LITHUANIAN QUARTERLY JOURNAL OF ARTS AND SCIENCES
Volume 18, No.2 - Summer 1972
Editors of this issue: Antanas Klimas, Ignas K. Skrupskelis, Thomas Remeikis
Copyright © 1972 LITUANUS Foundation, Inc.
FOLK METERS AND LATVIAN VERSE
VALDIS J. ZEPS
University of Wisconsin
This paper attempts to elucidate the relationship between Latvian folk and art meters, and, specifically, the attempts to use folk meters to compose contemporary poetry. Since the metric structure of Latvian folk verse is poorly understood, the paper begins with an exposition of the principles of Latvian folk meters.
There are three elementary terms that enter into the composition of Latvian folk meters the metric syllable (which is not necessarily identical with the phonetic syllable); the bridge (=), which is the obligatory absence of a word boundary; and the cesura (/) or the obligatory presence of a word boundary. It is crucial to note at the outset that stress is not an organizing principle in Latvian folk verse, and, in reciting, is assigned by arbitrary rules, ignoring normal prose stress.
The above three terms metric syllable, bridge, and cesura are used in three folk meters the so-called trochee, the so-called dactyl, and the so-called asymmetric verse. The terms "trochee" and "dactyl" are misnomers, as they imply organizing principles that do not in fact exist. They are, however, well established, and I have kept them here for ease of reference.
None of the three meters is organized quite on the phonetic level. In addition to the phonetic syllables, there are additional grammatical "syllables" that must be counted in some positions and may be counted in others. These grammatical "syllables" are linked to specific forms of the conjugation and declension. There are, for metrical purposes, two classes of word-forms those whose phonetic and metric syllable counts coincide, and those whose metric syllable count may or must exceed the phonetic syllable count by one syllable. Some examples from the declension and the conjugation follow:
Nom. sg. vīrs 'man' 1 or 2 metric syllables
Gen. sg. vīra 2 metric syllables
Dat. sg. vīram 2-3
Acc. sg. vīru 2
(1) Loc. sg. vīrā 2-3, etc.
Inf. nest 'to carry' 1-2
1st sg. pres. nesu 2
2nd sg. pres. nes 1-2
1st pl. pres. nesam 2-3
2nd pl. pres. nesat 2-3
3rd pres. nes 1-2, etc.
Note well that the metric length cannot have a phonetic definition, since words like vēl "he wishes" and kur "you set fire" have 1-2 metric syllables, whereas vēl "still" and kur "where" have 1 metric syllable only.1
With the above in mind, we can proceed to discuss the three meters.
The trochee is a binary meter, and its basic building block (colon) is a sequence of four metric syllables followed by a cesura the last two syllables are linked by a bridge, i.e., must belong to the same word. Four cola constitute a trochaic distich (by convention written in two graphic lines, hence the name), which is the next higher unit of organization. A song contains at least one, usually two, and sometimes more than two distichs.2
Ja man būtu / tā naudiņa, / kas guļ jūras / dibenā,/
Es nopirktu / Rīgas pili / ar visiem / vāciešiem /
(2) "If I had the money that lies at the bottom of the sea,
I'd buy the Castle of Riga, with all the Germans in it."
In the trochee, the "extra" grammatical syllable must be counted at the end of a colon, and may be counted internally. In (2), above, most forms have the same number of metric and phonetic syllables, except man (1-2), guļ (1-2), dibenā (3-4), visiem (2-3), and vāciešiem (3-4). Of these, dibenā and vāciešiem are cola-final and must be counted as four-syllabic, i.e., vāciešiem = x/ and dibenā = x (using "=x" to designate a bridge followed by the "extra" syllable; again, note that this extra syllable is more than a filler, since it anchors, as it were, one end of bridge, which, like the syllable count is essential to the definition of the meter). The form visiem (2-3) occurs at the end of a colon and must be counted as trisyllabic, i.e., ar visiem=x/, which is a correct four-syllabic structure, ending as it does in two metric syllables, linked by a bridge. The words man (1-2) and guļ (1-2) may be counted as monosyllables or disyllables in a non-final position. In example (2) they apparently are counted as monosyllables, since the alternative count, e.g., *ja man =x būtu, would cause an overrun of five metric syllables,3 or one syllable too many for a trochaic colon. By the same token, cola of the type *ar vāciešiem=x cannot and do not occur, since vāciešiem would have to count as four-syllabic at the end of the colon.
Thus, although a very simple meter in some respects, the trochee provides for a great deal of phonetic variety in its realizations. It is a subtle verse, deeply ingrained in the grammatical structure of the language.
The basic colon of the dactyl consists of three metric syllables, the last two linked by a bridge, followed by a cesura.4
(3) x x = x /
Like in the trochee, a distich (i.e., four cola) is the minimum free unit, although usually more than one is needed to make up a song. Unlike the trochee, however, there is reason to postulate a line as an intermediate metric construct. In a full definition of the dactyl we must make note that line-final cola can be shorter by one syllable.
(4) x = x /
and all, except for the distich final colon can be longer by one syllable.
(5) x x x = x /
or, summarizing the above, the dactylic distich is composed as follows:
(6) (x) x x = x / (x) (x) x = x /
(x) x x = x / (x) x = x /
The dactyl permits yet another option the second line of a distich may be realized without a cesura. In that case, three bridges are required:
(7) (x) (x) x = x = x = x /
i.e., the cesura-less line is effected by bridging two disyllabic structures, each in turn already requiring a bridge.
Since each option implies two alternatives, it can be seen that the dactyl provides for a very large number of metric subschemes.
The finch and the titmouse
Audeklu/ meta/ were stringing the warp;
(8) Atskrēja/ vanags/ along came the hawk
Samudžināja/ and tangled it up.
The above verse instance conforms to the following metric subschema, allowed by (6) and (7).
x x = x /
x x = x
(9) x x = x / x = x
x x = / x x = x [vanags (2-3), counted as (3)]
x x = x = x = x /
The Latvian folk dactyl is a dynamic meter, strongly associated with music, dance, and improvisation.
In the asymmetric verse, the line is the basic unit of organization. The line is six- or eight-syllables long. The six-syllable line is organized as follows:
(10) x x x x x = x /
The eight-syllable line is composed of two cola:
(11) x x x = x / x x x = x /
The "extra" grammatical syllable may (but needn't) be counted before a cesura; it must not be counted elsewhere. While other arrangements are possible, the 6-6-8-6 stanza is the more common:
Apkārt kalnu gāju /
Kalniņā uzkāpu /
(12) Ieraudzīju/ līgaviņu/
As I was going by the hill
I climbed the hill instead
And whom did I see but a pretty young bride
And the tears that she did shed
The asymmetric verse is least intimately connected with the grammatical structure of the language (hence my attempt to "translate" it into quasi-English folk verse); it is often the meter of recruit- (marching-) songs, and has a vaguely "common" connotation.5
None of the three meters organizes word-stress;6 to the contrary, normal prose stress is suppressed and replaced by an arbitrary (colon initial; and, in the cesu-ra-less dactyl as well as in 6-syllable asymmetric, line-penultimate as well) reciting or singing stress:
Normal prose stress Assigned verse stress
aiz kálniņa (trochee)
ar bráli (dactyl) ár brāli
(13) kápu kálniņā (dactyl) kápu kalniņā = x
kálniņā úzkāpu (asymm.) kálniņā uzkāpu
gáuži ráudājam (asymm.) gáuži raudājám = x
Art verse, in Latvian, organizes stress. Since Latvian is stressed on the initial syllable, this, in effect, means organizing word-beginnings, in direct contrast with the folk verse, where metric units are maximally constrained at their end.7
Thus in Latvian art verse it is possible, as one alternative, to organize word-onsets and to count syllables (syllabo-tonic verse), as, e.g., in the following iambs by Rainis:8
Mums visiem vienas sāpes
Un vienas cerības
Kad es jums savas sūdzu
Jūs dzirdiet savējās. (VII, 198)
(14) We share the same anguish
Our hopes are the same
When I put mine in words
You recognize your own.
The metric subscheme that underlies the above is undoubtedly something very much like (15):
x x x x x x x
x x x x x x
(15) x x x x x x x
x x x x x x
i.e., word onsets may or must occur in odd syllables only, with the accompanying stress as the major organizing factor, either in fact or at least in the mind of the poet.
Or, one can follow the other major European art verse tradition and organize stresses with respect to line and not count syllables between them [reciting stresses added by me VJZ]:
Tu vári mūs šķélt, tu vári mūs láuzt
Mēs sníegsim tāles, kur sáule áust.
(Rainis, III, 103.)
(16) Break and splinter us, if you will,
We'll reach the distant sunrise still.
i.e., the lines have been written with four stresses in mind, and stress here apparently means onsets of "dominant" words, reminiscent of certain English verse. There are one or two syllables between stresses, but they are not organized in any particular fashion.
This is not the place to discuss Latvian art verse in detail, nor am I prepared to do so. There are many differences between folk and art verse, but the difference between the two general principles is crucial art verse first and foremost organizes word onsets, while folk verse places maximal constraints on how metric units end. There are many consequences to this basic difference. Art verse can tolerate elision, while in folk verse elision would destroy the basis of its organization. Unlike folk verse, art verse cannot tolerate serious departures from prose stress, since there stressed word-onsets provide the basis of organization. The use of rhyme results in serious over organization in folk-verse, whereas in art verse the use of rhyme is expected, etc.
The first efforts to use folk meter for art verse were made by foreigners and were quite ineffective.9 More serious efforts were made during the Latvian national renascence in the latter half of the nineteenth century, when the growing Latvian intelligentsia turned, in part, to folk heritage for poetic and metric resources. Note, however, that Latvian intelligentsia were the worst possible candidates to write folk/art poetry. They spent their formative years in Russian and German schools; their intuitive control of folk verse was minimal; and contemporary theory of metrics and grammar did not provide them with the conceptual framework needed to analyze Latvian folk meters. The pseudo-classical theory of Latvian folk verse ("dactyls" and "trochees") that persists to this day, viewed Latvian folk verse as organizing stress and length (neither enters into Latvian folk meter, except as a dependent variable).
The lack of an adequate conceptual framework was no barrier to introducing and adapting classical and foreign meters. For one, these were organized on the phonetic surface, with metric and phonetic syllables being one and the same, and, thus, open to inspection (in effect, easier). Furthermore, the poets were to a degree free to develop their own canons of taste and acceptability in art verse; they did not have a comparable latitude in their attempts at folk verse, for there the canons had been set, and no degree of error was tolerable. Faulty folk verse was immediately recognizable as such.
Attempts toward a folk/art verse were, nevertheless, made, as the national-romantic appeal of a native art meter was strong. Thus, in manufacturing a "folk" epic from pieces of folktales and from mythological inventions, Lautenbachs-Jūsminš produced endless dactyls of the following type:10
Pagāja/ tiem drīz/ Laimiņas/ gadi/
Aizritot/ kā zelta/ Saulītes/ rati/
Pa tam/ dzemdēja/ Zalkš'līgaviņa/
Tam divus/ bērniņus/ dēliņu,/ meitu
(17) They soon spent the years that Laima had
As they rolled by like the golden chariots of the Sun;
Meanwhile, the bride of the Snake King
Bore him two children, a son and a daughter.
Besides being mawkish, inflated, and pedestrian, his dactyls here are quite acceptable. Other examples, however, show that Lautenbachs-Jūsmiņš, while he could produce any number of correct dactyls, could not rule out incorrect ones, e.g., in a nearby passage he has a colon of the shape vesti no 'to lead from' where the last two syllables are not linked by a bridge.
Probably for no inherent reason other than the ineptness of its practitioners, the imitation of folk verse soon fell into disrepute; not until Rainis' work of the second decade of the twentieth century did a major poet attempt to write in folk meters.
In his plays and poems, Rainis introduced several new dimensions in the use of the folk meters. He was first to realize that all three meters the trochee, the dactyl, and the asymmetric could be used for individual and special effects. He further experimented with introducing triple structures (three-line stanzas, not encountered in folk verse), with splitting lines and even cola between different speakers of a dramatic dialogue, etc.
It must be stated from the outset, however, that Rainis' intuitive and intellectual control of the principles of folk meter in no way exceeded that of his less talented predecessors, even though his own opinion of his competence was high.11 Along with good verse, he produced lines that are, from a metrical point of view, atrocious, to the point that only the sheer genius of other lines and the force of his ideas salvage the plays written in folk/art meters.
Visu cauru mūžu
Ar vien' daiļu meitiņ' (II, 202.)
(18) I played for and danced with
As if for a lifetime
With a beautiful maiden
In the above three-liner (meant to be a folk dactyl), anticipating with pathos the eventual tragedy of the protagonist-fiddler, the effect is marred by the jarring third line
(19) Ar vien' daiļu meitiņ'
where the unnecessary Germanism vien' one, a' forces the line to have the following incorrect metric structure
(20) x x x x = x / x x = x
i.e., there are too many syllables in the first colon, not salvaged by the use of the forbidden ellipsis (vien'). Metrically acceptable, but stylistically bad is the use of the diminutive (used very widely, but within limits, as an expansion device in all Latvian folk verse) in a stanza final colon, where the two syllabic meitu would have been preferable. As is, (19) reads like a doggerel, and unnecessarily so, since a correct line (ar daiļu meitu or ar daiļo meitu) is readily available. Yet some of Rainis' dactyls, as Rudzītis points out,12 are flawless in all respects:
Visa mana dvēsele
"All my soul
Iztvīkusi. (III, 14) thirsts."
(21) x x x = x / x x = x
x = x = x = x /
Rainis' trochees is, unfortunately, no better
Lelde skaistāka. Es zemei
Tādas dziesmas uzlaidīšu:
Smiedamies tās dziedās visi,
Tās no ļaužu atmiņām
Neizcirst ne kapa kaplim! (II, 44.)
(22) Yet Lelde is fairer! And as to Earth,
I'll fashion songs about her
from curses and jeers.
Mocking crowds will repeat them
till even the gravedigger's shovel
won't scrape them from their sculls!
There are only two elementary mistakes that one can make in Latvian folk verse the bridge may be missing, or the syllable count may be off. Rainis consistently makes both. In (22), where the fiddler Tots tries to intimidate Earth against accepting the lifeless Lelde, the first colon runs five syllables (rather than four). In line four, the first colon lacks a bridge (smiedamās tās), as does the first colon of the last line (neizcirst ne), and the last colon ends in an extra metric syllable, which must be counted (kapu kaplim = x), making the colon 5 metric syllables long. To Rainis' credit, he attempted to exploit the whole range of alternatives that he thought were made available by the native meters, rather than sticking to a few safe patterns. His mistakes are not those of a timid man.
Rainis is most effective in his use of the asymmetric verse, which least requires a measure of intuitive and intellectual understanding of its principles of organization. It is not, above all, a subtle meter it can be shouted as a slogan, as in (23); or, sung as an anthem, as in (24).
Zeme zeme, kas tā zeme
Ko tā mūsu dziesma prasa
Zeme tā ir valsts! (VIII, 131)
(23) Land, land, what's that land,
what does our song demand?
By land we mean a state!
(in the above example, valsts is counted as x = x, as expected).
Daugav' abas malas
i Kurzeme, i Vidzeme,
i Latgale mūsu. (VIII, 128)
(24) The two shores of Daugava
never shall part.
Kurzeme, Vidzeme, and Latgale
they all belong to us.
Later efforts to use folk meters in art poetry continue to show a profound misunderstanding of its nature, as the following line by Medenis illustrates (cesuras mine VJZ):13
"Labi, ka/ ievācām,"/ saiminieks/ atteica,/ "sausais gads/ klātu."
(25) "It's well, that we gathered it in," said the farmer, "lean times are upon us."
Here Medenis works with a metric innovation, namely, a six-cola (three-line) structure; whether or not this is acceptable can be argued. His very first colon, however, lacks a bridge. This latter point cannot be argued, since a bridge is essential to all three folk meters. Medenis, has written in a perfectly acceptable syllabo-tonic art dactyl (but not in "folk dactyl").
The relationship of Medenis' verse to folk and art meters is a very complicated one. At his worst, he tries to combine the two, and falls between the two stools of writing neither good art verse, nor good folk verse. This is a subjective judgement, i.e., I would hate to see a change in the standards of taste so as to make his verse fully acceptable. As I said, this is a complicated topic, and requires a full and detailed treatment of its own.
A few points, however, are instructive and can be made with minimum interruption of the main presentation. They concern the use of occasional structures based on folk meters in verse that is composed primarily in an art meter. In order to register as different from art verse, such structures must violate some condition of art verse (or be direct quotes of familiar instances of folk verse, or employ some stylistic device confined to folk poetry, as, e.g., the use of diminute for metric rather than semantic purposes, etc.).
Pákaļ tai šāvieni šķiļas kā krámi
. . . .
Skáts pēkšņi ápžilbst, pilns sārta tváika,
Lāsojas déniņi, kūp ritēdami (my bold VJZ)
Asiņu pilieni sméldzīgā liesmā!
Sasala jūriņa līdz dibenam!14
(26) Pursued by sharp reports of flashing flint. . .
Red mists then blind the eyes, the temples start to drip,
and drops of blood ooze forth in stinging heat.
The sea lies flash-frozen to its very bottom.
There is no question that here the basic rhythm is some sort of art dactyl, as the segments pakaļ tai and šķiļas kā in the first line illustrate; there is no bridge, and there is no way to read the first line as folk poetry. The underlined portions in (26), above, illustrate two instances of a folk dactyl within a poem that is otherwise composed in an art dactyl. The last line is a straight quote from a familiar folksong, and needn't concern us here. It registers as folk-verse, by it's familiarity alone. In the case of kūp ritēdami, since we have no direct access to the poet's intent, we have two options to read it as kūp ritēdami, i. e., with normal prose-/art-verse stress, and accept the extreme imbalance in the count of the unstressed syllables; or, to read it as kūp ritēdami, i. e., as a cesuraless folk dactyl, composed on the pattern
(x) (x) x = x = x = x /,
i. e., with a triple bridge and an arbitrarily assigned stress on the penultimate. I believe the latter reading is the intended one.
The above illustrates that, in mixed verse, art verse expectation is the dominant one; all that can be interpreted as art verse, is so interpreted, and only stretches that violate art verse conditions outright (or are familiar quotes, etc.) are interpreted as folk verse. This, of course, limits the role of folk meters in mixed verse to decoration, or "folk flavoring."
Returning to the main argument, we note that there are, of course, numerous instances of good folk verse in Latvian poetry. Thus, Andrejs Eglītis in his Thy land, O God, is in flames15 switches in the middle of the cantata to a (correct) folk trochee, for a very effective change in mood and sudden sobering and moving effect. His switching to a folk meter here is rather akin to switching back to one's native language in order to say things which cannot be expressed in an alien tongue; for that very reason, the effect is unfortunately lost in translation, and I will not attempt to illustrate this point.
Occasional successful lines and stanzas are, of course, no test of a poet's craft. The last sustained effort to write in folk/art verse that is known to me is the one-act tragedy Siege of Rakte by Zīverts, written in folk trochee with some passages in folk dactyl.16 With all due respect to our best contemporary playwright, the verse of the play is frequently flawed by the same elementary errors of false syllable count and missing bridges. Both occur in the passage below:
Varēji vēl brīdi paļaut
grāvī sārtu sakurināt
ķēžu vīrus pasvēpēt
(27) You could have let us build
some fires to smoke the
chain-mail wearers out.
The first colon (varēji vēl) lacks a bridge, the second (brīdi paļaut = x) and fourth (sakurināt = x) are too long. And yet metrically acceptable substitutes are readily available (my rendition VJZ):
Vēl varēji brīdi ļaut
(28) grāvī sārtu kurināt,
ķēžu vīrus pasvēpēt.
In fairness to Zīverts, he has a much better control of his meters than Rainis ever had Paradoxically, Zīverts' verse is the more painful of the two to the ear, inasmuch as Rainis' attempts were often so totally out of alignment with folk verse as to no longer invite comparison. Zīverts verse, on the contrary, is unmistakably meant to be in folk meter, and has to stand constant examination against the norms.
In view of all that has been said, one may ask should there be and can there be a successful native folk /art verse within the Latvian poetic tradition? In spite of some obvious difficulties, I would like to believe that the answer is a qualified yes to both questions.
The first question is, of course, a matter of values. What is it worth to the Latvian poet and playwright to write and be read in a truly native art form? And what is it worth to the nation to not let an existing and highly refined folk art form succumb to mere urbanization?
The second question is a much more intricate one, and the answer partly depends on the answer to the first. But even apart from it, there are some serious problems. Without doubt, the folk verse cannot serve all purposes of the modern playwright and poet. Could it do so in a modified form? There is no point in speculations along this particular line. Successful innovation surely must proceed from the basis of mastery over some technique, and not arise as the result of ignorance or lack of skill. It is very clear, however, that the initial obstacle to successful Latvian art/folk verse is the pseudo-classical theory of Latvian folk meter that is still taught in Latvian elementary and high-schools, insofar as the pseudo-classical theory provides at best incomplete, at worst false guidelines to Latvian folk meters. The situation has, after all, changed very little in the last hundred years the poet intellectual still goes through a long de-ethnicizing period of education, loses his intuitive basis for the native meters, and cannot compensate for this loss with formal criteria learned in school, since these criteria are based on the pseudo-classical theory of poetry.
Ultimately, of course, the future of Latvian folk meters in Latvian poetry rests upon the skill and resolve of the poets. There has, to date, been no lack of resolve.
1 The "extra" grammatical syllable is often actualized in singing or reciting, typically by some arbitrary vowel i.e., in (1)
nesti, nesami, viris', virami, etc. (the so called lapamais patskanis or "padding vowel"). In the pseudo-classical tradition of Latvian poetics this vowel has no organic basis in the language; it is used "where needed." In this paper, the "extra" syllable enters in the discussion from the very beginning, as it enters in the basic design of the meter. It is not, however, an ad hoc phenomenon; there is an independently motivated vowel loss rule in Latvian morphology (M. Halle and V. J.
Survey of Latvian Morphophonemics," M.I.T. Research Lab. of Electronics Quarterly Progress Report, 83 1966, p.p. 113-116), and the variable metric count in some grammatical forms is invariably the direct reflection of this vowel-loss rule. Technically speaking, Latvian meter in part organizes language material in a slightly more abstract morphophonemic representation than the phonetic. Thus, in a more precise sense, the meter does organize phonological material, and not arbitrary "extra" syllables.
2 V. J. Zeps, "The meter of the so-called trochaic Latvian folksongs," International Journal of Slavic Linguistics and Poetics, VII (1963), 123-8.
3 To show that words with a 1-2 count can be counted long colon-internally, consider the example
Briest rudzi/ briest mieži,/ briest jauna/ līgaviņa.
"The ears of rye are beginning to fill out, the ears of barley are beginning to fill out, and so is the young bride."
In this example briest (1-2) is counted long, as in the colon briest = x rudzi.
4 V. J. Zeps, "The meter of Latvian folk dactyls," Celi, XIV (1969), 45-47.
5 V. J. Zeps, "Latvian folk meters and styles," [Halle Festschrift (in press)].
6 The pseudo-classical notion that Latvian folk meters organize stress and length is in fact the commonly held view. For this view, consult, e.g. L. Bērziņš, Ievads latviešu tautas dzeja (Riga, 1940), esp. pp. 57-80, or, for that matter, any appropriate Latvian high-school text.
7 In this paper I have tried to present ray material in a way that no more unfamiliar terms are introduced than absolutely necessary. For the specialist, a more effective way would have been to present all verse (folk and art) in terms of alternations of maximally organized (constrained) stretches and minimally constrained stretches (and the actualizations of each). Interested readers can 'pursue this point in M. Halle and S. J. Keyser, English Stress: Its Form, Growth and Utilization in Verse (New York: Harper & Row [c. 1971]) and, with specific reference to Latvian, M. Halle, "On Meter and Prosody," Progress in Linguistics, M. Bierwisch and K. E. Heidolph, eds., (The Hague; Mouton, 1971), 64-80, and Zeps (in press) op. cit.
8 All references to Rainis will be to the standard edition, by volume and page: J. Rainis, Raksti, 17 vols. (Vasteras: Ziemeļblazma, 1952-1965).
9 L. Berziņš, pp. 27-31 and elsewhere.
10 Lautenbachs-Jūsminš, Niedrišu Vidvuds (Jelgava, 1891), p. 88.
11 K. Dzilleja in Rainis, I, p. 181.
12 J. Rudzitis, Raiņa Ritmi (Vasteras: Ziemelblazma, 1958), 47.
13 Cited after Berziņš, p 173.
14 J. Medenis, Mikelnicas (Vasteras; Ziemelblazma, 1952), 232.
15 A. Eglitis, Dievs, Tava zeme deg ([Copenhagen]: I. Reitmenis, 1948); previously published as "Jau tuvu ta stunda" in Uz vairoga (Jaunais Vards, 1947), 7-10.
16 M. Ziverts, Tirelpurvs un Rakte (Stockholm: Daugava, 1946), pp. 72-3.