Folk-Lore/Volume 4/Cinderella in Britain

From Wikisource
Jump to navigation Jump to search


Vol IV.]
[No. III.


THE first word anyone interested in folk-tales must say about Miss Roalfe Cox's remarkable volume of variants of Cinderella is one of congratulation. Her industry is scarcely more conspicuous than her taste. It required both tact and knowledge to pick out in the more elaborate analyses of the tales just those points of the original that deserved particular attention, and at every stage Miss Cox has shown that knowledge and tact. Then again, Miss Cox has obviously kept herself free from any parti pris, and her collection is thus absolutely and scientifically impartial in its tone and arrangement. We of the Folk-lore Society required a collection of variants of a single folk-tale "radicle" that should be tolerably complete, absolutely impartial, and conveniently arranged. We have got it.

I cannot say that we are altogether happy now that we have got our ideal collection. In the first place, it has become clear that some international plan must be arrived at for such a collection. It is impossible for a single person, however loyally assisted, as was Miss Cox, to cope with a problem which is essentially international. Even for the British Isles, Miss Cox has failed, as we shall see, in making her collection exhaustive of matter already printed, while the remarkable variant contributed at the last moment by Mr. Macleod (Cinderella, p. 534) will serve to show what rich harvests still remain to be gleaned from the folk-memory.

Again, Cinderella has proved not so desirable a choice for the exercise of Miss Cox's industry and skill as might have been desired. Among the most pressing problems that we should like to solve by means of such a collection are: (1) Has there been continued existence of folktales from pre-historic times to the present? (2) Are folktales with "savage" elements necessarily prior to the same without those elements, or have those elements been introduced? (3) Is India the sole or chief source of folktelling?

Now with regard to (1), Cinderella does not happen to be a good type of story to be used as a test. The essence of the tale is the rise in social position of a girl who makes a fortunate marriage. Possibly there are such cases in savage or in pre-historic societies; but the whole conception strikes one as mediaeval, almost as feudal. It would therefore be idle to look for its origin in societies where there was little variation of social position. Dr. Westermarck has indeed shown that girls have more freedom of choice in savage or semi-savage society than we had previously thought. But the monogamous condition which is at the root of the slipper-test does away with the probability that Cinderella arose in any but a tolerably advanced state of civilisation, and consequently its variants do not form a good subject for dealing with, or deciding our first question as to the comparative age and longevity of fairy tales.

Then as regards the vexed question of an Indian origin, Cinderella is specially unfortunate as a test case, since India is essentially a shoeless country, and the characteristic incident of the tale in its present form is the shoe test. We need not therefore be surprised that Miss Cox's collection gives a negative result as regards India. I, for one, have never contended that all fairy tales come from India; and M, Cosquin, in a private communication to me, points out that he has likewise guarded himself from any assertion of the exclusive Indian origin of folk-tales. I am quite prepared to admit the possibility of India borrowing from Europe, and the locale and character of the three Indian variants (Nos. 25, 235, 307) are sufficient to show the probability of such borrowing in the case of Cinderella. Miss Frere's collection was mainly from an ayah from Goa, whose family had been Christian for several generations; Salsette has long been open to European influence, and so has Bombay.

With regard, however, to the important methodological problem which I have placed second above, Miss Cox's collection has much instruction to give. The very fact that in its inception Cinderella, as we now have it, cannot have arisen in a savage stage of society, renders it certain that the "savage" elements in certain forms of it—animal parentage, dead-mother aid, bones together, and the like—may have been introduced into the story after it had obtained currency, or, if in the original form, may have been introduced as conventional episodes of the folk-tale which had a far more remote origin. The archaeological value of such incidents is accordingly much reduced by such considerations.

One thing, however, comes out quite clearly from Miss Cox's labours, and as it is a thing on which I have insisted throughout my own folk-tale studies, I am naturally jubilant over the result. Here we have 133 variants of type A—the Cinderella type pure and simple—scattered over all the lands of civilisation. Yet no one, I take it, would be prepared to contend that any single one of these was independently created, and was without relationship, cognate or agnate, to any one of the rest. The Borrowing Theory of explaining the similarities in folk-tale plots comes out triumphant as the sole working hypothesis that will explain the same story existing in so many lands. That in this particular case the borrowing is not from India does not affect the general question.

When, however, we come to the question who originated and who borrowed, we come to the problem of problems and the further research to which Miss Cox's labours lead us: hic labor, hoc opus. It would require more time than I could devote to the subject at present, more ingenuity than I could bring to bear on it at any time, to arrive at even an approximate solution of this intricate question. It is, in fact, a case for a European Concert, as indeed Miss Cox's book shows. The folk-lorists of each country might be called upon to determine from their local knowledge and further collections what was the original form in the particular country, and then our problem would be reduced to its simplest elements. We should perhaps be able to determine which was the original form of the tale, and where it exists at the present day in a form closest to the original. Whether this locality could then be fixed upon as the original home of the story would then have to be determined by various criteria. All this, however, is in the future, though, thanks to Miss Cox, it may be no distant future; for the present we may content ourselves with the first reduction of the problem so far as it relates to the British Isles. In other words, what was the original form in which the three types of story dealt with in Miss Cox's book—Cinderella, Catskin, Cap o' Rushes—appeared in these islands?

Before doing so, however, I would venture to point out one aspect of our subject which lends it considerable importance. We have to deal here with various versions of a series of incidents preserved by tradition and reduced to writing after many days. Now this, to compare great things with small, is exactly the problem of the Synoptic Gospels. It is not by any means improbable that folk-tale research, by arriving at the laws governing the transmission of narratives by tradition, may ultimately come to the aid of theological science in determining the relative age of the gospels and settling the amount and character of the alterations undergone by the narratives during the process of tradition. But this is a digression, and we must again turn to the particular case of folk-tradition we have before us in the diffusion of tales of the Cinderella type through Great Britain and Ireland.

A. "Cinderella, or the Fortunate Marriage of a despised Scullery-maid by aid of an Animal Godmother through the Test of a Slipper"—such might be the explanatory title of a chap-book dealing with the pure type of Cinderella. This is represented in Miss Cox's book, so far as the British Isles are concerned, by no less than seven variants, as follows:

(1) Dr. Blind, in Archæological Review, iii, 24-7, "Ashpitel" (from neighbourhood of Glasgow). (2) A. Lang, in Revue Celtique, t. iii, reprinted in Folk-Lore, September 1890, "Rashin Coatie" (from Morayshire). (3) Mr. Gregor, in Folk-Lore Journal, ii, 72-4 (from Aberdeenshire), "The Red Calf"—all these in Lowland Scots. (4) Campbell, Popular Tales, No. XLIII, ii, 286 seq., "The Sharp Grey Sheep." (5) Mr. Sinclair, in Celtic Mag., xiii, 454-65, "Snow-white Maiden." (6) Mr. Macleod's variant communicated through Mr. Nutt to Miss Cox's volume, p. 534; and (7) Curtin, Myths of Ireland, pp. 78-92, "Fair, Brown, and Trembling"—these four in Gaelic, the last in Erse. To these I would add (8, 9) Chambers' two versions in Pop. Rhymes of Scotland, pp. 66-8, "Rashie Coat," though Miss Cox assimilates them to Type B. Catskin; and (10) a variant of Dr. Blind's version, unknown to Miss Cox, but given in 7 Notes and Queries, xi, 461.

Now in going over these various versions, the first and perhaps most striking thing that comes out is the substantial agreement of the variants in each language. The English, i.e., Scotch, variants go together; the Gaelic ones agree to differ from the English. I can best display this important agreement and difference by the accompanying two tables, which give, in parallel columns. Miss Cox's abstracts of her tabulations, in which each incident is shortly given in technical phraseology. These abstracts have proved fully as useful and valuable as I anticipated in recommending them: it is practically impossible to use the long tabulations for comparative purposes without some such shorthand. For the purpose of our inquiry we will find it more convenient to arrange the incidents vertically, and not, as in Miss Cox's book, finish the tabulation of one story before beginning that of another. By this means we are enabled to display parallelism graphically.


Gregor. Lang. Chambers, I. and II.[1] Blind.
Ill-treated heroine (by parents). Calf given by dying mother. Heroine dislikes husband.[2] Ill-treated heroine (by stepmother).
Helpful animal (red calf). Ill-treated heroine (by stepmother and sisters). Hen wife aid. Menial heroine.
Spy on heroine. Heroine disguise (rashin coatie). Countertasks. Helpful animal (black sheep).
Slaying of helpful animal threatened. Hearth abode. Heroine disguise. Ear cornucopia.
Heroine flight. Helpful animal. Heroine flight. Spy on heroine.
Heroine disguise (rashin coatie). Slaying of helpful animal. Menial heroine. Slaying of helpful animal.
Menial heroine. Revivified bones.
Help at grave.
Dinner cooked (by helpful animal).
(Fairy) aid. Old woman advice.
Revivified bones.
Task-performing animal.
Magic dresses (given by calf). Magic dresses. Magic dresses. Meeting-place (church).
Meeting-place (church). Meeting-place (church). Meeting-place (church). Dresses (not magic).
Flight. Flight threefold. Flight threefold. Flight twofold.
Lost shoe. Lost shoe. Lost shoe. Lost shoe.
Shoe marriage test. Shoe marriage test. Shoe marriage test. Shoe marriage test.
Mutilated foot (Housewife's daugh.) Mutilated foot. Mutilated foot. Mutilated foot.
Bird witness False bride. False bride. False bride.
Happy marriage. Bird witness Bird witness Bird witness (raven).
House for red calf. Happy marriage. Happy marriage. Happy marriage.


Macleod. Campbell. Sinclair. Curtin.
Heroine, daughter of sheep, king's wife. Ill-treated heroine (by stepmother). Ill-treated heroine (by stepmother and sisters). Ill-treated heroine (by elder sisters).
Spy on heroine. Menial heroine. Menial heroine. Menial heroine.
  Helpful animal. Helpful cantrips. Henwife aid.
  Spy on heroine. Magic dresses. (+ starlings on shoulders). Magic dresses.(honey-bird, finger and stud).
Revivified bones. Revivified bones. Lost shoe. Lost shoe.
  Stepsister substitute. Shoe marriage test. Shoe marriage test.
Magic dresses. Golden shoe gift (from hero). Heroine under washtub. Mutilated foot.
Meeting-place (feast). Meeting-place (sermon). Happy marriage Happy marriage
Flight threefold. Flight threefold. Substituted bride Substituted bride (eldest sister).
Lost shoe (golden). Lost shoe. Jonah heroine. Jonah heroine.
Shoe marriage test. Shoe marriage test. Three reappearances. Three reappearances.
Mutilated foot. Mutilated foot. Reunion Reunion
  False bride.   Villain Nemesis.
Bird witness. Bird witness.    
Happy marriage. Happy marriage.    

Now in the "English" versions there is practical unanimity in the concluding portions of the tale. Magic dresses—Meeting-place {Church)—Flight—Lost shoe—Shoe marriage-test—Mutilated foot—False bride—Bird witness—Happy marriage, follow one another with exemplary regularity in all four (six) versions.[3] The introductory incidents vary somewhat. Chambers has evidently a maimed version of the introduction of Catskin. The remaining three enable us, however, to restore with some confidence the Ur-Cinderella in English, somewhat as follows: Helpful animal given by dying mother—Ill-treated heroine—Mental heroine—Ear cornucopia—Spy on heroine—Slaying by helpful animal—Tasks—Revivified bones. I have attempted to reconstruct the "English" Cinderella according to this formula in my forthcoming More English Fairy Tales. It will be observed that the helpful animal is helpful in two ways—(a) in helping the heroine to perform tasks; (b) in providing her with magic dresses. It is the same with the Grimms' Aschenputtel and other Continental variants.

Turning to the Celtic variants, these divide into two sets. Campbell's and Macleod's versions are practically at one with the English formula, the latter with an important variation which will concern us later. But the other two, Curtin's and Sinclair's, one collected in Ireland and the other in Scotland, both continue the formula with the conclusion of the Sea Maiden tale (on which see the notes of my Celtic Fairy Tales, No. XVII). This is a specifically Celtic formula, and would seem therefore to claim Cinderella for the Celts. But the welding of the Sea Maiden ending on to the Cinderella formula is clearly a later and inartistic junction, and implies rather imperfect assimilation of the Cinderella formula. To determine the question of origin we must turn to the purer type given by the other two Celtic versions.

Campbell's tale can clearly lay no claim to represent the original type of Cinderella. The golden shoes are a gift of the hero to the heroine which destroys the whole point of the Shoe marriage-test, and cannot have been in the original, wherever it originated. Mr. Macleod's version, however, contains an incident which seems to bring us nearer to the original form than any version contained in Miss Cox's book. Throughout the variants it will be observed what an important function is played by the helpful animal. This in some of the versions is left as a legacy by the heroine's dying mother. But in Mr. Macleod's version the helpful animal, a sheep, is the heroine's mother herself! This is indeed an archaic touch which seems to hark back to primitive times and totemistic beliefs. And more important still, it is a touch which vitalises the other variants in which the helpful animal is rather dragged in by the horns. Mr. Nutt's lucky find at the last moment seems to throw more light on the origin of the tale than almost the whole of the remaining collection.

But does this find necessarily prove an original Celtic origin for Cinderella? Scarcely. It remains to be proved that this introductory part of the story with helpful animal was necessarily part of the original. Having regard to the feudal character underlying the whole conception, it remains possible that the earlier part was ingeniously dovetailed on to the latter from some pre-existing and more archaic tale, perhaps that represented by the Grimms' "One Eye, Two Eyes and Three Eyes". The possibility of the introduction of an archaic formula which had become a convention of folk-telling cannot be left out of account when we consider our next type.

B. "Catskin, or the wandering Gentlewomen", now exists in English only in two chapbook ballads. But, as can be seen above, Chambers' first variant of Cinderella begins with the Catskin formula in a euphonised form. The full formula may be said to run, in abbreviated form—Death-bed promise—Deceased wife's resemblance marriage test—Unnatural father (desiring to marry his own daughter)—Helpful animal—Counter-tasks—Magic dresses—Heroine flight—Heroine disguise—Menial heroine—Meeting-place—Token objects named—Threefold flight—Love-sick prince—Recognition ring—Happy marriage. Of these the chapbook versions contain scarcely anything of the opening motifs. Yet they existed in England, for Miss Isabella Barclay, in a variant which Miss Cox has overlooked (Folk-Lore, i, App., p. 149), remembers having heard the Unnatural Father incident from a Cornish servant-girl. Campbell's two versions also contain the incident from which one of them receives its name. One wonders in what form Mr. Burchell knew Catskin, for "he gave the [Primrose] children the Buck of Beverland[4] with the History of Patient Grissel, the adventures of Catskin and the Fair Rosamond's Bower" (Vicar of Wakefield, 1766, c. vi). Pity that "Goldy" did not tell the story himself as he had probably heard it in Ireland, where Kennedy gives a poor version in his Fireside Stories.

Yet, imperfect as the chapbook versions are, they yet retain not a few archaic touches. It is clear from them at any rate that the Heroine was at one time transformed into a Cat. For when the basin of water is thrown in her face she "shakes her ears" just as a cat would. Again, before putting on her magic dresses she bathes in a pellucid pool. Now Prof. Child has pointed out in his notes on Tamlane and elsewhere (English and Scotch Ballads, i, 338; ii, 505; iii, 505) that dipping into water or milk is necessary before transformation can take place. It is clear, therefore, that Catskin was originally transformed into an animal by the spirit of her mother, also transformed into an animal.

If I understand Mr. Nutt rightly (Folk-Lore, iv, 133, seq), he is inclined to think, from the evidence of the hero-tales which have the unsavoury motif of the Unnatural Father, that the original home of the story was England, where most of the hero-tales locate the incident. I would merely remark on this that there are only very slight traces of the story in these islands nowadays, while it abounds in Italy, which possesses one almost perfect version of the formula (Miss Cox, No. 142, from Sardinia). It is at any rate an interesting result of the abstract analysis of the story that the whole has to be printed in Clarendon type as being entirely composed of the formula.

Mr. Newell, on the other hand (American Folk-Lore Journal, vi, 160), considers Catskin the earliest of the three types contained in Miss Cox's book, and considers that Cinderella was derived from this as a softening of the original. His chief reason appears to be the earlier appearance of Catskin in Straparola,[5] 1550, a hundred years earlier than Cinderella in Basile, 1636. This appears to be a somewhat insufficient basis for such a conclusion. Nor is there, after all, so close a relation between the two types in their full development as to necessitate the derivation of one from the other.

C. Cap o' Rushes is chiefly of interest as being similar to King Lear. Mr. Newell, l. c., suggests a direct relationship. Catskin, according to him, is derived from Godfrey of Monmouth. But the "loving like salt" formula (for which see Cosquin, i, 288) has a distinct folk-flavour about it, and I think it more likely that both Godfrey and Cap o' Rushes are derived from an English, perhaps British, folk-tale.

D, "Tattercoats," the original of which will appear in my forthcoming book, is of interest chiefly as being without any "fairy" or supernatural elements, unless the Herd-boy with his persuasive pipe be regarded as such an element. It is practically a prose variant of "King Cophetua and the Beggar Maid", and is thus an instance of the folknovel pure and simple, without any admixture of those unnatural incidents which transform the folk-novel into the serious folk-tale as we are accustomed to have it. Which is the prior, folk-novel or tale, it would be hard to say.

Our inquiries into the various forms of Cinderella and kindred types which have been observed by Miss Cox in Great Britain and Ireland have not led to any definite result, a result not to be wondered at. What is required is that similar investigation should be made for each country or linguistic area, with a view of ascertaining the earliest and most original form of each type in each country. We shall then be in a position, perhaps, to say where the story originated and how it got transmitted to other places.

And now a few words pro domo. Mr. Andrew Lang, in his Preface to Miss Cox's volume, has done me the honour of replying to some remarks of mine on his views, read before the International Folk-lore Congress, and published in its Transactions, pp. 76-86. I there "went for", with as cunning a mixture of vigour and courtesy as I could command, the view that the resemblances in folk-tales of distant countries is due to casual similarity arising independently, owing to the similarity of minds in a primitive stage. I was all for the resemblances having arisen in the most natural way, by nations borrowing one from the other: the other view seemed to me to overlook the improbability on the doctrine of chances of a complicated series of incidents occurring independently and casually in several localities. Thus, a story of twelve incidents could only occur casually with the same order of incidents in two different places once in 479,001,599 times; in other words, it is, roughly speaking, five hundred millions to one against its thus occurring alike by chance[6] in two different places. One does not want any greater certainty than that to be against the Casual Theory of the resemblances in folk-tales, and I therefore protested as vigorously as I could against it, and coupled with it the names of Mr. Lang and Mr. Hartland.

Well, it seems that, with regard to Mr. Lang, I was altogether unjustified in connecting such a theory with his name. He points out, fairly enough, that he has never unreservedly pinned his faith to the Casual Theory. He has "hedged" by granting that "something may be due to transmission", and now further supplements this by allowing that he should have said "much". Generally speaking, he claims to win on this point whether obverse or reverse turns up. But in making my strictures, I was not so much thinking of Mr. Lang's general remarks on this subject as his specific treatment of definite tales. He has given to the world some dozen delightful studies of special fairy tales: in only two of these. Puss in Boots and Jason ("A Far-travelled Tale" in Custom and Myth), has he allowed the possibility of borrowing, and in the latter case I still fail to gather whether he would allow that the Samoan variant must have been borrowed from abroad. In the other cases Mr. Lang was chiefly engaged in showing the underlying savage ideas which might have given rise to the story, presumably independently in different countries. It was this I was thinking of in fathering the Casual Theory on Mr. Lang, and in this I was far from being alone.

M. Cosquin took the same view of Mr. Lang's theories as I did. Professor Krohn shares the misunderstanding in his Bär und Fuchs. Here in England, among Mr. Lang's journalistic friends, there is nothing to be heard of but the Casual Theory. The young lions of the National Observer and the more elderly lioncels of the Saturday Review, are sublimely certain that resemblance in folk-tales is due to chance, not transmission. M. Sudre, in his recent study of the Reynard cycle, puts it that "l'anthropologiste Lang" is the author of the view "que tout conte est autochthone et a des représentants sur tous les points du globe parce que les idées primitives de l'humanité étaient partout semblables" (Les Sources du Renard, Paris, 1893, p. 8). M. Bédier, in his recent study of Les Fabliaux, is quite the casualist, and quotes Mr. Lang as his authority. Is it not too unkind of Mr. Lang to give away his English friends and French disciples with such a cœur léger? Nay, even after Mr. Lang has repudiated casuality and all its works, I observe that Lieutenant Basset, in reviewing the Cinderella volume, in which his palinode appears, sums up Mr. Lang's position naively: "Mr. Lang frankly acknowledges that he believes the details have been independently developed" (Folklorist, i, 177). It is clear that if I have misunderstood Mr. Lang, I have done so in good company. He will doubtless be deeply grateful to M. Cosquin and myself for giving him occasion to combat so widespread an error.

But is it an error? Is it not rather an essential adjunct of Mr. Lang's anthropological method of dealing with folktales to hold that the savage elements have existed everywhere, and that therefore the tales that embody them could have arisen anywhere independently? If the stories have been imported into civilised countries, the savage element in them cannot prove anything as to the primitive conceptions of those civilised lands, and the anthropological value of folk-tales is nil. I have already urged this objection in these columns (Folk-Lore, ii, 125), and I was not convinced by Mr. Hartland's reply in his Chairman's Address at the Congress. Mr. Lang seemingly yields his whole position in granting the probabilities of diffusion by borrowing, and we would like to know how far he has been convinced against his will.

It was mainly for this reason that I have urged the necessity of attacking the problem of diffusion first, as, till that is solved, the anthropological use of the stories is unjustified. Mr. Lang rebukes me, good humouredly enough, for not recognising his merits in pointing out the savage origin of the unnatural incidents of folk-tales. I willingly do so, though a word should be said for the interesting savage parallels drawn before Mr, Lang, by Mr. J. A. Farrer, in his Primitive Manners and Customs[7] But in emphasising these savage elements Mr. Lang has, in my opinion, diverted attention from the real nature of folk-tales, and the true method of dealing with him. By laying stress on the savage ideas in folktales Mr. Lang has associated them with myths and customs; they become with him and his followers in this regard, Mr. Hartland and Mr. Gomme, parts of primitive science. I contend that they are literature, folk-literature, if you will, but still literature, and so a part of savage or primitive art. It was for this reason that I ventured to express my surprise that Mr. Lang, a literary man par excellence, should have seemingly shown such little interest in fairy tales as literature. So far as his researches showed, he seemed interested in them not as gems of folk-literature, but as containing "survivals". Here, again, I appear to have misunderstood him, and he is indebted to me for an opportunity of disavowing such a heresy.

I know what Mr. Lang will reply to all this; he has so often explained his position that it is not difficult to apriorise the necessary deductions from that position. His chief concern was with the unnatural incidents in folk-tales. He had to rescue these from the mythological interpretations of the school of Kuhn and Max Müller. Instead of being degraded sun-myths, he has proved—it is not too strong a word—that they are "survivals" of savage customs. These he further considers to have existed among the European peasantry when they were in a savage state. With regard to the similarity in folk-tales, he is frankly an agnostic. Agnosticism is cheap to-day, as they say at the fruiterers. It may be scientific caution, but, on the other hand, it may be intellectual inertia. At any rate, it is particularly unfortunate that we should be made to halt between two ways on this question of diffusion, as upon it depends the whole value of the research after "survivals".

Mr. Lang is aware that for a certain class of folk-tales the problem of diffusion has been solved, for the derivation of a certain number of drolls from India has been, pace M. Bédier, definitely proved. Why may we not hope that we can also trace the paths of diffusion even when we are deprived of the aid of literary proof of transmission, as in the Indian cases? At any rate, it is in this hope that collections like those of Miss Cox are compiled. They may not trace folk-tales back to India; but they will certainly result in tracing each of them back to a probable birthplace, and it will be only for that birth-place that the doctrine of "survivals" will apply. For I cannot admit that because a peasantry receives and repeats a folk-tale with "unnatural" incidents, the peasants believe in the real occurrence of those incidents. It is of the essence of folk-tales that they are not believed to be true. Those that are so believed are myths, sagas, or legends, which are thus differentiated from folk-tales. Or is Mr. Lang of opinion that English children believe in speaking frogs or conversational tables because they enjoy The Well of the World's End, or Titty Mouse and Tatty Mouse?

  1. The second variant in Chambers does not contain the incidents marked in italics.
  2. The incidents marked in italics are clearly derived from some version of the Catskin type of story.
  3. Chambers, II, consists entirely and solely of these incidents.
  4. Who knows the Buck of Beverland nowadays?
  5. It is practically in Des Periers, Récréations, 1544.
  6. A modification would have to be made, however, when, as in most cases, the incidents are to some extent fixed in order. Thus, in Cinderella, the Happy Marriage cannot come before the Shoe Marriage Test. But Cinderella has seventeen incidents (supra, pp. 275-6), and these linkages would not reduce them to less than twelve complex incidents.
  7. Mr. Farrer is equally agnostic on the problem of Diffusion, Prim. Man., pp. 282-3