VOL. VIII

In the Rochester Democrat and Chronicle of October 31 last there was an account of the fifth quarterly meeting of the Monroe County Classical Association. The main subject discussed was the question, What Steps can be taken to introduce Greek into the Smaller Schools of Monroe County? Dr. Alfred C. Thompson, Principal of the Brockport Normal School, made a vigorous attack on the "present day vocational propaganda that leaves out of account the training of the thinking powers". In support of his declaration that the idea that a boy's ultimate vocation can be accurately determined when a boy is twelve to fourteen years of age is wholly fallacious, he presented the results of a special investigation.

About eighteen years ago, as superintendent of a school in Massachusetts, he inquired into the vocational preferences of 467 boys between the ages of 12 and 14. After using the resulting material for the purpose he had then in mind, he laid it aside and forgot about it. When, some five years ago, the material came suddenly to light, Dr. Thompson realized that he had in it the basis of an unusually valuable investigation.

He wrote to well-known citizens of the town in question and secured information regarding the actual vocations of 406 of the original 467 boys. Only five of the whole number had in fact entered the vocations for which, as boys of 12 to 14, they had expressed a preference.

Yet, if the ideas of some of the present-day vocational faddists had then been followed, those early and changeable preferences would have determined finally the boys' ultimate profession, and their future education would have had as its center some fleeting ambition, which, as can now be seen, was of no permanent influence on the lives of these boys.

Dr. Thompson further contended that, even if such a rigid determination of vocation in advance were possible, it is contrary to all the principles of democracy that any attempt should be made to divide our education into education for the masses and education for the classes. From several scores of manufacturers, from whom he sought to learn what qualities they most desired in their employees, he received, in varying forms, one answer-the ability to use their hands and the ability to think.

In this real sense of the term, Greek is preeminently a vocational subject. The narrow interpretation of the term vocational as meaning manual dexterity

No. 11

alone, a definition implicit in the ideas of many who are urging vocational studies with such intolerable insistence, cannot be defended. The ideal of education which it represents cannot maintain itself in this country.

It was announced at the meeting that hereafter the Normal Schools of New York State will not insist that students applying for admission to such schools must have three years of history, but will accept the last two years of a three year course in Greek in place of two of the three years of history. This ruling, reenforced by the giving of credit again in the Regents' examinations for one year of Greek, may help the cause of Greek in the last year or two of the High School.

The results of Dr. Thompson's special investigation reminded me of an investigation of matters relating to industrial training, to which reference was made in THE CLASSICAL WEEKLY 7.217. In that investigation Dr. Ayres, of the Division of Education of the Russell Sage Foundation, found that there has been little understanding of the extent to which "our national migratory instinct plays havoc with the conditions governing labor and education. Only one father in four was born in the city where he now lives, and only a few more than one-half of the boys were born where they now live". These facts "certainly upset the prevailing impression that schools should shape their courses with the predominant aim of preparing the children to enter local industries".

In conclusion, reference may be made to a very valuable article, Latin as a Vocational Study in the Commercial Course, by Mr. Albert S. Perkins, of the High School, Dorchester, Massachusetts, which appeared in The Classical Journal 10.7-16, a reenforcement of his earlier paper on Latin as a "Practical" Subject, The Classical Journal, Volume 8. Both articles deserve careful study. C. K.

A GRAMMATICAL EXCURSION* Many have been the attempts to cultivate that artificiosa memoria to which Cicero, the Auctor ad Herennium and Quintilian devote whole pages1. The father of the art, Simonides, is the hero of a gruesome after-dinner story of the Thessalian banquet, from which the poet was called to the door just in time to

*This paper was read at the Eighth Annual Meeting of The Classical Association of the Atlantic States, at Barnard College, April 17, 1914.

Cicero, De Oratore 2.350 ff.; Ad Herennium 3.28 ff.; Quintilian 11.2.

escape the fall of the roof, which mangled the rest of the diners so completely that the sorrowing relatives had no other means of identifying the fragments than their confidence in the mnemonic art of Simonides, which enabled him to name and place every guest2. Thus he was said to have discovered the value of the associative chain, or to have discovered, in Cicero's words, ordinem esse maxime, qui memoriae lumen adferret. But, though the method explained in the texts just cited might serve the budding orator, the boy in the thick of grammar could make small use of their loci, or imaginary stations, and their imagines, associated in thought with the loci, and so recalled as one travelled in mind from one locus to the next, like the pious pilgrim of to-day halting before the stations of the Cross. The school-boys' penance, at least under the later Empire, took the form of the memory-verse, not merely for the facts of grammar, but for historical material as well. So far as I am aware the earliest surviving specimens are those of Ausonius1 and of the Anthology. For Terentianus Maurus can hardly have intended that his entire treatise on meter in verse-form should be committed to memory. Yet he may have had this in mind for the more important statements.

In the mediaeval mind the versus memorialis found congenial soil, sprang up everywhere, and cast its shadow over many fields of learning. The moods of the syllogism were reduced to grotesque jargon in verse, simply that the proper 'record' might be always ready for the debating machine, the 'Victor' of the Schoolman. But it is with grammar alone that we are here concerned, and we turn at once to the humble priest who first versified the Latin grammar, not as a tour de force, but lest his readers forget.

Alexandre de Villedieu", better known as Alexander de Villa-Dei, was born about 1170, in Normandy, at the little town of Villedieu (Manche), a dozen miles northeast of Avranches, and about eighteen east of Granville. He studied in Paris, where he heard lectures on Priscian, and with two friends as poor as himself he began reducing the teachings of that grammarian to memory-verses. Later he was called away by the bishop of Dol, on the borders of Brittany, to instruct his two nepotuli in grammar. It was for these pupils that he completed his metrical rules; with the constant encouragement of the bishop they were published, probably in or about 1199. Alexander lived on into the middle of the century, and died as

2Cicero, De Oratore 2.352 f.; Quintilian 11.2.11 ff. a De Oratore 2.353.

'De Litteris Monosyllabis Graecis ac Latinis, in the Technopaegnion (Peiper, page 166); compare the Monosticha de Ordine Imperatorum, and other memory verses on the Roman emperors (pp. 184-193). The Ordo Urbium Nobilium and the Ludus Septem Sapientum may have been meant for memorizing (144 ff., 169 ff.).

Reichling, Das Doctrinale des Alexander de Villa-Dei, Berlin, 1893 (Monumenta Germaniae Paedagogica, XII). Compare Sandys, A History of Classical Scholarship, 1.532, 585, 640.

❝Studuerunt Parisiis et tum audierunt regulas Prisciani; componebant per versus. So states a gloss cited by Reichling, p. xxiii. Paupertate gravati. multa libris extracta metrice compilaverunt, states another gloss, ibidem.

canon of S. André, at Avranches'. By that time his Doctrinale had gathered to itself a mass of scholia in prose, as we see from the oldest extant manuscript, of 1259, now in the Laurentian library at Florence. Difficult as it is for the modern mind to comprehend how a work of this type could have become so popular, we are forced to admit that the mediaeval schoolmaster thought differently. For no less than 33 MSS of the XIIIth century are still extant, and 217 more of later date are registered by Alexander's modern editor, Reichling, with the certainty that his list is by no means exhaustive. By 1500, 163 editions had been printed, out of a total of 279 known to Reichling. No less than a hundred presses issued the work, among them many of the most celebrated.

The form adopted by Alexander was the favorite leonine hexameter, so much used in that day for the versus memorialis, as well as for hymns and other compositions. Here and there a line lacks the rhyme at the caesura. In this case, however, adjacent lines are often linked by end-rhyme. Thus it was hoped to reduce the difficulty of committing to memory-a most important consideration in view of the scale of the work. The total of verses reaches the formidable number of 2645, of which 1073 are in the first part, containing accidence, 476 in the second, on syntax, 732 in the third, on prosody, 364 in the fourth, on accent and figures. The rules for gender cover 195 lines. It must not be supposed that this was a beginner's book. Donatus was to be left in undisturbed possession of that field, while an introduction to the Doctrinale was provided in the Alphabetum Minus, a lost work of Alexander, who in the larger book assumes that the pupil has mastered the facts, so far at least that he can now make use of desiccated memory-verses in order to retain the rules and other statements. The aim was not to compose a didactic poem, after the manner of Terentianus Maurus, though there are lines which are sufficiently clear to be called didactic. The rest, that is the vast majority, were framed as mere aids to memory, to be recalled singly, or in groups, without context. Thus the apt pupil might provide himself with weapons to dazzle the adversary in any grammatical debate. The enquirer could at once be struck dumb with amazement at a mysterious response glibly reeled off. The cabalistic language of the oracle only added to the impression of superhuman knowledge, and we can imagine the uncanny impression created by these latter-day Alexandrians, who carried Priscian around in their heads. Not the least effective in reducing an opponent to silence would be those abracadabra lines which give lists of verbs reduced to their first syllables, and scanned with easy indifference to quantity9.

"Mentioned in verse 1346 of the Doctrinale (Abrincae, -arum.) Compare line 26.

Compare e.g. 700, cre. do. do mi. iu. sto pli. fri. so. ne. ve. la. se. cu. to a list of exceptions to avi in the perfect of the first conjugation. The succeeding lines give the clue by naming (in such order as the verse permitted) the verbs: crepo, domo, do, mico, iuvo, sto, plico, frico, sono, nexo, veto, lavo, seco, cubo, tono. Compare 759, 766, 778, 975.

The first chapter begins with the case forms of the first declension:

Rectis as es a dat declinatio prima,

atque per am propria quaedam ponuntur Hebraea, dans de diphthongon genetivis atque dativis. am servat quartus; tamen en aut an reperimus, etc. (29-32).

These lines fairly represent the general method, which permits many harsh constructions, not to mention liberties with the verse. Many lines are hopelessly obscure, until one has compared the pertinent passage in Priscian. But, as we have already seen, exposition is not the main purpose. But some of the sins which were ruthlessly scored by the humanists may be more charitably covered under the mantle of a mnemonic intent.

Of the rules for verb forms the following may be cited as typical:

tu geminans tundo tutudi dat sumque supino. dat tundo tutudi, donat tondere totondi; tondeo dat tonsum, sed praebet tundere tusum. datque cado cecidi, format rudo quoque rudi (829–832) ̧ An important part of the work is the syntax, from which, however, the whole subject of moods and tenses is omitted. The following may be taken as a specimen:

sunt ablativi plures rectore soluti: discere discipuli debent doctore legente.

atque loci nomen quandoque regente carebit: Roma, militia vel humo vel rure domoque

rex venit, etc. (1339-1343).

Next to Alexander stands his contemporary Eber hard of Béthune, variously known as Eberhardus or Ebrardus Bethuniensis, or Évrard de Béthune (a little town in Flanders). Of his life little or nothing seems to be known. Though he wrote against the Waldensians, and perhaps composed a poem on the woes of teachers of rhetoric, entitled Labyrinthus, he was chiefly known by his Graecismus1o, published probably in 1212. The work belongs to the same class with the Doctrinale, and follows the same methods. Its scope is ampler, including much of rhetoric, derivation of words, especially from the Greek. To this last feature it owes its very misleading name.

The Graecismus is preceded by a prose preface-a marvel of bombast-, setting forth the fatuitas asinina, the hybrid and impossible language of the ignorant, upon whom he had taken pity, in venturing to explain for them the real differences between the meanings of words. Following the order of Donatus he will treat of the noun, the pronoun, of figures, of barbarism and soloecism, etc., finally, (in alphabetical order), of words borrowed from the Greek. This done, the author enters upon an interminable series of versus memoriales, after the pattern of Alexander. After completing in 752 verses the subjects outlined in the preface, he takes a fresh start-the only break in the manuscript

10Wrobel, Eberhardi Bethuniensis Graecismus (Breslau, 1887), an elaborate critical edition; Sandys. A History of Classical Scholarship, 1.532, 622, 640; Reichling, op. cit. lxxiv f., lxxix ff.; Epist. Obsc. Vir. (Stokes), 209.

upon which I have based my study of Eberhard. This second part begins with the subject of gender, · to which 872 lines are devoted. Nouns often confused claim 462 verses, the four conjugations 571, verbs often confused 133, the different species of nouns 330, while at the end stands a brief chapter on syntax, 64 lines. The second part alone, with its total of 3794 verses, is considerably longer than the entire Doctrinale of Alexander (2645), while the whole book contains 4546 verses in Wrobel's text. In the derivations from the Greek the elegiac couplet prevails, elsewhere the hexameter with occasional pentameters. A number of identical lines show that Alexander and Eberhard probably drew in part upon older material, derived from their teachers".

It will have been noticed that, besides the matters usually to be found between the covers of a grammar, the Graecismus includes within its range much that belongs to the lexicon, dealing especially with groups of nouns and verbs likely to be confused by the student. Here too the intention was not to provide a work of reference, but simply a mass of memory-verses, to be learned by heart, in the hope that at the proper moment the pertinent verse would occur to the mind, just in time to prevent an error, or to insure a triumph over your adversary.

Eberhard's work circulated everywhere, and became only less well-known than that of Alexander. The Bibliothèque Nationale has eight editions down to 1500. In England the book was much less in vogue, and the British Museum has but two editions (1487 and 1490?). In this country it appears to be a very rare work, accessible to me at first only in a fourteenth century manuscript lent to me by Mr. George A. Plimpton from his valuable collection. The examples to be presently cited are accordingly printed as they stand in this manuscript, but references to chapter and line in Wrobel's Breslau edition (1887) are added. Naturally in a work of this type the MSS show many variant readings—even insertion of entire lines or groups of lines from Alexander or other sources.

Here is a specimen of a gender rule:

Feminei generis sunt haec que noscere quibis pax arx calx calcis cos et dos falx quoque falcis lux lex fraus fraudis simul et laus accipe laudis glos et frons frondis gens fax fex frons quoque frontis (p. 12; Wrobel, 6.9).

Here are some distinctions:

Liber id est Bacchus vel vir sine compede natus
at liber est codex vel raptus ab arbore cortex
(32; 9.171).

Dicas textoris radium radium quoque solis atque rote radius tibi sit non invenio plus (34; 9.224). Populus est arbor populus collectio gentis populus est aliquid sed nichil est populus (55; 12.14). Observe the insignificance of the many! Among amusing derivations one may find such as these:

"Reichling, op. cit. lxxxii.

Est mare sive salum pelagus quoniam sit amarum
a pello pelagus quia pellunt littora fluctus
equor ab equando cum nil turbatur in ipso

a potu pontus ideo citus occeanus (57; 12.67).
Here is a key to the group clava, clavus, clavis:
Clava genus baculi clavis sera clavus acutus
clava ferit clavis aperit clavus duo iungit
clavus remigium vestis quoque sit tibi clavus
(62; 12.208).

Here are more distinctions:

Est sompnus proprie dormitio continuata sompnia sunt ea que per sompnum nocte videmus est imber pluvia nubis caligine tetra

est nimbus tenuis quo nos aspergit aprilis (73; 12.369).

Te vestis pulcrum reddit natura decorum (77; 13.19) Finally I cite:

Explicit ebrardi grecismus nomine XRi

qui dedit alpha et o sit laus et gloria XRo (27.62).

For three centuries, Alexander, and in a more limited way Eberhard, had almost complete possession of the Latin grammar market. The attacks of the humanists had little more effect at first than to call forth new editions of Alexander, with alteration or omission of defective lines, difficulties smoothed out, and other concessions to the demands of a more critical age.

As an example of these Renaissance revisions I may cite that of a Dutch scholar, Hermannus Torrentinus, in a Tübingen edition of 1514, containing the first part of the Doctrinale, reduced by the omission of 171 lines; also the second part unabridged. It was necessary to atone in an Apologia for the editor's temerity in laying hands upon a text hallowed by three hundred years of almost undisputed use. In this we learn that by the efforts of Aldus Manutius and others the Doctrinale has been 'exploded' from the schools of Italy a task which he regards as still hopeless in his own country12. He has bitter memories of his own boyhood, and the woeful task of memorizing Alexander. Nor can he believe that the younger generation should 'toil as we have done'. Why be jealous of those who may travel on wheels where we went on foot? Now that the revival of learning has come, one should not be confined with the minotaur in Alexander's labyrinth, hoc est frivola discere, et nunquam perdiscere. In fact he claims by his revision to have slain the beast and to have made the labyrinth a highway13.

From the Dutch scholar who admits that, in the age of Maximilian and Henry VIII and Louis XII, Alexander is still so firmly entrenched in the schools of the

12Quorum opera effectum est, ut Alexandro doctrinale quod vocant italiae scholis explosum sit. Quod ipsum quum nostris in regionibus factu difficile ac pene impossibile videtur, etc.

Ipsi namque olim pueri non sine malo nostro experti sumus, quantus labor sit Alexandrum ediscere et ad unguem intellegere, quod plaerique intra septennium vix possunt. At dicet aliquis laborent adolescentes ut nos fecimus, nec ocio marceant. Quid hoc aliud est quam si doleas alios illac carpento vehi, qua tu pedibus iter fecisti?. Nunc autem deo propicio humanitatis artes revixerunt. ..ut nunc non oporteat barbariem legere, non oporteat in Alexandri labyrintho apud minotaurum includi, hoc est frivola discere, et nunquam perdiscere. Nullus igitur Alexandrinum quem dixi labyrinthum vereatur ingredi, quem Thesei exemplo pervium reddidi, et latitantem in co minotaurum iuvenesque devorantem necavi.

North that a mild revision alone is possible, we turn to the printer-scholar, Aldus Manutius (1450–1515), and his Institutionum Grammaticarum Libri Quatuor, the preface of which is dated at Venice in 1507. He has himself been brought up on Alexander, and hence protests all the more eloquently against the whole method of laboriously learning what is quickly forgotten, instead of memorizing standard texts such as Cicero and Vergil14. His own grammar has the catechism form, and contains memory-verses, but they are not numerous relatively. It is a totally different plan from that of Alexander. As an example may be cited the following for third declension adjectives of three endings:

Syluester. uolucer. celer & campester. & acer.
Atque alacer. celeber. mediocre. saluber. equester.
Sicque paluster, item sic inclinato sequester.
Hic sed & haec alacris, celebris, mediocris & acris
Atque per e neutrum dices genus. Adde salubris
(p. 10 recto, ed. Ven. 1549).

A rule for -im in the accusative of the third declension nouns fills ten hexameters, while twenty-three are needed for the plural genitives, -ium vs. -um (11 v.-12 r.; 14 v.). A reminiscence of Eberhard de Béthune is perhaps to be detected in the use of elegiac couplets occasionally, as those for heteroclites (p. 23 r.), ending with a sentiment which might have served as a motto for the whole book,

teneris pauca notasse sat est.

Evidently the gender rules and exceptions of Alexander had caused Aldus most of his boyish woe, for he has himself no memory-verses for gender. Some of his metrical rules have to do with principal parts, as this, which, in its bisection of plico, may recall the abracadabra already cited from Alexander:

Cum domo iunge mico. cubat, & neco cum frico. cum pli.

Et tonat, atque sonat. secat, & crepo, nexo: vetoque, Doque, dedi, & laui lauo. sto, stetit. & iuuo iuui (p. 41 v.).

There are metrical rules for forms and construction of verbs, for lacking supines and participles, for grammatical figures. But on the whole the book bears out the doctrine laid down in the preface, and reduces enormously the burden upon the memory of the pupil.

North of the Alps progress came more slowly. Not only was Alexander revised, as by Torrentinus,

14Inst. Gram., Pref.: nequid nisi doctissimorum authorum ediscere cogatis adolescentulos. imo ne grammaticas quidem regulas, nisi compendia quaedam brevissima, quae teneri facile memoria queant, laudo eos ediscere, sed tamen ut illas assidue, accurateque legant, nominaque et verba declinare optime sciant. Nam dum lucubrationes nostras vel carmine, vel prosa oratione etiam de arte commendare memoriae eos cogimus, erramus, (ut mihi quidem videtur), multis modis. Primum quod quae summo labore edidicerunt, dediscunt paucis diebus, quod ego et puer olim, et iuvenis, compositis etiam a me regulis, sum saepe expertus, nam cum generum regulas praeteritorumue summa cura memoriae mandassem, perbrevi eas obliviscebar. . . .Tum eo ipso tempore, quo nostra ediscunt, facilius meliusque uel Ciceronis_aliquid, uel Virgilii, aliorumue illustrium possent discere. . . .Equidem puero mihi, cum Alexandri carmen de arte grammatica, praeceptore cogente, memoriae mandabam, non ita contigisse, plurimum doleo. Addite quod cum incultos, et barbaros discimus, tales, ac potius incultiores et ipsi euadimus.

already noticed, but upon the Doctrinale was based the grammar of Despautère, which held sway in the French schools for more than a century, and in some of them continued in use until after the middle of the XVIIIth century.

Jean Despautère was born in Brabant (ca. 1460), at Ninove, and hence was called Jean le Ninivite. He taught at Louvain and elsewhere, dying at Commines in 1520. The oldest edition of his Grammar accessible to me is that of Ascentius (Lyons, 1536), in two thin volumes, now in the library of Teachers College. In his preface to the first part Despautère speaks of those who have endeavored to dethrone Alexander, naming Aldus among others; also of those who, as Torrentinus, have improved him. Result, a bewildering variety of text-books in the schools! All of which is non sine maximo adulescentulorum damno: qui ad scholam aliam quoties mittuntur: priora frustra didicisse censetur (sic) omnia: nec erit rei tam pernitiose finis ulla: donec unus reliquis preferatur: hic quis erit: novit deus. This preface contains a few slings at Alexander, and the claim that the metrical rules concerning this first part of the grammar are briefer than Alexander's by four hundred lines, and citra obscuritatem. The rules of Despautère are in hexameters, and the entire first book (just 100 lines) is devoted to gender, beginning:

Omne viro soli quod conuenit esto virile.
Omne viri specie pictum vir dicitur esse.
Esto femineum recipit quod femina tantum.
Femina dicatur facie pictum muliebri.

Quicquid vtrique datur: commune locatur vt exul. The second book takes up the declensions, the third heteroclites, the fourth comparison, the fifth, sixth and seventh verb forms. Whole of diffuse pages prose comment are interlarded among the metrical rules, not without attacks upon the correctness of Alexander's statements. For this temerity he defends himself at the end in an Epistola Apologetica. In the preface to the Syntax (1513) he sides with Alexander against his carping critics, and charitably quotes the judgment of Quintilian on Accius and Pacuvius: Illum videlicet rudem fuisse non sua culpa sed temporrum, etc1. The Syntax consists of a long eighth book De Concordantia Et Regimine, and a few pages (book nine) on word-order. The versified rules are few in number comparatively, and almost lost in voluminous comment and scholastic question. An Ars Versificatoria on similar lines makes a tenth book in some editions of Despautère, who was early edited and abridged, as by Sebastian Novimola of Duisburg on the Rhine (Cologne, 1534, Teachers College library). This was evidently a drastic revision, and the editor, while paying a tribute to the general excellence of the book, adds1 that the author has neglected Horace's advice to be brief: Cum autem tot scholis sit receptus, & vix ullus, tum discipulus, tum etiam praeceptor

15Compare Quintilian 10.1.97. 16In his preface to the Syntaxis (Dillingen, no date, bound with the last).

reperiatur, quin ex immodica commentariorum diffusione nauseam contrahat, diu summis desideriis aliquem expectaui, qui Syntaxin diligenter accisam, in communem vtilitatem exhiberet". Another abridgement of Despautère is that of Pellisson18, 1529, accessible to me in a Venice edition of 1613. Although it is difficult to see how Despautère is less arid, or easier to commit than old Alexander or Eberhard, he exerted a great influence all through the XVIth century. We may leave him with the epitaph written by the witty physician, Gui Patin, more than a century later:

Grammaticam scivit, multos docuitque per annos, Declinare tamen non potuit tumulum.

In England the history of the grammatical versus memorialis is bound up with the name of William Lily, the first headmaster of Dean Colet's foundation, St. Paul's School. With Lily's Grammar as a whole, and with the contributions of Colet and Erasmus to some of its parts, as with its subsequent numerous revisions, as King Edward VI's Latin Grammar, and much later as the Eaton Latin Grammar, we are not here concerned. But the hexameters which prescribed the rules of gender, of preterites and supines, etc., filled a large and painful place in the school memories of many generations of Englishmen. Who has not heard of Propria quae maribus and As in praesenti, not without a mental picture of a cane rapidly vibrating between a mortar-board indecorously askew and a writhing piece of humanity in an Eton jacket? Lily first separates proper nouns from common: Propria, quae maribus tribuuntur, mascula dicas, Ut sunt Divorum, Mars, Bacchus, Apollo, virorum, Ut Cato, Virgilius, fluviorum ut Tibris, Orontes, Mensium, ut October, ventorum, ut Lips, Notus, Auster.

He then lumps nouns of all declensions together, making the genitive ending the criterion for gender, a singular system which makes it necessary to enter many nouns as exceptions to an exception, and hence to provide copious footnotes. The rules for preterites begin with:

As in praesenti perfectum format in avi Ut no nas navi, vocito vocitas vocitavi. Syntax is presented in plain prose, without the assistance of memory-verses. This part of the book appeared separately with Colet's preface in 1513, as De Octo Orationis Partium Constructione. It was sometimes reprinted in the same volume with Aldus's Institutiones, and wrongly attributed to Erasmus, in spite of that scholar's Epistula Candidis Lectoribus prefixed19.

The fashion of the hexameter memory-verse in the Latin Grammar died a very slow death. New ones were still produced in the XVIIth century, by no less

"Epist. Nuncupatoria (Cologne, 1533). 18 Jean Pellisson of Condrieu. 19So in a Columbia University copy of Aldus (Venice, 1526), in which Erasmus's name has been deleted.