a half-savage people out of about eight hundred radical word-stems with the aid of about two scores of affixes. The Maya language proper is not polysyllabic, consisting of words of one, two, or three syllables.

Perez worked on his dictionary from 1835 to 1859, compiling his great work partly from the living language of Yucatan and partly from the ancient or antiquated terms found in the manuscripts written by Maya scholars 350 years ago; and the Maya language is so conservative, as Dr. Brinton tells us, that the educated Maya of our days is fully able to comprehend the language of those old manuscripts, perhaps some obsolete terms excepted.

After Perez's untimely death, Dr. Behrendt revised and completed the dictionary, and it was published in Merida, Yucatan, in 1877.

The writer of the present paper has translated Perez's dictionary into English, and, while thus at work, he observed a great number of affinities to the Koptic language; so many, in fact, that far more than one-third of the Maya roots or radical word-stems come under the scope of this affinity.

The space in the PROCEEDINGS allotted to this abstract of the essay does not admit a recital of the equations; to recite few examples would be of little value, because in an investigation of this kind the convincing argument and proof rests in the accumulation of a large number of equations or examples, which are worth more than all speculation; but for those it will be necessary to refer the critical reader to the essay itself, which will probably appear in extenso in some European journal and perhaps also in an American periodical in the course of this year.

Suffice it to say that the Comparative Vocabulary brings 334 equations of radical stems, in which a Koptic word is covered by a Maya equivalent analogous in phonetic form and showing affinity in signification.

In gathering those equations, only such examples have been selected in which the phonetic differentiation is insignificantly slight and the affinity is apparent even to the less observant reader; otherwise the number of examples might have been largely increased.

The affinity does not end with the vocabulary. Not much 66 grammatical" analogy can be expected in two languages, when the branches of a people, who spoke a common original language, were forever separated before grammar, in the stricter sense of the word, was established; again, a large stock of radical terms or so-called roots of foreign origin may be imported into a language, enriching its vocabulary.

These roots, being assimilated, may thrive with a luxuriant growth in the adopting idiom; they will then be treated like the indigene roots according to the native grammar, so that finally they can hardly be recognized as being of foreign origin. Such, I think, was the process with the Koptic words in the Maya language, which doubtless already had its own grammar when the foreign element intruded.

Yet there are some grammatical analogies common to the two compared languages and peculiar to them, as compared with other tongues. We will enumerate a few of those:

1. The sounds are almost identical in the two languages, and where there is dialectic differentiation, the same is analogous in both; the Koptic r, which does not appear in the Maya proper, reappears in the Quichè dialect; so does the Bashmuric substitute an / for the Sahidic r. Koptic f is represented in Maya by

por b; so does the Sahidic and Bashmuric foster those latter sounds for the Memphitic f. d and g are missing in Koptic as in Maya.

2. Both languages love gemmination of the radical vowel, and, what is very characteristic, both use such gemmination or lengthening for the formation of the passive verb. [Compare Schwartze, Koptic Grammar, edited by Steinthal.]

3. In the formation of the plural, the Maya suffix ob, Quichè om or ob, corresponds to the Koptic plural suffix ov. Both languages know a plural by duplication of the stem.

4. In both languages, the simple stem is generally monosyllabic and tri-literal, sometimes bi-literal. Now, since a tri-literal root is never primitive, neither of the two languages can be considered as a primitive language.

5. In both languages the simple verbal stem is also used for a noun; the Maya does this without special prefix, while the Koptic distinguishes the noun by the affixed article; but the old Egyptian had no article, and did as the Maya does.

6. Both languages have the peculiarity of reduplicating the verbal stem; in the Koptic, generally, the full syllable with change in the radical vowel, the Maya retaining the latter in the reduplicated syllable but generally dropping the final consonant. The Maya forms frequentative verbs in this manner, the Koptic does the same, and it forms plurals by duplication. [See Schwartze, Gram., § 91, P. 372.]

7. The Koptic has a real, original indefinite demonstrative ah and another ash, which are used for forming verbal nouns. [See Schwartze, Steinthal, Gram., pp. 353, 362, 364.] The Maya forms certain "personal nouns from verbal roots, employing for that purpose the prefix ah- for males and ish- for females, f.i. cambal, to be instructed; ah-cambal, a disciple; ish-cambal, a female scholar.

8. The Koptic forms whole classes of verbal or abstract nouns by prefixes; the Maya has a very elaborate system of expressing a special manner of action such as: forced, sudden, slow, swift, etc., etc., by a simple prefix to the verbal

stem.

9. The Koptic forms ordinal numbers and it counts -times, -turns, -fold, -parts, days of month, hours of day, by special affixes. The Maya works a similar system for all it is worth: -times, -fold, -parts, -days, -tierces, -bundles; -flat or round or long or large things are discerned by special affixes to the numeral.

10. Both languages use ma, m, em, for strong negation, and both form comparative or superlative by particles of comparison only. These examples do not pretend to exhaust the affinities.

Now, since the Koptic is the youngest form of the Egyptian people's language, and since the oldest Koptic manuscripts accessible to us date from about 250 A.D. and are written in the Sahidic dialect, and since the Koptic words (found in the Maya language) belong to that dialect, we have a right to the following conclusion as the result of our investigation:

I. A large number of Egyptian words has been imported into the Maya language; those words were taken from the Egyptian language, as it was spoken by the common people in Upper Egypt, about the dawn of the Christian era; viz. in the Sahidic dialect, which at that time was, or shortly afterward became, the written language [Schriftsprache] of Egypt.

Now consider in connection with this the following well-known facts:

a. The Maya chronicles relate of an immigration of a number of people in

long robes, who, coming from sunrise, landed in Yucatan under a prince Votan, and, ascending the Usumacinta River, founded Na-chan [Palenque], and, taking native wives, naturalized and became the teachers of art and science; and the same chronicles fix the date of these occurrences at about the stated epoch, 250 A.D., according to calculations of P. Perez, Dr. I. Valentine, and others.

b. The peculiar arrangement of the Central American pyramids, similar to the terraced pyramid in Sokara in Upper Egypt; the special features of Maya sculpture reminding us of the Eastern style, although executed in a wild fantastic ornamentation.

c. Certain customs, ceremonies, and notions, such as embalming of the corpses, use of incense for worship, and others common to the people of Egypt and those of Central America, as enumerated in extenso in my paper,— then we cannot fail to arrive at the second conclusion: :

II. Those Egyptian words were brought to Yucatan by Egyptian (Koptic) emigrants, who formed a colony and communicated to the natives as much civilization as they themselves possessed. And since we are able to trace some of those Egyptian words into the Nahuatl of Mexico, into the Dakota (Sioux) and Algonquin, and again into the Quichna of Peru, our third conclusion will be:III. The influence of that immigration spread from the Maya to the wandering American nations; traces of such influence can still be traced. Consequently, the principal North American nations, as well as some of the more civilized South American, especially the Incas of Peru, must have had some connection and intercourse with the Maya nation at some time after the Egyptian immigration had occurred.

As proof is brought of a migration from the eastern to the western continent within well-defined historical times, the gap between the two continents is bridged, not by a fabulous Atlantis, but by the seafaring enterprise, audacity, and restlessness of man, the constant wanderer, carrying with him his virtues and vices, his myths and legends. The veil is lifted from the mysterious existence of Eastern lore on American soil; the wonders disappear before the light of knowledge, and for this we have to thank comparative philology.

22. Studies in Greek Agonistic Inscriptions, by Professor Edward Capps, of the University of Chicago.

This paper is printed in full in the TRANSACTIONS.

23. Etymologies of Some Latin Words of Will and Desire, by Professor Charles H. Shannon, of the University of Tennessee (read, in the author's absence, by Professor A. G. Laird).

The following etymologies are taken from a yet unpublished study of Words of Will and Desire in the Indo-European Languages.

Studeō, 'am zealous, eager.'

Phonetically, a connection of studeō with Gr. σñeúdw‚1 ‘am zealous,' cannot be defended; and its comparison with Gr. σreûμai,2 make as if I would,' is at least

1 Prellwitz, Etym. Wtbch. d. Griechischen Sprache, p. 297.

2 Persson, Wurzelerweiterung und Wurzelvariation, pp. 141, 144.

doubtful. I would suggest a comparison of studeō with Lat. tundō, 'strike, importune'; Skt. tudáti, 'he strikes.' In fact, Alb. štüń <*stūd-nio and Goth. stautan, 'strike,' which are recognized1 as belonging with Lat. tundō, Skt. tudáti, agree with studeō in showing initial s. Moreover, N.H.G. Stoss, blow, impulse,' in addition to the phonetic agreement, shows an approximate correspondence in sense. But, for the meaning, compare especially Eng. hammer at, labor at assiduously, earnestly.' From N.H.G. Stoss and Eng. hammer at it can be seen how could have arisen in studeō the idea of zeal,' which is the prevalent meaning throughout.

Amō, 'love, desire.'

A connection with Lat. emō, 'take, buy'; Goth. niman, 'take,' is, on the side of phonetics, far from plausible. The difficulty arises from the a in amō as against the e in emō; for the proportion maneō: μévw, while admitted, is not understood;2 and the assumed relation of amō to emō would not seem to be necessarily parallel, inasmuch as in this case the vowels in question are at the same time initial and before a nasal.

A second etymology proposed for amō is that, originating in a nursery word, a Lallwort, it is to be connected with Skt. ambā,3 'mother'; but, as regards meaning, there is no satisfactory parallel for the development from such a source of a word of passionate desire like amō.

I would explain amō as follows. In the Indo-European languages words of desire frequently develop from words that denote movement towards. Compare Lat. petō, make towards, rush at': 'beg, seek, desire'; Gr. ópoúw, 'rise and rush forward': 'am eager.' It is, further, perfectly natural that the idea of movement towards should show both a friendly and a hostile side; and this is abundantly evident in Latin itself. Compare, again, petō, 'rush at, desire': impetus,' attack, ardor'; petitio, 'attack, blow': 'request, beseeching.' Now amō may very well represent one side the friendly side- of such a double development from a word of movement towards; and the other side may be found in Skt, am, amīti, 'he presses on, harms'; dma-s, onset, impetuosity'; Avest. am, 'go.' To this explanation no objection can be made on account of the connection of Gr. 8uvvu, 'swear,' with amiti; for the ablaut &: a, which would have to be assumed in ŏuvvμ : Lat. amō, must be recognized in other words also. Compare oкpis, 'a jagged point,' with Lat. acus,5' needle.'

Latin amita, aunt,' possibly represents I.-E. *ame-tå, the fem. of *ame-tós, 'beloved,' a verbal adjective from the root of amō. For the meaning compare O. Irish fine, 'cognatus,' from the root uen seen in Lat. venus, 'love,' and Skt. van, vanéti, he loves, desires.'

Gr. ǎμOTOν, eagerly, insatiably,' for which no acceptable etymology has been offered, I would refer to the same root as amō, Skt. amīti, Gr. buvvμ. For the ablaut of ἄμοτον : ὄμνυμι compare ἄκρος, ‘at the point, with ὄκρις;7 and for the

1 Brugmann, Grundriss, I, pp. 113, 726.

2 Brug. Grundriss, I3, p. 120 f.; Lindsay, Latin Language, pp. 222, 274 f.

3 Uhlenbeck, C. C., Etym. Wtbch. d. altindischen Sprache, under amba. Zimmermann,

KZ. 34. P. 584.

♦ Brug. Grundriss, 12, p. 154; Aufrecht, Rh. M. 40, p. 160.

5 Brug. Grundriss, 12, p. 486.

7 Brug. Grundriss, 12, p. 486.

Brug. Grundriss, I2, p. 326.

recessive accent of ἄμοτον compare ἄμητος, a gathered crop'; βίοτος, ‘life, and the like. The development of meaning would be the same as in amō.

The root to which the foregoing words are to be referred probably had the following forms:

With these last thematic forms, especially Skt. dma-s, may be compared Lat. amāre, which implies directly an a-stem, *amā, and indirectly an o-stem, *amo-s (cf. Skt. dma-s). It is commonly the o-stem which stands in Latin beside the denominative verb in -äre. Compare regnu-m: regnāre; dominu-s : domināri. The root meaning, as above suggested, seems to have been movement towards, impulsion.8

ōrō, 'beg, beseech.'

As rhotacism does not take place in Oscan, the common view regards it as necessary either to separate ōrō from ōs, ōris, ‘mouth,' and compare it with Osc. urust, 'oraverit,' or to separate it from urust and compare it with ōs, ōris. A third possibility has been suggested, namely, that Osc. urust may have been borrowed from the Lat. ōrō, in which case, of course, it would be possible to look upon ōs, ōris, as the original of both. But there seems to be no sufficient reason to assume a borrowing on the part of Oscan.

I would follow those who see in urust the weakest form of the root uer, comparing Gr. elpw, say, speak, tell,' and the dh-extension of the same root in Lat. verbum and Goth vaúrd, 'word.' In this root the idea of 'speaking,' seen in Lat. verbum and Goth. vaúrd, is in all probability more nearly original than that of 'asking.' The assumption, moreover, that Osc. urust contains the weakest form of the root yer does not necessitate the separation of urust from ōrō; for the relation of these two words can be simply and naturally explained by seeing ur, the weakest form of uer in both. The phonetic Latin *uro, by a folk etymology, under the influence of an inevitable association with ōris and other cases of ōs, could readily have become ōrō.

24. The Formation of Substantives from Latin Geographical Adjectives by Ellipsis, by Professor John C. Rolfe, of the University of Michigan (read in abstract, in the author's absence, by Professor W. K. Clement).

This paper appears in full in the TRANSACTIONS.

1 Bartholomae, BB. 17, p. 111 f.

2 Brug. Grundriss, II1, p. 947.

This would account for the meaning of ourvue: cf. Eng. urge, drive on, press upon; asseverate: so in' He urged that this was true.' There seems to me to be less evidence that the root of oμvvμ, Skt. amiti, meant be hard, make hard,' as Aufrecht, Rh. M. 40, p. 160, takes it, comparing also ὠμός.

4 V. Planta, I, p. 520.

V. Planta, I, p. 520; Kluge, Etym. Wtbch, under Wort.