[In the essay below, Roche argues that Petrarch consciously utilized Renaissance concepts of numerology in the structuring of the Canzoniere.]

The purpose of this essay is to argue that the ordering of the three hundred and sixty-six poems in Petrarch's Canzoniere is numerologically oriented and that one of the main structures in this ordering is a calendrical framework that places the Canzoniere unequivocally in the context of fourteenth-century Christian morality. Without even referring to recent studies about numerological composition in English poetry of the Renaissance or to the overwhelming evidence of Biblical commentaries from earliest times through the seventeenth century,¹ we have the figure of Dante, whose numerological structuring of La Vita Nuova and La Divina Commedia has never been called into question. Behind Dante are the Platonic and Pythagorean theories of the mathematical basis of the universe. From the time of Plato's Timaeus and later of Boethius' Arithmetic (both works used as text books throughout the Middle Ages and the Renaissance) men learned that God had created the world in number, weight, and measure, and these philosophical texts for the Christian reader merely corroborated the evidence from the Book of Wisdom: "Omnia in mensura et numero et pondere disposuisti" (Liber Sapientiae, XI: 21). By incorporating numerical proportions into his poems, a poet would be eliciting relations already existent in the world and thereby enhancing his poem and making it more "real."

In such a world Petrarch clearly saw beyond mere coincidence when he insists that his first sight of Laura was the same hour and day and month as that of her death:

Laurea, illustrious for her own virtues and long celebrated in my poems, first appeared to my eyes about the time of my early manhood (sub primum adolescentie mee tempus) in the year of the Lord 1327 on the sixth day of the month of April, in the church of St. Clare in Avignon, in the morning (hora matutina); and in the same city, in the same month of April, on the same sixth day, at the same first hour, in the year 1348, her light was withdrawn from this light, when I by chance was then at Verona, alas, unaware of my fate. Moreover, the unhappy news reached me through a letter of my Ludovico in the same year, in the month of May, the nineteenth day early. That most chaste and most beautiful body was placed in the church of the Franciscans on the very day of her death at vespers.…²

In a world governed by number, one might well begin to believe that a conspiratorial Providence had been numbering her days. I do not want my last punning remark to be taken too lightly, for there is evidence that this kind of numbering of days was familiar to Petrarch. He writes about his grandfather, Ser Garzo, who predicted the hour of his death and having reached the ripe age of one hundred four, when the predicted hour came, lay down quoting the Psalmist: "I will lay me down in peace and sleep" and died. Also Petrarch's earliest surviving poem, written when he was fourteen or fifteen, is thirty-eight Latin hexameters on the death of his mother. She was thirty-eight years old.

I do not want to stress these more lugubrious aspects of numerology, but we have grown away from the belief in Providence and the patterning of human life so far and in so many ways that we can hardly think ourselves into a time when men believed in a God of Love, Who promised that good would eventually emerge from evil, Who proved His Providence in the book of His Revelation and in the trials of His Chosen People, and from Whom the living would expect no less. A man could derive emotional comfort from the power of number to rationalize loss or to memorialize the departed creature through the very same power used by God to create in number, weight, and measure.

For the reasons just indicated and for others that I will develop later I cannot accept the late Professor Wilkins' assumptions about the structuring of the Canzoniere:

Miss Phelps has shown that within each of the two parts [of the Canzoniere] the poems are as a whole arranged with great artistic care upon three principles: (1) the maintenance of a generally but not strictly chronological order; (2) the securing of variety in form; and (3) the securing of variety in content. In accordance with the second and third principles, for instance, canzoni are so placed as to prevent the existence of long series of sonnets, and political poems are so placed as to prevent the existence of long series of love poems.³

I do not find these assumptions convincing as aesthetic principles. Like the outmoded theory of "comic relief" in Renaissance tragedy they do not take into account the extraordinary intellectual analogies that exist between the various poetic forms and topics that constitute the completed Canzoniere, and they fail to see the real and valid relation between the love poems and the political poems.⁴ Even more to the point, these assumptions do not begin to probe the complexity of Petrarch's innovation, discussed by Wilkins later in his book:

Of the 366 poems of the Canzoniere, 29 are canzoni, nine are sestine, seven are ballate, four are madrigali, and 317 are sonnets. In the Canzoniere these forms are not kept separate, but are so mingled as to afford a pleasing variety. In view of the consistent practice of the separation of canzoni and sonnets in MS collections of pre-Petrarchan lyrics, Petrarch's procedure in mingling canzoni and sonnets is clearly seen to constitute a notable poetic innovation.⁵

Wilkins' magisterial studies of the Canzoniere, for which I have nothing but admiration, are primarily genetic; they quite rightly emphasize the "making" of the sequence and of necessity do not concern themselves with the completed work. Without denying the validity of Wilkins' researches (in fact, using them to test the validity of my own assumptions) I want to examine the evidence of structural patterns in the Canzoniere to see what they tell us about the meaning of the sequence and about Petrarch's innovation of mingling canzoni and sonnets.

Let us take the simple example of the four madrigali, all occurring in the first part of the sequence. They are numbers 52, 54, 106 and 121. We know that 121 (Or vedi, Amor) was a late addition in Petrarch's manuscript and that it replaced the ten line ballata Donna mi vene, which appeared in most of the early manuscripts and even some of the early printed texts.⁶ Is it mere coincidence and even some printed of the early of that the number the first two (52 and 54) add up to the number of the third (106) and that canzone 53 contains 106 lines? Perhaps. But if we look at the alternation of forms in the two segments 38-54 and 105-121, we may see a reason why Petrarch replaced the ballata with that fourth madrigale.

The formal pattern of 38-49 is repeated in 107-118, and the admittedly short pattern of 53-54 is repeated in 105-106, the two patterns being reversed. The facts that these two segments form a pattern embracing all four madrigali and that the groups of twelve sonnets are the only two in the Canzoniere suggest that the pattern I am describing may have been intentional. It would seem all the more likely in that 55-63 is another formal pattern of ballate and sonnets.

My suggestion about the pattern of the madrigali becomes even more probable if we consider the integration of that pattern with the grouping of the first seventeen canzoni. Only five times in the entire sequence does Petrarch group canzoni together: 28-29, 70-73, 125-129, 206-207, 359-360. The first three groups are respectively two, four, and five canzoni. If we consider the placement of the first seventeen canzoni, we will almost see the simple process of adding one more canzone to each succeeding group.

The pattern is broken by canzoni 4-6 (37, 50, 53) and canzoni 11-12 (105-119), all but one of which is part of the pattern of the madrigali, but we should note that the missing group of three canzoni is in fact supplied by 37, 50, and 53, which cannot be grouped together if the madrigali pattern is to be maintained. The same line of reasoning will explain the apparent intrusion of canzoni 11 and 12 (105 and 119) into the simple additive grouping of canzoni. It should also be pointed out that the largest group of canzoni (125-129) is followed by the one hundredth sonnet, a fact that Petrarch himself noted in his manuscript.⁷

Structural symmetries are not the only use made of numerological composition. The first canzone in the sequence is number 23. The first canzone of Part II is number 264, and it is the twenty-first canzone of the sequence. Are we to ignore the facts that Petrarch was 23 when he first saw Laura and that she died 21 years later after he first fell in love with her? I think not, but I do not want to argue at this point about the meaning to attach to this numerological order or about the difficulties arising from the fact that 264 was probably written before Laura's death in 1348. I simply want to suggest the probability of a more patterned structure in the completed sequence.

And we need not confine ourselves to merely formal patterns; verbal repetitions also help to define the structure. The first poem in Part II and the last sonnet have remarkably similar first lines that are meant to recall each other:

I' vo pensando, e nel penser m'assale (264)
I' vo piangendo i miei passati tempi (365)

A more complicated verbal structuring occurs in the segment 70-81.

The group of four canzoni, as I have already mentioned, is the second longest grouping of canzoni in the sequence. Number 70, which is also the seventh canzone, is a poem of five ten-line stanzas, the last line of each being the first line from poems by Arnaut Daniel, Guido Cavalcanti, Dante, Cino da Pistoia, with the exception of the last stanza where Petrarch quotes the first line of his own 23, the first canzone in the sequence. Numbers 71-73 are the justly famed "Tre Sorelle" canzoni, so called from the comiato of 72:

Canzon, I'una sorella è poco inanzi,
E l'altra sento in quel medesmo albergo
Apparechiarsi; ond'io piú carta vergo.

Each of these poems has the same stanza and rhyme scheme. The underlined words in the comiato of 73 are picked up and used as the first line of 74, the first of that six sonnet segment:

Canzone, i'sento gia stancar la penna
Del lungo e dolce ragionar co llei,
Ma non di parlar meco i pensier mei. (73)

Io son giá stanco de pensar si come
I miei pensier in voi stanchi non sono.…
  (74)

Virtually the same line is picked up once more as the first line of the very next block of sonnets: "Io son sí stanco sotto 'I fascio antico …" (81).

There are many more isolated instances of formal structuring in the Canzoniere, but I do not want to discuss them before establishing the basic calendrical structure. Suffice it to say that my proof of this structure will be based on just such formal considerations as we have been discussing.

It is well known that the Canzoniere consists of 366 poems, divided into two parts, poems 1-263, called In vita di Laura, and poems 264-366, called In morte di Laura, since they deal with the time after her death. Furthermore, there can be no question that Petrarch wanted the major division of the Canzoniere to occur at 264.⁸ Both the evidence from MS 3195 and Wilkins' definitive studies of the accretions to both parts of the sequence (see chart in Wilkins, Making, p. 194) prove that Petrarch had this division in mind as early as 1347.⁹ Nevertheless, there is a genuine problem here since the division, presumably to mark the death of Laura, occurred at least one year before her death. Wilkins summary of the problem is worth quoting in full:

Those who have been troubled by the division have thought of it from the point of view of the reader rather than from the point of view of the creating poet; and have assumed that the division was made after Petrarch knew of the death of Laura. And indeed, if the division had been made after Petrarch had that knowledge, his decision to begin Part II with 264 rather than with 267 [first poem to announce the death] would have been extraordinary. But this consideration in itself suffices to indicate that the decision to begin Part II with 264 was made before Petrarch knew of the death of Laura. If this fact is realized, the making of the division at this point is no longer difficult to understand. 264 is—as it has been called in a previous section of this chapter—a very great and distinctive canzone, expressive of the fundamental conflict in Petrarch's inner life and of a desired reorientation. The canzone was probably written in 1347; it is highly probable that it was Petrarch's intense experience in the writing of this poem that led him, during the composition of the poem or very soon afterward, to decide to use it to mark a major division in the Canzoniere. In any case, the decision, once made, was too firmly fixed in Petrarch's mind to be altered even by the death of Laura.¹⁰

This seems to me an astonishing conclusion. A man who spends a lifetime memorializing the life and death of a woman in poetry, who divides his sequence in vita and in morte, does not ignore one of the two major events of that lifetime in favor of one poem, no matter how intense the experience of writing it, unless there is some poetic priority. The implications of Wilkins' final statement diminish the importance of Laura both for Petrarch's life and for his poetry in ways that cannot be supported either by the text or by his prose statements.

Petrarch insists again and again that he first saw Laura on the sixth of April 1327 and that she died on the sixth of April 1348. We have already quoted the memorial inscription in his copy of Virgil, and he is no less insistent in the Canzoniere:

Mille trecento ventisette, apunto
Su l'ora prima, il dí sesto d'aprile
Nel laberinto intrai; né veggio ond'èsca.
(211.11-14)

And again,

Sai che 'n mille trecento quarantotto,
Il dí sesto d'aprile, in l'ora prima,
Del corpo uscío quell'anima beata.
(336.11-14)¹¹

There is no reason to dispute Petrarch's facts, although the remarkable coincidence of his first sight of her and of her death need not even be true historically. The dates are part of the meaning of the myth he created for himself. On the basis of evidence from the 1336 letter to Giacomo Colonna, from the third book of the Secretum, from the Letter to Posterity, from the memorial inscription in his Virgil, one can say with certainty that Petrarch is telling us that he fell in love with a woman.¹² We can date the occurrence and place it, but we can say very little more. Who is she? What was her name? What was her situation? Questions of this sort we cannot answer from life records, and this has driven critics to the fiction of the Canzoniere to find out the facts, with the result that those supremely metaphorical poems have been squeezed and drained of their vitality to produce biographical fact. This will not work; this is not proof. Laura has a reality that comes from Petrarch's statements and from the poems, but this reality does not require the actuality of existence to make it real. If one is a Christian, one does not have to see the actual Jerusalem to know what the real Jerusalem is. That city has been built aere perennius in the minds of generations of men through words that defy actuality and the temporal and posit a reality undisturbed by actuality because essential. Laura may have existed. Laura may not have existed. Neither one of these possibilities will affect the reality of the myth that Petrarch created. In the passages refered to above, Petrarch seems to indicate the actual existence of Laura, but this actual existence is always subsumed into the greater myth of fourteenth century Christianity, as is evident in the conclusion of the memorial in his Virgil:

But her soul has, I am persuaded, returned to the heaven whence it came, as Seneca says of Scipio Africanus. As a memorial, afflicting yet mixed with a certain bitter sweetness, I have decided to make this record in this place of all places, which often falls under my eyes, that I may reflect that there can be no more pleasure for me in this life, and that, now that the chief bond is broken, I may be warned by frequently looking at these words that it is time to flee from Babylon. This, by God's grace, will be easy for me when I think courageously and manfully of the past's vain concerns and empty hopes and unexpected outcomes.¹³

If Petrarch could accommodate his love and his loss to this Augustinian Christian view, then we too as critics should be able to do as much.

Nevertheless, underlying most of the discussions of the problem of the division of the Canzoniere are the assumptions that Laura actually existed and that she actually died on the sixth of April 1348, and these assumptions force the critics to consider the division at 264 as a real problem. If one is starkly logical about the problem, either Laura existed as a woman or she is a symbolic fiction, either the 1327-1348 dates are real, or one or both of them are fictions. If Laura and the dates are real, then there is an indecorum in not starting Part II with 267 because the structure of the sequence does not follow the facts. If, on the other hand, Laura and the dates are fictions, then the indecorum is reduced to a mere bumbling ineptitude in naming the two parts, because Petrarch could have changed at will the dictates of his fictive world. Neither one of these explanations is satisfactory—nor even binding on the poems.

A third possibility exists: Laura was an actual woman, and Petrarch is writing fiction about her. If we now make the assumption that this fiction depended on a plan for ordering the poems, we will escape the problem that Wilkins' genetic approach imposes on him. What could be so "firmly fixed" in Petrarch's mind except some plan, some intellectual program, that would encompass the events of his experience? Let us start with the assumption that Petrarch did have such a plan when he began the Canzoniere, that this plan became fixed when in 1347 he decided that 264 would begin the second part, that this plan could accommodate even the unexpected death of Laura, and that this plan found its completion in the writing of MS 3195 over the next twenty-six years. This plan, I am suggesting, is the use of a calendrical structure to order the poems, and it is evident only in the completed MS 3195.

If we return now to Petrarch's insistence that he first saw Laura on the sixth of April 1327, we find another instance where a strictly biographical approach to the poems lands us in difficulty. On the evidence of poem 3 (Era il giorno) commentators connect this sixth of April date with Good Friday, but then quickly accuse Petrarch of inaccuracy because the sixth of April 1327 was the Monday of Holy Week and not Good Friday. Nonetheless, we need doubt neither Petrarch's memory nor his calendar. As Carlo Calcaterra has brilliantly shown, Petrarch was not speaking of the annual liturgical celebration of Christ's crucifixion, he was speaking in terms of absolute time, and his scholarly sources told him that Christ was actually crucified on the sixth of April.¹⁴ If the April sixth date of Petrarch's first sight of Laura actually occurred, a man of Petrarch's learning could not have avoided relating this date to other occurrences on the feria sexta. According to some traditions, as Calcaterra has shown, on the sixth day man was created, and on the sixth he fell, and on the sixth Christ's death redeemed that fall.¹⁵ Petrarch would have us believe that this same sixth of April was equally important in his life: on the sixth of April 1327 he first saw Laura; on the sixth of April 1338 he first was inspired to write the Africa;¹⁶ on the sixth of April 1348 Laura died.

The implications of all these sixes leads me to discuss one more April sixth, which I feel is of equal importance to our reading of the Canzoniere as the 1327 and 1348 dates. Good Friday fell on the sixth of April only four times during the fourteenth century—1319, 1330, 1341, and 1352. The third of these—1341—is a most important part of the myth Petrarch was constructing for himself.

On the morning of the first day of September 1340 Petrarch received from the Roman Senate an invitation to be crowned laureate, and on the afternoon of that very same day he received another such invitation from the chancellor of the University of Paris.¹⁷ He chose Rome and asked for the sponsorship of King Robert of Naples, for which purpose he travelled to Naples, reaching there sometime in February 1341. Petrarch was duly examined by King Robert as to his eligibility for this honor and satisfied the requirements. King Robert wanted the ceremony to be performed at Naples, but Petrarch requested that his coronation take place in the Senatorial Palace on the Capitoline in Rome. The request was granted, and armed with a "robe of honor" from King Robert, Petrarch set out for Rome, arriving there probably on the sixth of April. On Easter Sunday, the eighth of April 1341, Petrarch received the triple crown of poeta laureatus from the hands of Orso dell'Anguillara and after the ceremony proceeded to St. Peter's where he placed his crown on the altar.¹⁸

We do not know whether Petrarch chose this precise date, but it is entirely possible that he was most particular about this most spectacular event in his life. Not since antiquity had a poet been crowned in Rome, and only two other laureations had occurred anywhere since antiquity.¹⁹ Could Petrarch not have seen some relation among all these April sixth dates, each one of which was a turning point in his life? At any rate that is the assumption I shall make, and it is an assumption that has profound implications for our understanding of the structure of the Canzoniere.

We are now in a position to discuss the calendrical structure of the Canzoniere, and my hypothesis will, among other things, offer a better reason for Petrarch's dividing the sequence at 264 and will place the whole sequence in that Christian context, without which the poems can only be poorly understood. The hypothesis is quite simple: if we number each poem with a day of the year, beginning with the sixth of April, we will find that when we reach 264 we have also reached the twenty-fifth of December, Christmas Day. In short, I am suggesting that the division of the Canzoniere is based on two of the three most important events in the Christian calendar. Part I, dealing with the inception and growth of his love, begins with the death of Christ; Part II, dealing with the death of his love, begins with the birth of Christ.²⁰

Instead of seeing Petrarch's grieving love as a kind of formalized autobiography, a reading of the sequence fiercely refuted by the Secretum, one might better view it as part and parcel of a fourteenth-century Christian outlook. There is an esthetic distance between Petrarch the lover and Petrarch the poet. It seems to me that Petrarch the poet is saying that he has conceived a passion for Laura that is essentially selfish, that is not the vera amicitia of the philosophers, as at least one of the early commentators agrees.²¹ Petrarch indulges his longings without success, until he learns that Laura has died, when—deprived of the physical object of his desires—he learns to love her truly for her virtues by the end of the sequence. It is thoroughly medieval: subtly intellectualized and probing deep into the nature of man's desires, whether these be for a woman or for glory. The burden of the third book of the Secretum, to which Petrarch gives assent, is that he is still bound by the two golden chains of love and glory, which keep him from the proper love of God, and this subject, it will be recalled, is the major subject of poem 264. The inception of his love on the day that Christ died counterpoints the old Augustinian distinction between caritas and cupiditas, a point made over and over again by the early commentators.²² Beginning Part II of the sequence on the anniversary of the birth of Christ proclaims as well the possibility of rejuvenation and a truer understanding of both earthly and heavenly love. Petrarch did not have to change his plan for the division of his sequence when Laura's death occurred in 1348 because her death simply enriched the metaphorical significance of the basic opposition of caritas and cupiditas: death of Christ—birth of love for Laura; birth of Christ—death of Laura. And that is why Petrarch did not alter his decision to begin Part II with 264.

To test my hypothesis about the Good Friday-Christmas dating of 1 and 264, let us consider other times of the Church year as possible structural devices. This can best be done by considering the placement of nonsonnets in the sequence (i.e., canzoni, sestine, ballate, and madrigali). There are eleven nonsonnets in Part II, and they are by my calculations associated with the following dates:

Let us begin with 270 since we have already made a case for 264; and 268 is clearly a meditation on the death of Laura announced in the preceding poem. Poem 271 begins a segment of 52 sonnets. The date I attach to 271 is January first, New Year's Day, and I would like to suggest that symbolically those 52 sonnets form a year of mourning. This block of sonnets, the second longest in the sequence, is followed by the three non-sonnets 323-325, another structural break introducing the last forty poems of the sequence. I cannot believe that it is mere coincidence that 326 is associated with 25 February, which was the date of Ash Wednesday in 1327. The last forty poems form a symbolical forty days of Lent, leading up to 366, that great hymn to the virgin, which is associated with 5 April, which in 1327 was Palm Sunday, the day of Christ's triumphal entry into Jerusalem.

Within that forty-poem segment there are two significant formal breaks: 331-332 and 359-360. Number 331 is associated with the first of March, an alternative beginning of the year, and it is followed by a sestina and 26 sonnets, or a half-year cycle matching the 52 sonnets from 271-322. Number 359 is associated with 29 March, which was Passion Sunday in 1327, the beginning of Passiontide.

Now it may be objected that I have not accounted for all the non-sonnets in Part II, as indeed I have not. The reasons for this neglect are two-fold. One, not all the formal breaks are to be explained by the calendrical form. Two, I have simplified the scheme I believe Petrarch was using. Thus far I have pointed out only those correspondences that would have occurred in 1327. But as I have suggested earlier, Petrarch had in mind at least two other years in which the sixth of April was important: 1342, the year of his coronation, and 1348, the year of Laura's death. These three years are counterpointed against one another to complicate the formal breaks in the sequence. In the following chart I will try to show how this counterpointing is sometimes significant and sometimes not.

It must first be pointed out that of the three years in question only 1348 was a leap year, a reason for the Canzoniere's having 366 poems rather than 365. It will be seen from the chart that the two Ash Wednesdays of 1327 and 1341 frame, as it were, the three non-sonnets 323-325. It will also be seen that the late Ash Wednesday of 1348 does not participate in the formal structure.

A similar use of the three years can be demonstrated in the first two formal breaks of the sequence.

No one has ever advanced a reason for Petrarch's making 11 and 14 ballate, but it would not be out of the question to suppose that there is a change in form at these points to call attention to the occurrences of Good Friday and Easter in 1327 and 1348, dates which frame the ballate as the two Ash Wednesdays framed 323-325.

Another use of the liturgical calendar is Advent, the little Lent, the season of preparation for Christmas. Again, a chart will help.

The two sestine frame the Advent of 1327, and the other two years, so close, are not used in the formal structure.

One final example may help to clarify the reasons why Petrarch sometimes uses more than one date and sometimes does not. We began our discussion of the formal structure of the Canzoniere with what I called the "simple example" of the madrigali pattern, 38-54 and 105-121, in which I tried to show that there was a formal symmetry that obtained between the two segments containing the four madrigali and also how this symmetry was integrated with the arrangement of the first seventeen canzoni. This pattern is, I believe, also integrated into the calendrical structure of the sequence for the feasts of the Ascension and Pentecost.

It will be seen from the chart that 52-55, which is the longest group of non-sonnets of various forms in the sequence, begins with the Pentecost of 1341 and that the segment of three ballate separated by groups of three sonnets is signalized after the first and third ballate (56 and 64) by the Pentecosts of 1327 and 1348. The occurrence of the Ascension dates with 42 and 46 seems to me insignificant, although the coincidence of 54 with the Ascension of 1348 may have a significance of which I am still unaware. What is of importance in this segment is that it is the first time that there is some verbal hint about the correlation between the poem and the date. Poem 53, the canzone that comes between the first two madrigali, whose number of lines adds up to the total of the two surrounding numbers (52 and 54) and is matched by the number of the third madrigale (106), comes immediately after the poem associated with the Pentecost of 1341 and begins with a phrase never fully explained by the commentators, "Spirto gentil." What is this Spirit if not the Holy Spirit That rules the members of Christ's church on earth and tries to guide the restoration of the ancient seat of that church in Rome through the agency of Cola di Rienzo?²³

Spirto gentil, che quelle membra reggi
dentro a le qua' peregrinando alberga
un signor valoroso, accorto et saggio,
poi che se' giunto a l'onorato verga
colla qual Roma et suoi erranti correggi
et la richiami al suo antiquo viaggio
io parlo a te, però ch'altrove un raggio
non veggio di vertú, ch'al mondo è spenta,
né trovo chi di mal far si vergogni.
(53.1-9)

In conclusion, one must say that the calendrical frame-work will not explain all the formal breaks in the sequence, nor will it assist (with the possible exception of 54) in the reading of individual poems. Nevertheless, it must be conceded that the probability of mere coincidence of poem and day is offset by the large number of parallels (approximately 85 percent of the non-sonnets) that adds a new dimension to our reading of the Canzoniere, and that this new dimension sets the Canzoniere firmly in the context of fourteenth-century Christian morality. The Good Friday-Christmas division of the Canzoniere sets up a pattern of parallels between the two parts. In Part I the two ballate (11 and 14) are meant to emphasize the Good Friday-Easter dates of 1327 and 1348. Part I ends with the two sestine (237 and 239) emphasizing the beginning of Advent followed by 24 sonnets in preparation for Christmas. In Part II 270 sets up the segment of 52 sonnets beginning on the first of January, which leads to the non-sonnet group 323-325, framed by the two counterpointed Ash Wednesdays of 1327 and 1341, which leads us into that symbolical Lent of the last forty poems, a parallel to the Advent segment of Part I. Thus the beginnings and the ends of each part would seem to be formally structured around four major events in the Christian year: Good Friday, Advent, Christmas, and Lent. Against this annual cycle Petrarch counterpoints the agony of his earthly love and his growing awareness of the disparity between it and the heavenly love he ultimately sought.

Notes

1. See especially Alastair Fowler, Spenser and the Numbers of Time (London, 1964) and Triumphal Forms: Structural Patterns in Elizabethan Poetry (Cambridge, 1970); Silent Poetry: Essays in Numerological Analysis, ed. Alastair Fowler (London, 1970); Christopher Butler, Number Symbolism (London, 1970). For the earlier period see the admirable summary study of Edmund Reiss, "Number Symbolism and Medieval Literature," Medievalia et Humanistica, n.s. I (1970), 161-74.
2. Inscription in Petrarch's Virgil now in the Ambrosiana in Milano. It is reprinted in Carducci and Ferrari, Le Rime (Firenze, 1899), p. 370: "Laurea, propriis virtutibus et meis longum celebrata carminibus, primum oculis meis apparuit sub primum adolescentie mee tempus, anno Domini M° IIJ° XXVIJ° die VJ° mensis aprilis, in ecclesia sancte Clare Avinoni, hora matutina; et in eadem civitate, eodem mense aprilis, eodem die sexto, eadem hora prima, anno autem M° IIJ° XLVIIJ°, ab hac luce lux illa subtracta est, cum ego forte tunc Verone essem, heu! fati mei nescius. Rumor autem infelix per litteras Ludovici mei me Parme repperit anno eodem, mense maio, die XIX° mane. Corpus illud castissimum ac pulcerrimum in loco Fratrum minorum repositum est ipso die mortis ad vesperas.…" A slightly different translation from mine appears in Morris Bishop, Petrarch and His World (Bloomington, 1963), p. 62.
3. Ernest Hatch Wilkins, Life of Petrarch (Chicago, 1961), pp. 2, 5.
4. Ernest Hatch Wilkins, The Making of the "Canzoniere" and Other Petrarchan Studies (Roma, 1951), p. 93, hereafter cited as Wilkins, Making. Ruth Shepard Phelps, The Earlier and Later Forms of Petrarch's Canzoniere (Chicago, 1925).
5. See in particular Aldo S. Bernardo, Petrarch, Scipio, and the "Africa" (Baltimore, 1962), Chapter IV, "Scipio vs. Laura," pp. 47-71.
6. Wilkins, Making, p. 266.
7. Wilkins, Making, p. 120.
8. The division of the Canzoniere at 264 is evident in Petrarch's manuscript, MS 3195, from the facts that (1) there are some blank sheets left between poem 263 and 264 and (2) only poem 1 and poem 264 have elaborate initials. We know further that Petrarch was consciously aware of placing poems in some order because he left some blank spaces in MS 3195, which were later filled in by a different hand and in a different ink. This system of transferring poems to specific places in MS 3195 is corroborated by another Vatican manuscript, MS 3196, entirely in Petrarch's hand, which was apparently one of his workbooks. In this manuscript he often makes a carefully dated notation that he has finally transcribed this poem in ordine in MS 3195. We know also that Petrarch was consciously aware of the placement of the various kinds of poems he included because he has placed a C opposite the one hundredth sonnet, a CL opposite the one hundred fiftieth, a CC opposite the two hundredth and so forth (for the misnumbering of later sonnets see Wilkins, Making, p. 122). One might also point out that 183, the midpoint of the entire sequence, is also the one hundred fiftieth sonnet. Or we could point to the regular occurrence of groups of six sonnets in conjunction with a sestina: 30 with 31-36, 74-79 with 80, 136-141 with 142 and 143-148, 208-213 with 214.
9. One such small formal structuring appears even if we accept Wilkins' and Miss Phelps's assumption about a roughly chronological ordering of the poems. Of the self-dating poems (30, 50, 62, 79, 101, 107, 118, 122, 145, 212, 221, 266, 271, 278, 364) all are sonnets except for the first two. Number 30, a sestina, is the seventh non-sonnet in the sequence and contains the phrase "oggi ha sett'anni," the seventh anniversary of his first sight of Laura. Number 50, the fifth canzone and the ninth non-sonnet of the sequence, contains the phrase "ben presso al decim'anno." Counting a poem both as canzone and non-sonnet may seem like double-dealing, but the fifteenth and sixteenth century editors sometimes discriminated among the various non-sonnet forms in numbering and sometimes labelled all non-sonnets "canzoni." For our purposes it is entirely justifiable to call a poem a "sestina" in calculating one formal structure and to count it merely as a non-sonnet in calculating another formal structure because, as I intend to show, the various structures overlap.
10. Wilkins, Making, p. 193.
11. See also Triumph of Death, 1. 133 and Canzoniere, 325. 13.
12. The letter to Colonna is Fam. II. 9 and is translated and abridged in Morris Bishop, Letters from Petrarch (Bloomington, 1966), pp. 29-33. The Secretum was translated by William H. Draper as Petrarch's Secret (London, 1911). The Letter to Posterity was translated by Morris Bishop, Letters from Petrarch, pp. 5-12, but see esp. p. 6. The Latin texts with Italian translations can be found in Francesco Petrarca Prose, ed. Martellotti, Ricci, et al. (Milano, n.d.).
13. Translated by Morris Bishop, Petrarch and His World, pp. 62-3. The Latin text as cited in Carducci is: "animam quidem eius, ut de Africano ait Seneca, in celum, unde erat, rediisse mihi persuadeo. Hec autem ad acerbam rei memoriam amara quadam dulcedine scribere visum est hoc potissimum loco qui sepe sub oculis meis redit, ut scilicet cogitem nihil esse quod amplius mihi placeat in hac vita, et, effracto maiori laqueo, tempus esse de Babilone fugiendi, crebra horum inspectione ac fugacissime etatis existimatione, commonear: quod, previa Dei gratia, facile erit, preteriti temporis curas supervacuas, spes inanes et inexpectos exitus acriter ac viriliter cogitanti" (p. 370).
14. Carlo Calcaterra, La "Data Fatale" nel Canzoniere e nei Trionfi del Petrarca (Torino, 1926). Reprinted in Nella Selva del Petrarca (Bologna, 1942).
15. Calcaterra, p. 30.
16. The commentators make the same mistake about the inception of the Africa as they do about the April date in Canzoniere. Morris Bishop, Letters, p. 8, translates the day as Good Friday, but the Latin in Posteritati is "sexta feria": "sexta quadam feria maioris hebdomade" (p. 12). It should also be pointed out that in 1338 the sixth of April was the Monday of Holy Week and not Good Friday, exactly as in 1327.
17. The letter is Fam. IV. 4 and is translated by Bishop, Letters, pp. 51-2. See the account in Bishop, Petrarch and His World, pp. 160-71 and Wilkins, Life, pp. 24-9, but note that Wilkins erroneously reverses the order of arrival of the two invitations.
18. See references in preceding note and Wilkins, Making "The Coronation of Petrarch," pp. 9-69. Petrarch's oration is translated by Wilkins, PMLA, LXVIII (1953), 1241-50.
19. Albertino Mussato had been crowned in his native Pavia in 1315, and Dante had been crowned posthumously.
20. The Good Friday-Christmas division of the Canzoniere was first pointed out by Angel Andrea Zottoli, "Il numero solare nell'ordinamento dei 'Rerum vulgarium fragmenta,'" La Cultura, VII (1928), 337-48. Zottoli's argument is radically different from the one presented here. I am indebted to him only for the coincidence of Christmas and 264.
21. Antonio da Tempo in his commentary on poem 1. His commentary was often reprinted following the commentary of Filelfo, e.g., the edition of Venice 1503.
22. The early commentaries have not been studied with sufficient attention to their impact on directing the reader's response to the poems, but see Bernard Weinberg, "The Sposizione of Petrarch in the Early Cinquecento," Romance Philology, XIII (1960), 374-86. I am at present at work on a study of the fifteenth and sixteenth century commentaries and academy lectures as a guide to understanding the reading habits of the Renaissance.
23. A useful discussion of the difficulties of this opening passage may be found in the notes of Chiorboli's edition of the Canzoniere (Milano, 1924).