version of two orations of Lysias, of Aristotle's Rhetorica ad Alexandrum (really the Texn of Anaximenes). I lent you this codex in the time of duke Visconti. I made these versions when professor at Florence. (Dec. 15, 1453: first mention by F. of fall of Constantinople.) Nobody here in Milan has a copy of Plutarch's Parallel Lives: I have Latinized Numa and Lycurgus. 1454: Francesco Barbaro is dead in Venice: combination of Greek and Latin scholar

ship (as in him) is very rare. Get my Greek codices from his heirs.

Now (1454 sqq.) writes many letters to the rich and great in Italy and France, bespeaking money for travelling Greeks who are collecting funds for Turkish ransom. Are there any Greek codices for sale in Turin? F. wishes to employ a domestic copyist, an expert in abbreviations (notae). 1456: To Pope Calixtus III (see p. vii fin.). Pity that Nicholas V failed in his desire to have a version of Homer made. That pope, after fall of Constantinople, sent his envoys through all that part of Europe and Asia to buy up Greek codices. Argyropulos is among the Greeks now collecting funds for ransom. 6 Expects a librarius' from Mantua, to be employed in his house at Milan. 1458: Have not been able to find any book on Greek quantities. I need a Greek copyist. He thanks new Pope, Pius II (Humanist), for a codex of Plutarch.

1460: F. has had copied for himself Porphyrio on Horace. To his son Xenophon at Ragusa in Dalmatia: Are there any ancient monuments at Ragusa? any old inss. with name of town? Keep a lookout for Greek codices. To Alamanni at Florence: F. hears that many Greek codices have reached Florence from the shipwreck of Constantinople. Find out who has that codex of Silius Italicus, once acquired at Montepulciano from the estate of a man who was secretary to Pope Martin V († 1431): 'nam codices omnes, quotquot illo exemplari exscripti sunt, depravatos corruptosque invenio.' I desire a copy

made of the Latin translation of Aesop: nam auctor ipse periit apud Graecos.' 1461: To his faithful correspondent, Palla Strozzi, the Florentine exile (in Padua): I hear there are for sale there: Palaiphatos repl waλaiwv iotopiŵv; Cornutus Tepláλλŋyоpɩŵv; Syncellus: have copies made at my expense, or send the codices for copying. In a ferocious diatribe against his minor fellowhumanist Candido Dicembre F. cites, for quantity, Donatus, Servius, Priscian; for geography he quotes Strabo, Ptolemy, Pliny N.H., Polybius. We close this abstract with Filelfo's own words which he wrote for publication (as we now would say) or at least for the public: Unus Filelfus audet affirmare, (vel insaniente Candido) neminem esse hac tempestate nec fuisse umquam apud Latinos, quantum constet ex omni hominum memoria, qui praeter se unum idem unus tenuerit exercueritque pariter et Graecam pariter et Latinam orationem in omni dicendi genere, et prosa et versu.'

And in spite of this fanfaronade we may well accept the judgment of his biographer Rosmini: ‘la sua vasta erudizione per que' tempi maravigliosa.

3. The Titles of Caesar's Works, by Professor Francis W. Kelsey, of the University of Michigan. This paper appears in the TRANS

ACTIONS.

Remarks were made by Professor Sihler.

4. Futures in -bo in modern Hindu Dialects, by Professor E. Washburn Hopkins, of Yale University (read by Dr. Magoun).

It will doubtless surprise the classical scholar to learn that modern Hindu (Aryan) dialects possess futures in -bo. They have in fact futures in -bo and futures in -am, so that, as striking parallels to ibo or dabo and dicam, they show yabo, 'I will go,' from yā, 'go,' and kaham, ‘I will speak,' from kah, 'speak.' I hasten to add that between the Hindu and Latin terminations there is no genetic connection.

But these forms of the Hindu verb are worth noticing. They show, when their history is traced, first, that pronominal endings occur as verbal suffixes in purely Aryan verbs, and second that precise tense-meanings may develop out of a merely adjectival verbal form. The history of terminations in Greek and Latin is doubtful; the greater value attaches to a growth which can be followed back to its beginning.

For the modern material here referred to and for the earlier middle Indic -bo forms I am indebted to the great thesaurus of Dr. Grierson, whose Linguistic Survey of India is a mine of wealth to the student of modern Hindu dialects.

The forms in -bo are, in a word, only the latest development of adjectival forms parallel to Greek adjectives in -Téov plus a pronoun-ending. As λuTéov compared with Auróv has passed from a general meaning to one prevailingly gerundive, so the Hindu adjective in -tavyà has passed from a general adjectival (infinitive) meaning to one prevailingly gerundive. In Greek itself the -Tóv form has also the sense of the -Téov form: ȧκovσтóv is audible'; πрактóν, what 'may be done'; TIμNTéos is a man to be honored,' venerandus. In Sanskrit, the -tavyà form gradually receives an almost stereotyped gerundive meaning: han-tavyà is not merely about to be killed' but 'killable.' But in the earlier Vedic use, as found in (jātám) janitavyàm, the sense is merely 'to be born.' At this stage no subject is needed, but when found it is in the instrumental case. But the potential note becomes pronounced very early, and this, again, is paralleled by the closely related adjectival forms in -tva (tua). Thus kṛtāni are things done,' as opposed to kártvāni, things 'to be done,' and in ydj jätáṁ yác ca jántvam, 'going to be born,' is simply future by implication. The notion of possibility, the potential idea, comes next, as in the Greek σrvyvós, 'hateful,' as well as 'hated,' or Latin invictus, unconquerable' as well as unconquered.' So (RV.) jayātu jētvāni, 'conquer what can be conquered,' (RV.) nántvāni, 'conquerable,' (RV.) snåtvam (udakam), 'bathable' (water). Another Greek parallel may be found in such forms as ayıos, Sk. yajyd, the latter being a noun as well as an adjective. So yujya is not only iunctus, but a friend,' just as máytos is 'fixed,' 'firm.' Latest of all is the moral gerundive sense, nd brāhmaṇd hinsitavyàḥ, AV. v, 18. 6, ‘inviolate is the priest.'

The connection with the infinitive, to which the -tavyà form is an adjective, makes certain the indefiniteness in meaning of the Sanskrit form. Thus kartavyà is to be done' or 'to do' (cf. the inf. kártave), and the infinitive itself is used in the same way: pra'ndhám . . . cákṣase kṛthaḥ, 'ye have made the blind to see'; yád im ušmasi kártave, 'what we wish to be done.'

...

Now when the middle Indic dialects made their future they operated in part with this verbal adjective like -tavya. First they took car or cal, 'go,' for exam

ple, and using the verbal adjective to express futurity, made, with regular phonetic change, (calitavyam) calidavvam and caliavvam. This in turn became caliba, or, in Bengali, calib, not by any means at first a personal future, but impersonal, for any person and number. For this reason, when a Bengali had to say 'I shall go,' he added to calib (eundum) the word a, which is pronounced, and often phonetically spelled, ō, that is, the word 'me' in the agent-case; so that calibo or yabo is Latin, eundum-mihi, ibo, maribo is 'I shall strike' ('to strike by me'). Here then we have a verbal ending which in reality is nothing but a personal pronoun.

In exactly the same way, māris is 'he struck' and the s, although to all appearance a verbal termination, is in reality the final reduction of the third personal pronoun, meaning 'by him,' while the first part of māris is a phonetic reduction of the past passive participle, mărita, mārida, māria, mārya, māri. In both of these cases the impersonal form was preferred, as it is in Sanskrit. So māry-am is struck by me' in the Sauraseni form, am being 'by me' (in Hindī, mārilam māridam has the same origin). The last stage is exemplified in Rājasthānī, where the agent-case has been supplanted by the nominative. Thus wah uthio, 'he rose' (instead of usă, ‘risen by him'). This is a recent development, showing that all sense of the impersonal origin has been lost and the verbal has at last attained to the state of a completely inflected verbal form, the nominative pronoun replacing the agent-case. So the Hindi has hữ mãio, ‘I killed,' instead of 'killed by me.'

To the adaptationist, who repudiates such synthesis as un-Aryan, such a development within comparatively recent time should furnish food for reflection. On the other hand, the genesis of a tense of precise future meaning out of the indefinite (infinitive) meaning supports the view that the direction of development of meaning is toward precision. The vague and general becomes exact and specific. But further. Future forms may serve as subjunctives. In one group of dialects the same form in one dialect is future, in another subjunctive. But this futuresubjunctive is really an indicative filled out with the copula 'be.' Thus in Rājasthānī the subjunctive differs in the first person from the indicative only in adding to the latter (I) am.' 'I go' is indicative, 'I am go' (going) is subjunctive, as in karu and karữ hãi (hãi, ‘I am'). Both future and subjunctive are expressed by hu mārūgō, ‘I shall (or may) beat,' hй uthugō, 'I will arise'; as in the preterite jito ha gio (gio, 'went') is literally 'I (be) came alive.' But in one dialect of this group, hu mārugo serves only as future, 'I shall beat' and hu mārū hai, 'I am beating,' serves as subjunctive. The usual future is made with gō, hữ us-tah kahugō, 'I will speak to him.' 1

In all this we are reminded of the Tibeto-Burman verb (or lack of it), where there is only a verbal noun and the future is made by adding a post-positive for,' in the sense of 'in order to.' In Burmese we find 'ga,' 'with,' added to a stem to serve as a future-sign, just as with' (ge-) makes a German perfect. In the Lushai dialect, 'do,' thwa, is added to a stem to make an imperative, suggesting that in to, Sk. idhi, we have really do come,' √dha. In the Hindu-Kush the

1 The Sanskrit scholar will observe how the ablative' sense has yielded to that of the locative in us-tah, him-to.' But really location in general is expressed by tas even in Sanskrit. Thus itaḥ is here,' even to here' as well as 'hence,' as in Śak. ito dattadṛṣṭiḥ is 'with look directed hither,' on this side.'

deliberative subjunctive is made by adding to the indicative the interrogative a, which, when added to a stem in a, makes ā, just as hanti, indicative, becomes hánati, subjunctive, and bhárati becomes bhárāti. As another example of an ending which is a word, some dialects make the subjunctive by adding kyah, 'perhaps,' to the indicative; thus kudddu is 'I strike,' and kuďádukyah is ‘I may strike.' The passive in Yidghah is made by adding kshiyah to the verbal stem, and kshiyah is the word for 'go.' The present as a future may be illustrated by the Bhojpuri of Palaman, which regularly uses present as future, yai, 'I will go,' kahi, 'I will say,' instead of yaib (yābo) and kahab (kaham). This indicative is the Nāipālī future. By adding lā, ‘gone,' to the present indicative (subjunctive) a future is produced like that of Hindusthānī (with gā, dekhūgā, 'I am going [that] I see,' that is, 'I shall see'); thus dekhùlā is ‘I shall see' in Nāipālī, but the Bhojpuri uses the same form, dekhīlā, as a present. Eastern Maithili has here dekhibo (o), like Bengali -ibā (ō). And what is the 'inserted ¿' in dekh-i-bo? It appears also in one of the Eastern Hindi dialects (called by Dr. Grierson Surgujia), where there is "a tendency to pronounce [i.e. insert] a final or unaccented short i in the preceding syllable," best illustrated by ka-i-r for kar, ma-i-nase for manise. A 'tendency' of this sort may be enough to explain the same phenomenon at an earlier date.

Hindi is analytic, Bengali is synthetic. Thus ghara-kā or ghara-mă of Hindi becomes gharak, etc. So in modern Aryan we have just the conditions which would have produced 'endings.' Hindi gharak is a combination of two words (the kā is reduced from kṛta in oblique form as 'for'), the latter of which has become a mere ending, but was once a separate special word with a definite meaning. Why should we doubt that in the same way it was of old quite Aryan (as it is now) to possess analytical forms reducible to synthetic combinations? Further, as regards the subjunctive idea, it is plain that there is no a priori necessity for deriving it from a volitive through a deliberative into a prospective notion, as is now generally thought to have been its course.1

5. The Relation of Accent to Elision in Latin Verse, by Professor Albert Granger Harkness, of Brown University. This paper will be found in the TRANSACTIONS.

In the discussion Professors Bennett, Radford, H. F. Burton, Fitch, Knapp, Dr. Magoun, and the author participated.

6. Some Linguistic Principles and Questions involved in the Simplification of the Nomenclature of the Brain, by Professor Burt G. Wilder, of Cornell University; read by invitation.

1 On the connection of the -tavyd forms with the infinitive, see Brugmann, KVG. § 809; Gr. Gr., § 583. The gerundive meaning, even after it is fully established in Sanskrit, occasionally lapses back into the infinitive-potential sense. Thus in Mbh. vii. 54. 37, yady evam etat kartavyam mayā na syād vină prabho, means only "if this cannot be done without me, O Lord," (not to do'). The Sanskrit future stem is also employed to make verbal adjectives. The oldest case is yani kariṣyá kṛṇuhi, ' do what (things) are to be done' (kartavyāni) RV. i. 165. 9, according to Sāyaṇa; but this may be a false reading, as is now generally assumed. Later we find janisya as in Rām. vii. 24. 5. 58, na jāto na janisyo vā, as in the older phrase jāto janitavyo vā.

The object of Simplification of the Nomenclature of the Brain is to render the knowledge of the structure and functions of that most complex organ more easy to advance, to record, to teach, and to disseminate among the laity.

The leading ideas of this paper were stated in an address, “Paronymy versus Heteronymy as Neuronymic Principles," read by me, as President, before the American Neurological Association twenty years ago. It was published in the Transactions of the Association, Journal of Nervous and Mental Disease, vol. XII, No. 3, July, 1885, and reprints were somewhat generally distributed among contemporary neurologists and anatomists. In some respects it would have been read more appropriately before this Association. However, several members of this Association have already directly assisted me. An early colleague, Isaac Flagg, suggested paronymy in 1885. A later colleague, now President Benjamin Ide Wheeler, for nearly twenty years, often at the cost of interrupting his own studies, has promptly and helpfully responded to my etymologic queries. It is therefore fitting that from Dr. Wheeler should have come the proposal that I be invited to read this paper.

Simplification of anatomic language in general was advocated by me in 1871, and in 1880 attention was called to the special need with respect to the brain. The magnitude and difficulties of the task were not fully realized until 1884, in dealing with the neurologic portion of Foster's Encyclopaedic Medical Diction ary. From works and periodicals in all languages there was compiled an alphabetic list of about 10,500 names for the (at the most) 500 known parts and features of the brain.1

The need of some sort of classification of this host of terms was literally forced upon me, and there was gradually evolved a dichotomous arrangement substantially identic with the subjoined Table.

Recognition of the labors of predecessors, and acknowledgment of the coöperation of contemporaries, here and abroad, are recorded in the following papers: "Paronymy versus Heteronymy" (mentioned above); "Neural Terms, International and National," Journal of Comparative Neurology, vol. VI, pp. 216352, Dec., 1896; "Some Misapprehensions as to the Simplified Nomenclature of Anatomy"; address, as President, before the Association of American Anatomists, Dec. 28, 1898; Proceedings of the Association, eleventh session, pp. 1539; also Science, n.s., vol. IX, April 21, 1899, pp. 563-581. The paper last named discusses the objections and adverse criticisms that have been offered. Of these the most accessible in this country is a "Review" in Science, n.s., vol. VII, May 20, 1898, pp. 715-16; the printer's blunder ("chippocamp" for hippocamp), involving injustice to both parties concerned, was corrected by the reviewer at the end of Science for June 3.

1 Of these more than 3000 (an average of at least six for each part) were Latin and thus ostensibly international. Many, however, were more or less completely restricted to certain countries, institutions, or writers. The Report (embodied in the B. N. A.) of the "Nomenclatur-Commission," adopted in 1895 by the Anatomische Gesellschaft, while defective in many respects and practically ignoring the previous labors of English-speaking anatomists as individuals and as committees, "buried” a large number of “dead or dying" terms; those who aim at still further improvement of neuronymy may now confine their attacks to a smaller number of names, exchanging, so to speak, the shot-gun for the rifle.