27 John Nicholson, "Some Old-Time Shetlandic Customs," Orkney & Shetland Miscellany (Publications of the Viking Club, Vol. 5, 1912), p. 124.
28 Walter Gregor, Notes on the Folk-Lore of the North-East of Scotland (Publications of the Folk-Lore Society, Vol. 7, London, 1881), p. 85.
29 William Henderson, Notes on the Folk-Lore of the Northern C ounties of England and the Borders (London: Longmans, Green & Co., I866), p. 79.
30 Op. cit., p. 78, n. I5.
31 Ibid., p. 80, n. 23.
3 Both the first and second versions of this variant give this line in the second stanza.
33 John Harrington Cox, Folk-Songs of the South (Cambridge: Harvard Univ. Press, 1929),
p. IIo.
34 Ibid.
35 Arthur Kyle Davis, Jr., Traditional Ballads of Virginia (Cambridge: Harvard Univ. Press, 1929), p. 346.
36. Cox, op. cit., p. 355.
37 Ibid., p. IIo.
38 Ibid., p. 354.
39 Cecil J. Sharp, English Folk Songs from the Southern Appalachians (London: Oxford
Univ. Press, 1932), 2 vols. II, 84.
40 Davis, op. cit., p. I64.
41 Bayard: op. cit., p. 76, n. 8.
42 Moreover, the "screaming" and "crying" commonly attributed to the maiden in these ballads are not clearly analogous to the mourning of the Gaelic banshee. According to John Gregorson Campbell's Superstitions of the Highlands and Islands of Scotland (Glasgow: James MacLehose, I900, p. 41), "The songof the Fairy woman foreboded great calamity, and men did
not like to hear it. Scott calls it 'The fatal Banshi's boding scream,' but it was not a scream, only a wailing murmur ... of unearthly sweetness and melancholy."
43 Bayard, op. cit., p. Ioo.
44 Second version, stanza 2.
45 Grundtvig, op. cit., II, 114, stanza 12. (Translation mine.)
46 Op. cit., p. I02.
47 Liestol and Moe, op. cit., III, i6 stanza 4. (Translation mine.)
4S Grundtvig, op. cit., II, x16, stanza 14. (Translation mine.)
49 Op. cit., p. 99.
50 Ibid., p. 94.
51 Ibid., p. 84.
52 Ibid., p. 101.
53 Ibid., p. 100.
54 Ibid., p. 98.
55 Journal of American Folklore, 52 (1939), 47.
56 Bayard, op. cit., pp. 74-75.
57 Sharp, op. cit., I, I99.
58 Child, op. cit., I, 382.
59 The Fairy Mythology, Illustrative of the Romance and Superstition of Various Countries (London: Bohn, 1850), p. 63.
60 Child, op. cit., I, 323, 37I. ("Thomas Rymer," Version A is from Alexander Fraser Tytler's Brown MS; "Clerk Colvill" is from William Tytler's Brown MS.)
61 Journal of American Folklore, 31 (I918), I88-190.
62 Quoted in G. F. Black, Examples of Printed Folk-Lore concerning the Orkney and Shetland Islands ("County Folk-Lore," Vol. 3) (Publications of the Folk-Lore Society, Vol. 49, London:
David Nutt, 1903), pp. I88-I90.
63 Selkies were also part of the folklore of the Faeroe Islands, where substantially the same
story is told of them; and it is possible that the mermaid, either in propria persona or in the
conjectural intermediate form of a selkie, could have entered the "Elveskud" ballad there. But
this is unlikely, since the versions collected as late as the I870's still persist in having elves.
64 Black, loc. cit.
65 (Edinburgh: Grant & Murray, 1932.)
66 The rare instance is that of the "Kunal-Trow" or king trow. There are no female Kunal- Trows, so that these creatures must take human wives to propagate their species. But such matches are not eagerly sought on either side; for the human mother dies as soon as the baby trow is born, and the father dies as soon as his trow son comes of age. (Saxby, op. cit., p. 128.)
67 Hugh MacDiarmid, The Islands of Scotland: Hebrides, Orkneys, and Shetlands (London:
Batsford, 1939), p. 2.
68 Stanza 13.
69 Stanza 14.
70 Stanza 27.
71 (Chicago: Univ. of Chicago Press, 1928), p. 286.
72 University of Nebraska Studies in Language, Literature and Criticism, No. 8, Lincoln,
Nebraska, 1927, p. 78.
73 (Edinburgh: Bannatyne Club, 1883), 3 vols. II, Part I, p. 27.
74 Ibid., p. 537.
75 Ibid.
76 Op. it., p. 40.
77 Op. cit., IV, 865. (Translation mine.)
78 Op. Cit., I, 374.
79 (Oxford: Oxford Univ. Press, I929), p. xxvi.
80 Ibid., p. xxiv.
81 (London: 1928), 2 vols. I, xvi.
82 Ibid., p. xvii.
83 Ibid., p. xci.
84 Op. cit., pp. 80o-i82.
85 Jakobsen, op. cit., I, xvii.
86 Grundtvig, op. cit., IV, 849. (Translation mine.)
87 Pitcairn, op. cit., I, Part 3, p. 377.
88 Ibid., p. 393.
89 Ibid., pp. 386, 376.
90 Ibid., p. 393.

91 As given in John Pinkerton, A General Collection of the Best and Most Interesting Voyages and Travels in All Parts of the World (Qondon: Longman, Hurst, etc., 1809), 6 vols. III, 790. (Brand states that the crime was allegedly committed by "four brethren of the name of Sinclar." Two "Sinclaris" are stated to have been among the accomplices of John Stewart in conspiring to murder Earl Patrick (Pitcairn: op. cit., p. 374). No other Sinclairs are mentioned in Pitcairn's records connected with the Colvile murder. Perhaps these two represent a link between John Stewart and that crime.)

92 Op. cit., II 906