From Peruigilium Veneris, The Eve of Venus, incerti auctoris, carmen de vere, in Latin and in English, edited and translated with a Commentary, by R. W. Postgate, Houghton Mifflin Company, Boston and New York, 1924; pp. 23-35.

      This online text is made available with the permission of Professor John Raymond Postgate and his newphew, Daniel Postgate, two remarkably kind men, both as notable in their own right as their illustrious ancestor.

Cover Title, 'The PERVIGILIVM VENERIS' in a diamond shaped scroll border.






IT WOULD not be possible to discuss here the various re-arrangements of the Peruigilium attempted by scholars. Mr. Clementi in his Studies in the Peruigilium (1913) gave five re-arrangements in tabular form, which were all intended to restore to the poem a strophic structure. They none of them agreed in any but the most trifling particulars. (His own re-arrangement was the simplest and probably the least violent). Dr. Mackail cut up and re-arranged the poem into quatrains, and wrote in verses of his own to fill out the poem where his scheme impaired it. Since then, Mr. Fort has added a new scheme, postulating a written and a sung version of the poem — also with insertions. To this list Mr. Clementi added a further table of eleven more re-arrangements which were not based on a strophic plan, and he could have increased that number.

These alterations, some of which are really astounding, are based on two assumptions :  firstly, that the refrain cannot occur at irregular intervals, but that a plan or scheme must be hidden; secondly, that the MSS. are in a very corrupt condition and there are difficulties that show that large transpositions have occurred.

The first assumption has never been proved, and is probably not provable. Against it are the facts, firstly, that according to the MSS. the refrain does occur at irregular intervals; secondly, that seventeen different attempts at least have been made to “re-introduce order” into the poem, and they none of them agree at all. It may be that the poem had originally a strophic or quatrain construction, but scholars have not succeeded in discovering what it was, or agreeing upon it. It is true enough, as Mr. Clementi and Dr. Mackail argue, that after they have made certain of their changes the “outline of a structure begins to appear.” He would indeed be a clever poet who could write a poem such that no determined editor could, by [26] reckless transposition, force a rhythmic character on it. Until a convincing re-arrangement is brought forward, it is wisest to hold as near as is reasonably possible to the manuscript arrangement, and admit that the refrain may have occurred at irregular intervals in a poem which is half-way between ancient verse with its strophae and medieval hymns with their quatrains.

Now, as to the manuscript text. Are the MSS. very corrupt, and are large transpositions necessary? A formidable list of apparent errors is given in the list below, which I have endeavoured to make as full as possible. But really there are few evidences of deep corruption. If the erratic spelling of the two manuscripts (Codex Salmasianus, “S,” and Codex Thuaneus, “T”) is reduced to any accepted standard, a good half of the errors disappear. There are also one or two obvious blunders, for the copyists were in both cases singularly illiterate. Omitting such blunders and misspellings, there are only twelve lines, in this supposedly grossly corrupt poem of ninety-three lines, where the MS. reading demands alteration. It is not to be denied that the present condition of both MSS. is, on the face of it, unsatisfying, but that the manuscripts (or manuscript, as is likely) from which they were copied, was seriously damaged, is by no means proved. There is indeed only one line (22) where I suspect deep corruption :  apart from this, the lines of the poem made good sense with a minimum of alteration.

If we consider their order (so far as the two questions can be considered separately) a similar observation is true. One transposition has certainly occurred :  line 58 should follow line 39. Where one transposition has occurred, others, of course, may have. But that does not give us free licence to move lines about at our pleasure unless, independently of any a priori strophic theory, the sense of the poem requires it. Nor is it sufficient to argue — as the gentlemen did who wrote the notes I quote on line 90 — that [27] the poem is artistically improved by such alterations. We must be convinced that the changes restore what the poet wrote, not what he ought to have written. Richard Bentley, in editing Paradise Lose, made vast improvements, as he thought, by such methods, but the great doctor’s edition of Milton is generally held out as an example to be feared rather than imitated.

It is not possible to deal with all the transpositions which have been made. Two of the most frequent require mention. Lines 73—74 and 71—72 were transposed by Bährens. They also require emendation, and, emended, give sense as they stand. Transposed simply, they do not appear to me to give very good sense. The question is one which any reader can answer according to his own taste.

More serious is the question of line 9, in which it is argued that the word tunc has no meaning. The suggestion by Mr. Owen that it refers to uer in lines 1—2 has been rejected on the ground that the distance is too great between the lines. It is also argued that it is odd that the author in a poem of 93 lines should “postpone till line 59 his explanation of why the coming day will be so high a festival” (Clementi). Hence it has been a common practice to transfer to this place some or all of the passage beginning cras erit quom primus Aether (line 59). Once this beginning has been made the floodgates are open, and the editors go right on to rearrange the poem in a strophic and often catastrophic manner. All this depends on the assumption that tunc has no meaning. This hardly seems just. Tunc, I wish to submit, has precisely its proper meaning :  “at that point of time, on that date.” Tunc is not tum. It refers, as is natural, to cras in the line before. It offers the explanation why to-morrow will be “so high a festival.” The three lines mean, briefly, “To-morrow was Venus’ birthday.” It is no odder that fecit should follow tunc when it refers to cras than that copulauit, also in the perfect tense, should follow cras erit quom in line 59.


I append a list of readings. Wherever either or both of MSS. differ from the text I have given the variant, even if it is only an obvious misspelling. I have also given an arbitrary selection of the more interesting conjectures. Pithoeus is Pithou, who published the editio princeps. Dousa is the younger Dousa, who palmed off for fifty years an indecent fragment of a fake Peruigilium on the world. S is the Codex Salmasianus, T the Codex Thuaneus. Both these MSS. are in the Paris Librairie Nationale.

[Note: This browser window will not display the text below properly if your screen size is too small. So, if the line numbers bleed into the text, widen your screen !!! — Elf.Ed.]

HEADING from S only

 (1)S   amabit quique amauit:  T   amauit quitque cras amauit amet.

 (2)S   uere natus Iouis :   T   uer natus orbis.

 (3)T   nubent.

 (4)T   resoluet.

 (5)T   amorem.

 (6)S   gaza uirentis . . . mirteo :   T   gazas uirentes . . . myrteo :   Pithoeus casas.

 (7)S   fultas sublime trono.

 (8)T   quinque.

 (9)S   superbo :   T   tuno quiuore de superhuc spumeo pont’ de glouo.

(10)ST   cerulas :   T   uipedes.

(11)de maritis imbribus is probably a mistake from line 3  :   Riuinus read marinis which is lame.

(13)T   gemmas :   ST   floribus :   Rigler floridis.

(14)T   surgentes . . . sparitu (altered to spuritu) :   ST   faboni.

(15)S   notos penates ipsas :   T   totos pentes ipsa :   Scriuerius nodos :   Lipsius tepentes :   Bährens feraces.

(16)T   relinquid :   ST   tumentis :   Pithoeus umentes.

(17)S   et micanat lacrimas :   T   et mecanat lacrime . . . cadū :   Lipsius en micant :   Salmasius emicant.

(18)S   urbe . . . sustine :   T   preceps orbe . . . sustinet.


(19)S   in pudore florulente . . . purpore :   T   in pudorem florulente . . . purpure :   Schultze en.

(20)S   humor :   T   umor.

(21)S   tumenti :   T   papilla.

(22)S   manet tute :   T   ip . . . mane tuae :   Dousa ut udae :   Orelli totae (“all”) :   Mackail nudae, expanding the line into two. The line presents two further difficulties (1) “iussit” is used elsewhere in the poem but with its normal construction (2) “ipsa . . . facta” appear to refer to the same person, whereas, of course, the second line refers to the rose. The corruption lies probably very deep. Cf. note on line 63.

(23)S   facta prius :   T   fusta prius . . . decque :   Conjectures are : Facta Cypridis, Veneris, Paphies, de ipsius.

(24)S   purporis.

(26)S   unica marita noto :   T   unica marito nodo . . . pudent :   Bergk uoto (“marriage”) :   Bucheler unico Noto marita.

(28)ST   nimfas :   T   loco.

(29)ST   et :   T   comis :   Pithoeus it.

(31)S   in te nimfe :   T   ite nymfe.

(32)S   iussus e . . . nudos :   T   est . . . durus.

(33)T   quid acuneo . . . quid digne :   ST   lederet.

(35)S   inermis sidem :   T   inermis idem :   Pithoeus in armis :   Scriuerius est in armis totus (for metrical reasons but the trochaic metre had a strong accentual element).

(36)S   amit :   T   qui amauit.

(37)S   unam res :   T   una re.

(39)S   tragibus.

(40)T   omits altogether :   S   uellit erogare.

(41)S   uellit.

(42)feriantes and feriatos have been conjectured.

(44)S   myrteo.

(45)S   baccus :   T   baccas . . . deas.

(46)S   detinente . . . peruiclanda :   T   detinent et:   Mackail [30] de tenente (“unceasingly”) . . . perviglanda :   Clementi, after Owen detinenda est tota noctis pervigilia canticis (pervigilia em. Sing.vigil”).

(49)S   hyblei :   T   ybleis.

(50)S   presens . . . dicit adsederunt :   Dousa praeses . . . dicet, adsidebunt.

(51)S   hibla totus fundet quidquid annos adtullit :   T   annis.

(52)ST   hibla :   S   superestem :   T   rumpereste :   Crusius sume uestem :   S   etnec :   T   ethne, Enna is a common variant for Aetna, which is not remarkable for flowers :  cf. Ovid, Met. V 385, Columella X 270, Sil. Ital. I 93 etc.

(53)T   t for uel.

(54)S   quae locus :   T   quaeque locos :   ST   montes :   Scriuerius fontes.

(55)T   alitas :   Bucheler mater alitis dei.

(56)T   nullo.

(58)ST   rigentibus :   S   uirentes duoad umbra :   T   uirgentes ducat umbras :   the anonymous Leipzig edition (1852) transferred this line to its correct place, after line 39.

(59)S   quo (for quom) :   T   qui :   S   ether copolauit niptias.

(60)S   creauit :   ST   uernis :   Pithoeus uernus.

(61)S   alme :   T   fluctus almae et.

(62)S   flaetus . . . omnis alteret :   T   ut fletus . . . aleret.

(63)T   uernas :   S   adque :   T   permeante spū. Notice that ipsa here as always in this poem is the goddess, the mistress of the festival (suamque norat ipsam), being almost a proper name. It is easily understood and should not be used as an excuse for shifting the passage elsewhere on the ground that it refers to coniugis. The passage must remain here because of the sense, which is:Earth and Ether made all living bodies, but Venus controls them.

(64)S   gobernat procreatis.

(65)S   per quem . . . per quem.

(66)S   tenderem.


(67)T   nosce.

(69)T   nec potes . . . latino.

(72)S   romuleas ipsas :   T   rumuleas ipsa . . . Saumis.

(73)S   samnes:   T   ramnes.

(74)ST   romoli matrem crearet :   Bucheler Romulum patrem :   Fort Romuli parem creauit.

(76)T   facundat.

(77)ST   dione :   S   natu.

(78)T   perturirit :   S   sinum.

(79)S   deligatis

(80)T   amat for second amauit.

(81)ST   aonii :   Pithoeus tauri.

(82)S   tutus quo tenetur coniugali federe :   T   tuus . . . cum iugali :   Mackail coetus continetur.

(83)S   gregis :   T   ualantum gregum.

(84)S   canores.

(85)T   cigni :   S   stangna quinni.

(86)S   popoli :   T   adsonante aerei puelle supter.

(87)S    mussico :   T   putas.

(88)S   queris :   T   eet.

(88A)Refrain inserted by Clementi (and others, following Riuinus) in pursuance of his strophic arrangement. I have retained it because the whole structure of the poem demands a break of some kind here.

(89)S   quan :   ST   uir.

(90)S   fiam ut :   T   faciam ut . . . taceret :   ST   celidon :   Rivinus suggested “uti chelidon,” but he preferred “quando fies muta, aedon, ut tacente te canam,” meaning, “When will you be silent and let me have my turn?” Wernsdorf believed, on the strength of an inscription, that Chelidon was the name of the poet’s wife and the line meant, “When shall I talk as much as my wife?”

(91)T   perdidimus antacendonĒ :   ST   foebus :   Thomasius Apollo.


(92)S   amiclas . . . perdedit :   T   amidas . . . perdedit :   ST   taceret :  emendations tacebant, tacerent. I have bracketed this line and struck it out of the translation :  this is the only liberty that I have permitted myself upon my own unsupported judgment. It was, indeed, the annoyance of seeing this wretched interpolation at the magnificent end of a magnificent poem, turning up unchallenged in every edition that made me decide to offer this edition. Consider the position of the line in the whole poem and its relevance. The writer of thisleaf lost from the closing manuscript of Roman poetryhas been dwelling upon the new life that is rising all around him, because spring has come. But the song of the nightingale, happy and not sorrowful though it is, makes him break off and remember that for him there is no spring :  the utter neglect of art and beauty has brought its natural consequences :  thedivine afflatus,” as it would have once been called in English, has gone.

     Perdidi musam tacendo, nec me Apollo respicit. So the poem ends on a note of sadness, as, indeed, it should. For this is the last of the Roman singers :  it is not, it is true, of the barbarians from over the frontier that he is thinking, but of a sister symptom of decay, the dead artificiality of the art and poetry of his day. Then, to this perfect ending somebody has added the fatuous comment, “So Amyclae, when it was silent, was destroyed by silence.Whoever wrote in the margin his “Sic Amyclas perdidit silentium,” to show that he had read Virgil Aen. X 564, was not an inspired commentator, but the introduction of the redundant, ungrammatical, and prosaic “cum taceret” to make the line scan, has made his inappositeness worse. And what has Amyclae to do with it, anyway? According to the scholars, there were two Amyclae, one in Laconia, one in Italy. The story about silence is obscure, and it is not certain to which town it belongs. But it appears to have been a variant of the “crying Wolf! Wolf!story. What relevance an allusion of such a kind has here is more [33] than any scholar, so far as I can find, has attempted to say.

     Finally, and this is important, Mr. Fort, in his edition dated 1922, has pointed out that thePeruigilium” excludes — with this exception — all words longer than three syllables at the end of the line. Where this rule is not observed, as he proves from examples, the introduction of quadrisyllables and longer words makes a startling change in the quality and music of the verse. He himself retains this verse, because he requires it for an elaborate reconstruction of the poem, but unless we accept his quatrain arrangement there is no need to retain it.

(94)Explicit, etc., given only in T.

Tail-piece engraving of a red diamond made of scroll-work.


Pelican Publisher's Logo.


Keep rare books coming!