¹ Among the first who recognized that bellido and vel(l)ido are merely variant spellings was E. Gorra, Lingua e letteratura spagnuola delle origini 376, 424 (Milan, 1898).

² F. Diez, Etymologisches Wörterbuch der romanischen Sprachen 564-5 (Bonn, 1853).

³ P. Rajna, Una versione rimata dei Sette Savi, Romania 7.49.

⁴ G. Körting, Lateinisch-Romanisches Wörterbuch §1111 (Paderborn, 1891); 2nd and 3rd edd. §1310 (Paderborn, 1901, 1907).

⁵ R. Lanchetas, Gramática y vocabulario de las obras de Gonzalo de Berceo 178 (Madrid, 1900). The formation is labeled as diminutive.

⁶ R. Menéndez Pidal, Cantar de Mio Cid: texto, gramática y vocabulario 891 (Madrid, 1908-11).

⁷ C. Michaëlis de Vasconcelos, Glossário do Cancioneiro da Ajuda, RL 23.92.

⁸ J. J. Nunes, Crestomatía Arcaica² 602 (Lisbon, 1921).

⁹ J. Huber, Altportugiesisches Elementarbuch 32-3 (Heidelberg, 1933). The author considers the possibility of a Castilian loan to Old Portuguese.

¹⁰ C. de Figueiredo, Novo Dicionário da Língua Portuguesa⁴ 2.913 (Lisbon, ca. 1925).

¹¹ C. C. Marden, El Libro de Apolonio, Elliott Monographs 11-12.179; J. Keller, Contribución al vocabulario del Poema de Alixandre 39 (Madrid, 1932).

¹² G. Cloëtta, Crenu, Romania 14.571-2.

¹³ J. Cejador y Frauca, La lengua de Cervantes 2.1131 (Madrid, 1905-6).

¹⁴ Thus, in the 15th edition (1925) vellido is followed by the gloss ‘velloso’.

¹⁵ L. Spitzer, Notes étymologiques, RFE 12.231-2. Notice the critical comment on the Old Italian parallels adduced by Menéndez Pidal.

¹⁶ P. F. Monlau, Diccionario etimológico de la lengua castellana 418, 1126, s.w. bello, vello (Buenos Aires, 1941; reprint of 1881 ed.).

¹⁷ P. Förster, Spanische Sprachlehre 217 (Berlin, 1880).

¹⁸ J. Alemany Bolufer, Tratado de la formación de palabras en la lengua castellana 77 (Madrid, 1920; reprint from BRAE).

¹⁹ The Diccionario de Autoridades 6.436 bases its interpretation on a not quite transparent passage from Don Agustín Moreto's play El poder de la amistad (and is, as usual, supported by A. de Pagés, Gran diccionario 5.682). Yet even if that explanation were correct, it would apply to the word in its dying stage, when its meaning, no longer properly understood, was exposed to distortion; it would have no bearing on OSp. vel(l)ido, which alone interests the etymologist.

²⁰ On the survival of -ītus, see, in addition to Cloëtta's note, the excellent chapter in G. Cohn, Die Suffixwandlungen im Vulgärlatein und im vorliterarischen Französisch 180-205 (Halle a.S., 1891).

²¹ W. Meyer-Lübke, Grammatik der romanischen Sprachen 2.517 (Leipzig, 1890-1904). Vellido is here likened to dolorido.

²² J. Leite de Vasconcelos, Antroponímia Portuguesa 53 (Lisbon, 1928).

²³ F. Hanssen, Gramática histórica de la lengua castellana 132 (Halle a.S., 1913).

²⁴ He qualified the derivation of vellido from bello as dubious in REW¹ (the fascicle appeared ca. 1915) and completely omitted it in REW³ (ca. 1931), forgetting to delete the reference in the index. A student of Meyer-Lübke, E. Gamillscheg, also expressed his doubts in a lecture course in the University of Berlin, ca. 1936.

²⁵ P. Meyer, Fragment d'une Vie de Saint Eustache, Romania 36.15.

²⁶ W. von Wartburg, Französisches etymologisches Wörterbuch 1.320a. For further documentation, see Godefroy, Dictionnaire de l'ancienne langue française 1.617b; Tobler-Lommatzsch, Altfranzösisches Wörterbuch 1.911; K. Nyrop, Grammaire historique de la langue française² 2.328, 458 (with bibliography). On Old Provençal bellaire, bellazor, see C. G. Grandgent, An Outline of the Phonology and Morphology of Old Provençal 97 (Boston, 1905); O. Schultz-Gora, Altprovenzalisches Elementarbuch⁴ 72 (Heidelberg, 1924).

²⁷ Notice the contrast with sella > si(e)lla ‘saddle’ and with the suffix -ellu > -i(e)llo, as discussed by Menéndez Pidal, Orígenes del español² 168-76, 579 (Madrid, 1929).

²⁸ See REW³ §1027; FEW 1.321. Further study must establish whether bel(lo) is to be classed as a retarded formation (semilearned), as a borrowing from Provençal, or as an autochthonous word subject to apocope and failure to diphthongize, much like dominu > don (in contrast to dueño), because of its pretonic position. To explore this latter possibility, see the data assembled by Hanssen, Gramática histórica 32, and M. A. Zeitlin, La apócope de la -a final átona en español, HR 7.242-5.

²⁹ Were it not for the authority of scholars like Meyer-Lübke and Wartburg, it would be idle to document bello and its cognates from Old Spanish sources. For bel, see Vida de Santo Domingo, quatrain 335; Crescentia, chapter 4; Carlos Maynes, chapters 29, 37; Enperador Ottas, chapters 23, 30, 48. For bello, -a, see Vida de Santo Domingo, quatrain 234; Fernán González, quatrain 403; Historia troyana en prosa y verso 196; Carlos Maynes, chapter 3; Enperador Ottas, chapter 41; Juan Ruiz, quatrains 129, 522, 1025; Confisión del amante, fol. 216ro. Beltad, beldad: Loores, quatrain 205; Santa Catarina, fols. 15vo, 18vo, 19vo, 22vo; Rey Guillelme, fol. 37ro; Crescentia, chapters 1, 2, 7, 11, 12, 17, 20; Carlos Maynes, chapter 1 ; Enperador Ottas, chapters 13, 17, 31; Juan Ruiz, quatrains 596, 643, 725, 727, 1322, 1613. On bellamientre see V. Fernández Llera, Gramática y vocabulario del Fuero Juzgo 120-1 (Madrid, 1929). For further sources, see the Dicc. Hist. Many, though not all, of these texts go back (directly or indirectly) to Old French originals; intermediary versions were in Old Galician or in Old Leonese, as has been demonstrated by Mussafia and Pietsch. However, from the preference given by translators to belilo) over fermoso, fremoso in translating OFr. biaus it does not necessarily follow that bel(lo) was an imported word. On the reduction of bellus in Gascony and Aragon to the status of an indefinite pronoun, see G. Rohlfs, Le gascon: études de philologie pyrénéenne 130 (Halle a.S., 1935).

³⁰ V. García de Diego, Elementos de gramática histórica gallega 190-1 (Burgos, [1909]), quotes nogueirido ‘nut-tree grove’, teijido ‘yew-tree grove’, and interesting toponymic data like Lourido, Acibido, Carballido, Cerquido, Abelaindo < Abelãido. On the various products of -ētum, -ēta, see also J. M. Piel, A Formaçāo dos Nomes de Lugares e de Instrumentos em Português, BF 7.31-47.

³¹ On Old Navarran betereiscitu from veterēscere ‘to grow old’ (found in Glosas Silenses, no. 7), see Menéndez Pidal, Orígenes del español² 377. On the participles in -ido, see, in addition to the standard manuals, A. Gassner, Das altspanische Verbum 190-5 (Halle a.S., 1897). On the survival of -ido (instead of -ecido) in verbs which, in their finite forms, had shifted or were shifting by 1250 from the -ir to the -ecer class, see Studies in Philology 38.433. The occurrence of forms like amortido, atordido, embegido, empobrido, enflaquido, enloquido, establido, estordido, finido in Berceo and earlier writers can easily be established with the help of V. R. B. Oelschläger, Medieval Spanish Word List (Madison, [1940]).

³² Pertinent examples are (a)modorrido, (de)sabrido, desequido, (des)famnido, (des)-marrido, dolorido, ensangostido, entelerido, escaldrido, (es)pavorido, malbaylido, regordido, resequido, revejido, analyzed in excursus A. Isolated formations are fardido ‘bold’, borrowed from Old French hardi, and not yet fully transparent sencido ‘intact’ in Berceo (see García de Diego, RFE 7.117, whose explanation satisfied A. G. Solalinde, as appears from his note to Milagros, quatrain 2; also Spitzer, RFE 13.113-28).

³³ Amply documented by V. García de Diego, Elementos de gramática histórica castellana 194 (Burgos, 1914); J. Alemany Bolufer, Tratado de la formación de palabras 76; E. Benot, Diccionario de asonantes y consonantes 827-31 (Madrid, [1893]) ; C. von Reinhardstoettner, Grammatik der portugiesischen Sprache 137 (Strassburg, 1878); J. H. D. Allen Jr., Portuguese Word Formation with Suffixes 52 (Language Dissertation 33; Baltimore, 1941). Oelschläger, Medieval Spanish Word List 17a, supplies interesting excerpts from documents issued in Caparroso, Guadalajara, Peñafiel, Sahagún, Salvador, Sepúlveda, and Tudela from the 11th to the early 13th century, to illustrate the early use of apellido. Alarido is equally old, occurring in the Cantar de Mio Cid. Perplexing is the use of bocedo ‘crying’ in Duelo, quatrain 195 (confusion between -ītu and -ētu?). See Excursus A.

³⁴ On Gallo-Romance abelir, rabelir, embelir, rembelir, parembelir, desembelir, debelir, and, with sporadic and distinctly secondary apheresis, belir, see FEW 1.320. For Italian and Catalan, see REW³ §1027 and the standard dictionaries. On apheresis, see E. Muret, De l'aphérèse, in Mélanges A. Thomas 309-17 (Paris, 1927).

³⁵ See Gassner, Das altspanische Verbum 63. I have been unable to identify embellecer in the oldest texts.

³⁶ Afeitar (not of indigenous stock): Duelo, quatrain 50; Milagros, quatrain 515. Ptg. enfeitar has been arrived at through prefix change. Afermosar: Las Siete Partidas, López de Ayala, Crónica de D. Pero Niño (Dicc. Hist. 1.239a) ; and Primera Crónica General 378b. Afermosear: Juan de Mena, Gordonio (Dicc. Hist.). Afermosiguar: Enrique de Villena (ibid.). Fermosear seems to belong to a later stratum; hermoseamiento is documented from Juan de Pineda by F. Rodríguez Marín, Dos mil quinientas voces castizas 202 (Madrid, 1922).

³⁷ Compare Faya Bellida ‘Beautiful Beech’ with the name of a village in the province of Salamanca : Robledohermoso ‘Beautiful Oak Grove’ Cf. J. J. Nunes, A Vegetação na Toponimia Portuguesa, Boletim da Academia d. Sciências 13.131-75.

³⁸ Thus spelled in Primera Crónica General de España 509b-512a, ed. R. Menéndez Pidal (Madrid, 1906). Adolffo is, of course, Germanic; see J. J. Nunes, O Elemento Germânico no Onomástico Português, in Homenaje ofrecido a Menéndez Pidal 2.587 (Madrid, 1925). Free combination and even contamination of Latin and Germanic names was not exceptional, as has been pointed out by W. Meyer-Lübke, Einführung in das Studium der romanischen Sprachwissenschaft³ 248 (Heidelberg, 1920).

³⁹ Under this name, and on the understanding that he was the son of Dolfos Bellido, the putative murderer of King Sancho the Second appears in the romancero. See Cancionero de romances (n.d.), fol. 158; Cancionero de romances. (1550), fols. 148, 150; Silva (1550), vol. I, fol. 80; Primavera y Flor (1856), nos. 45, 47. For brief comment, see S. G. Morley, Spanish Ballads 122-3.

⁴⁰ See the dictionary of M. Aguiló i Fuster 8.194. Also, there exists in Catalan the verb abellitarse ‘enginyarse per fer les coses de manera que donin bon resultat’; see A. M. Alcover and F. de B. Moll, Diccionari català-valencià-balear 1.37a (Palma de Mallorca, 1930-), corresponding to Sp. habilitarse. To explain it, possible influence of abellir or abillar has been suggested ; the existence of abellidar ‘to embellish’ in Old Spanish seems to recommend a different interpretation.

⁴¹ Leite de Vasconcelos, Antroponímia Portuguesa 53, 352, with reference, for a rich collection of variant forms, to A. A. Cortesão, Onomástico Medieval Português, in O Arqueólogo Português (Lisbon, 1912 ff.), a study not accessible to me.

⁴² Documentation is supplied by Menéndez Pidal and Leite de Vasconcelos. Add A.C. Jennings, A Linguistic Study of the Cartulario de San Vicente de Oviedo 207 (New York, 1940), and Oelschläger, Med. Span. Word-List 135b, s.v. monneca.

⁴³ Compare the cases of Lerdo and Lozano. As will be shown in forthcoming essays, lerdo ‘sluggish’ derives from en-lerdar ‘to be lazy’, based, in turn, on glīs, glīris (*glēre, through Italic influence) ‘dormouse’. Lozano (OSp. loçano, Ptg. loução) signifies fundamentally ‘proud, overbearing’, involving a Cothic stem and a Latin formative.

⁴⁴ So stated by Menéndez Pidal. The word is not listed in the glossary to the Academy edition of the Cantigas (1889).

⁴⁵ Line 6191 of the edition by C. Michaëlis de Vasconcelos, according to her glossary, RL 23.92. Velido is here understood to mean ‘enfeitado, lindo, bonito.’ Similarly, H. R. Lang, Das Liederbuch des Königs Denis von Portugal 170 (Halle a.S., 1894) renders velido by ‘anmutig, hübsch, schön'. C. Michaëlis, in quoting parallels, supplies references to the numbers of poems as they appear in the edition of the Cancioneiro da Vaticana by E. Monaci (Halle a.S., 1875).

⁴⁶ Using the data of Monaci, C. Michaëlis, and Nunes, I have prepared the following list of occurrences of velido in the Cancioneiro da Vaticana: Nos. 170, 172, 195: El-Rei Don Denis (corresponding to lines 1839, 1881, 2345 of the edition by Lang); Nos. 259, 262, 263: Airas Corp'ancho; No. 292: D. Joan Soares Coelho; No. 342: D. Afonso Lopez de Bayan; No. 401: Pae Gomez Charinho; No. 415: Pedro Eanes Solaz; No. 435: Garcia Soarez; No. 462: Airas Nunes; No. 488: Roy Fernandiz; No. 712: Pero d'Ardia; Nos. 739, 741, 742: Joan Servando; Nos. 754,761 : Joan Zorro; Nos. 792, 793,797: Pero Meogo; No. 855: Lopo (jograr); No. 889: Martin Codax; No. 898: Joan de Requeixo; No. 1173: Pero de Ponte. Some poets seem to have consistently avoided velido; cf. A. Gassner, Zwanzig Lieder des Joan Ayras de Santiago, Miscelânea de Estudos em Honra de D. Carolina Michaëlis de Vasconcelos 385-418 (Coimbra, 1933), where dona fremosa, senhor fremosa are used throughout (C. V., Nos. 534, 540, 542, 543, 548). In the following notes, the numbers refer, for convenience sake, to the poems in Nunes, Cantigas d'amigo dos trovadores galego-portugueses (Coimbra, 1926-8).

⁴⁷ No. 43, line 1; No. 92, line 1; No. 95, line 7; No. 96, line 10; No. 249, line 6; No. 369, line 2; No. 371, line 1; No. 462, line 15.

⁴⁸ No. 20, line 1; No. 123, line 3; No. 175, line 4; No. 236, line 1; No. 258, line 3; No. 347, line 9; No. 372, line 1 ; No. 390, line 1 ; No. 414, line 3; No. 415, line 2; No. 419, line 1 ; No. 503, line 14.

⁴⁹ On the etymology of this word, see note 43.

⁵⁰ No. 383, lines 3, 6, 9,12; cf. No. 369, line 7: madre velida e senhor.

⁵¹ No. 18, line 1 ; No. 220, line 26; No. 496, line 2. Compare the personification underlying OPtg. o corpo velido ‘the beautiful maiden’ with OSp. la barva vellida ‘the gallant hero’.

⁵² R. Academia Gallega, Diccionario gallego-castellano 312a (La Coruña, 1913-26) : corpo belido ‘cuerpo bonito, elegante, delgado, flexible, gracioso, airoso’, marked as obsolete; J. Cuveiro Piñol, Diccionario gallego 37 (Barcelona, 1876) : belido ‘hermoso, bello’, also marked as antiquated. The word is not listed in the glossary appended to the Old Galician Cronica Troyana.

⁵³ The use of belido in the journal A Nosa Terra, No. 225, is recorded by Nunes, Cantigas d'amigo 3.700. The sporadic appearance of velido in the writings of Lopes Vieira is mentioned by Leite de Vasconcelos, Antroponímia 53.

⁵⁴ At least, I suspect that L. Carré Alvarellos, Diccionario galego-castelán 1.250 (La Coruña, 1928), has manufactured that derivative.

⁵⁵ Whenever possible, his references have been checked against the latest editions available.

⁵⁶ Menéndez Pidal was doubtless justified in interpreting this spelling as intended to represent vellido. Nevertheless, if the etymology here advocated is acceptable it is theoretically conceivable (though, to be sure, unlikely) that a variant from velido, with [1] instead of [], should have existed, cf. melindre < mellīgine ‘honey-like substance’. There has been considerable vacillation in Imperial Latin between indigenous mel, mellis ‘honey’ and imported μέλι, μέλιτος. The result was the coinage of numerous hybrids and the absorption of several Hellenisms. See ThLL 8.619 for illustration of meliphyllon, melisson, melissophyllon, melisūrgica, melitenus, melitērus, melitītēs, melittaena, melizōmus; J. Sofer, Lateinisches und Romanisches aus den Etymologiae des Isidorus von Sevilla 100-1 (Göttingen, 1930), s.v. melimēlum - mālomellus; J. Svennung, Untersuchungen zu Palladius und zur lateinischen Fach- und Volkssprache 153-4 (Upsala, 1935). In the case of mellīgō, recorded vacillation between cālīgō and callīgō ‘mist, vapor’, fūlīgō and fūllīgō ‘soot’, lolīgō and lollīgō ‘cuttlefish’ may have had its share in the simplification of the geminate consonant; see W. A. Baehrens, Sprachlicher Kommentar zur vulgärlateinischen Appendix Probi 78 (Halle a.S., 1922). Notice that line 274 of the Cantar (enclinó las manos la barva vellida) contains a felicitous emendation of the editor.

⁵⁷ Libro de Alexandre, ed. Willis, O, quatrain 1415c (vellida); P, quatrain 1557c (bellida). F. J. Wolf and C. Hofmann, Primavera y flor de romances 84 (Berlin, 1856) ; also reproduced by A. de Pagés.

⁵⁸ Libro de Alexandre, O, quatrain 2087b; P, quatrain 2229b. Libro de Apolonio, ed. Marden, quatrain 315b. Maestro Gonzalo Correas, Vocabulario de refranes 397a (Madrid, 1906) : Por sus ojos bellidos, dícese con desdén cuando se niega y no hay obligación de hacer algo por uno; bellidos es ‘hermosos’, es lo mismo que decir: ‘por su cuerpo gentil’.

Another refrán cited by Correas reads: Dámele vestido, dartele he bellido.

⁵⁹ See the comment of Spitzer on the quotation of the saying by J. Lamano y Beneite, El dialecto vulgar salmantino 282 (Salamanca, 1915).

⁶⁰ Gonzalo de Berceo, Santa Oria, ed. F. Janer, quatrain 29d: vírgines . . . bellidas; Libro de Alexandre, O, quatrain 365d - P, quatrain 373d; Libro de Apolonio, quatrain 370c: moça . . . vellida.

⁶¹ See the strongly Castilianized and late version B, ed. H. Morf and M. Schmitz, RF 11.315-411, 623-7, quatrains 224c (tomólo de la mano y onrrólo el bellido), 225c (mi ermano el bellido). The earliest editors misread the word as valido. The rendering of v- by b- is normal in aljamiado texts; see R. Menéndez Pidal, Poema de Yúçuf : materiales para su estudio, RABM 6.114 (tercera época); A. R. Nykl, A Compendium of Aljamiado Literature 189-90 (Paris - New York, 1929; reprint from RH 77). Bellido does not occur in the Judaeo-Spanish Coplas de Yóçef, nor is it in use among Moroccan Jews (BRAE 13.522-5, 14.577).

⁶² F. J. Simonet, Glosario de voces ibéricas y latinas usadas entre los mozárabes 22 (Madrid, 1888). Possibly more information is given in M. Asín Palacios, Glosario de voces romances registradas por un botánico anónimo hispano-musulmán (Madrid-Granada, 1943), not yet accessible to me.

⁶³ Quoted by Aguado, Glosario sobre Juan Ruiz 263, from Un nuevo poema por la cuaderna vía: el Libro de la miseria, ed. M. Artigas (Santander, 1920; reprint from BBMP 2), quatrain 35d.

⁶⁴ The passage from La Coronación is quoted by the Dicc. d. Aut.

⁶⁵ I owe this reference again to Aguado. B. Díaz del Castillo, a native of Medina del Campo, lived from 1492 to 1580. The word occurs in Verdadera historia de los sucesos de la conquista de Nueva España, chapter 20: Comenzó de pulir y abellidar en su persona mucho.

⁶⁶ See Juan de Valdés, Diálogo de la lengua, ed. J. F. Montesinos, p. 11 (Madrid, 1928).

⁶⁷ The copy consulted was the reprint (New York, 1928) of the Vocabulista arávigo en letra castellana.

⁶⁸ This applies not only to the standard dictionaries of C. de las Casas, Sir R. Percivale and J. Minsheu, C. Oudin, G. Vittori, L. Franciosini, and the like, but also to the rhyme dictionary of J. Díaz Rengifo, Arte poética espñola con una fertilissima sylva de consonantes (princeps cf the year 1592). The printing of the earlier rhyme dictionary by P. Guillén de Segovia seems to have been interrupted by the Civil War in Spain.

⁶⁹ To remain within the word family here studied, compare ‘honey’ with Attic ‘I collect honey’.

⁷⁰ It is questionable whether prōmulgō, -āre ‘to make public’ is a mere variant of prōvolgō, -āre, as W. A. Baehrens, Sprachlicher Kommentar zur vulgärlateinischen Appendix Probi 99-100 seems to suggest (following the clue of Paulus Festus) ; Ernout-Meillet, Dictionnaire étymologique² 814, link it to mulgeō, -ēre Ho milk, to drain, to squeeze'. But the confusion between globus ‘sphere’ and glomus ‘ball of thread, tangle’, doubtless prompted by the outward similarity of the two objects involved, is on record in Appendix Probi, No. 71; see Baehrens, loc.cit.

⁷¹ W. Meyer-Lübke, Grammatik der romanischen Sprachen 1.479 (Leipzig, 1890-904); P. E. Guarnerio, Fonologia romanza 606, 622-3 (Milan, 1918). These and other earlier writers classed this permutation with cases of assimilation and dissimilation. Today, this permutation is more likely to be interpreted as an example of acoustic equivalence, to use the term proposed by A. Alonso, BDHA 1.440-55, and accepted by Menéndez Pidal, Manual de gramática histórica española⁶ 194-201. Epigraphic material, available for the wavering in Hispano-Latin between b and v, is not extant for the alternation between m and b, v; see A. Carnoy, Le latin d'Espagne d'après les inscriptions² 128-41 (Brussels, 1906).

⁷² See G. Rohlfs, Baskische Kultur im Spiegel des lateinischen Lehnwortes, in Philologische Studien aus dem romanisch-germanischen Kulturkreise, Karl Voretzsch . . . dargebracht 58-87 (Halle a.S., 1927). The following cases are here assembled: bacillu > makila, balneu > mañu, balteu > malzo, barca > marka; *fascula > mas̃kla, fēnuculu > mihuli; pālu > maru, pentēcostē > mendekoste, persicu > mirtchika, pīca > mika;vāgīna > magina, vāscellu > maskelu, venerārī > meneratu, vindicāre > mendekatu. According to Rohlfs' (by no means exhaustive) list, the shift in Basque has been unilateral, with b-, f-, p-, and v- yielding to m-. In Ibero-Romance, the exchange has been in either direction; admittedly, the shift m>b has been less frequent than the reverse. See G. Baist, Spanische Sprache, Gröber's Grundriss der romanischen Philologie² 1.896; J. Cornu, Portugiesische Sprache, ibid. 1.964, 984; T. Navarro Tomás, Cuestionario lingüístico hispanoamericano: fonética, morfología, sintaxis 31-2 (Buenos Aires, 1943).

⁷³ A. Kuhn, Der hocharagonesische Dialekt, RLiR 11.98-9.

⁷⁴ In addition to Kuhn's monograph, which includes information on Gascon, see Rohlfs, Le gascon: études de philologie pyrénéenne 28,114 (Halle a.S., 1935).

⁷⁵ A. Steiger, Contribución al estudio del vocabulario del Corbacho, BRAE 10.168-70, with quotations of earlier studies by Tallgren-Tuulio, Leite de Vasconcelos, Spitzer, and Menéndez Pidal. Cases like bandíbula ~ mandíbula, bengala ~ mengala, bielgo ~ mielgo were known as early as C. Michaëlis, Romanische Wortschöpfung 281, 293 (Leipzig, 1876).

⁷⁶ See BDHA 1.152-3, 164-5, with reference to earlier publications of Baráibar y Zumárraga, Baruch, Batres Jáuregui, Bayo, Cuervo, Echeverría y Reyes, Espinosa, Gagini, García de Diego, García Icazbalcete, García-Lomas, Krüger, Lamano y Beneite, Mangels, Membreño, Montori, Múgica, Pichardo, Ramos y Duarte, Riofrío, Rodríguez Marín, Román, Sánchez Sevilla, Sevilla, Subak, Sundheim, Wagner, and with original material illustrating Argentinian usage. For philological evidence, see the discussion of medicāmine > OSp. vedegambre by Menéndez Pidal, Etimologías españolas, Rom. 29.374-5, and García de Diego, Contribución al diccionario hispánico etimológico 118 (Madrid, 1923).

⁷⁷ Much material has come to light on vīmen, -inis and its plural form vīmina; to the quotation of binbre by Hanssen, Spanische Grammatik 59, from Fuero de Navarra and Primera Crónica General, add the reference to Pero Guillén de Segovia, see O. J. Tallgren [-Tuulio], La Gaya de Segovia 47 (Helsinki, 1907), and to López de Úbeda and E. de Salazar in C. Fontecha, Glosario 47. The same form has been produced by A. Castro, Glosarios latino-españoles de la edad media (Madrid, 1936), E 1804. Blima, bringa are reflexes recorded by B. Acevedo y Huelves and M. Fernández y Fernández, Vocabulario del bable de occidente 38 (Madrid, 1932); blim(b)a, blim(b)al, and further derivatives are registered by J. M. Canellada, El bable de Cabranes 122 (Madrid, 1944); additional products are supplied by V. García Rey, Vocabulario del Bierzo 30 (Madrid, 1934). Bimbrero has survived in Mexico; see P. Henríquez Ureña, Datos sobre el habla popular de Méjico, BDHA 4.291. Notice palma bien avimada in Berceo, Milagros 39.

Here are some other cases of permutation of initial labials from sources recently made accessible. Muriones beside buriones ‘caracoles marinos’ has been listed by J. González Campuzano and E. de Huidobro, Apuntes para un vocabulario montañés, BBMP 2.10. Canellada, op.cit. 262, 266, documents wavering between marmullar and barbullar, registers metrinariu ‘veterinario’. Acevedo y Huelves, op.cit. 142, records magurento beside valorente, Malvina beside Balbina; assembles the offshoots of vermis, -is; plur. vermina (based on *vermen) ‘worm’ : verbene, vierven, verme, vermino, viermo. In Murcia, there exist doublets like baluarte ~ malguarte, bandurria ~ mandurria, vendimia ~ mendema, veneno ~ meneno, vagancia ~ magancia, verienda ~ merienda; see J. García Soriano, Vocabulario del dialecto murciano, p. lxxv (Madrid, 1932). A supplementary example is adduced by P. Lemus y Rubio, Aportaciones para la formación del vocabulario panocho 223 (Murcia, 1933) : borceguíes beside morceguiles. From the Catalan frontier, V. Ferraz y Castán, Vocabulario del dialecto que se habla en la Alta Ribagorza 34 (Madrid, 1934) reports the use of brená ‘merendar’. On the American scene, R. Lenz's record of Chilean distortion of bueno to muön is now available in BDHA 6.170. Henríquez Ureña has listed bembrillo, desboronar, baraña - moñuelo, muñiga, manguardia, espamento ‘aspaviento’ as peculiar to folk-speech in Mexico, see BDHA 4.291, 299. Borondanga and morondanga are heard side by side at the coast of Colombia, see P. M. Revollo, Costeñismos colombianos o apuntamientos sobre el lenguaje costeño de Colombia 54 (Barranquilla, 1942).

Supplementary illustration has been drawn from other sources: OPtg. Almafacem ~ Albofacem (name of a Moorish king): Nunes, Crest. Arc.² 54-5. Cat.-Arag. batahalua, Ptg. balafaluga, OSp. matafalva < Arab, (Castro, RFE 10.115). Biermen (FArag.), bierben (ibid.), berbezón (FNav.): Tilander, Los Fueros de Aragón 274. Borondanga (P. Espinosa) ‘morondanga’: Rodríguez Marín, Dos mil quin, voces 59-60. Burg. camota ‘cabeza de un alfiler’, Val.-Cat. cabota ‘cabeza de un clavo’ < caput (García de Diego, Contr. 43). Caramanchel, carabanchel, camaranchel ‘cueva en el campo’ (Corominas, AIL 2.183-4). OPtg. enciba < encima ‘above’: Crest. Arc.² 211, 567. Fisgar ~ Transm. misgar ‘ver com muita atençáo’ (RL 12.120; García de Diego, Contr. 120). Sp. marbete < Fr. brevet (Spitzer, AIL 2.29); melesca ~ bresca (ibid.); Ptg. melharuco ~ abelkaruco (L. Freire, Grande e Novíssimo Dicion. 3376). OSp. monclura < OFr. embouclure (Spitzer, A1L 2.30). Ptg. mungir ~ espongir ‘to milk’ < ēmulgēre [?] (García de Diego, Contr. 69). Ochamero ‘ochavero’: Alhóndiga, a.d. 1170 (Med. Span. Word-List; influence of Septimus not unlikely). Prominco ‘propincuo’ : Campó, a.d.* 1219 (Docum. ling. de España, No. 23). Finally, notice in García de Diego's Contribución: mala folia > Arag. barfolla, Gran. parfolla, Murc. perfolla, possibly under the influence of valv- (114-5) ; the interaction of merga and ventilāre (119); Burg. mollejo < molliculu beside pollejo, conceivably suggested by polio (122); Ast. pozelles < *mocellas (125); OPtg. provinco, OSp. prominco < propinquu (133-4); Ptg. t(a)ramela, OSp. tarabilla < *trabella (167); Ast.-Cast. muermo < Germ. wurm (186). See also Steiger, RFE, 7.382; D. Alonso, RFE 27.172.

⁷⁸ See V. García de Diego, Elementos de gramática histórica gallega 33, 35 (Burgos, [1909]) : bacalau ~ macallau, bonifate ~ monifate; J. Cornu, loc. cit.

⁷⁹ M. Leumann in his revision of Stolz-Schmalz, Lateinische Grammatik⁶ 227 (Munich, 1928) proclaims mellītus modeled after pellītus, from pellis ‘skin, hide’, and this sequence of events is accepted in Walde-Hofmann, Lateinisches etymologisches Wörterbuch 2.61-2 (1939).

⁸⁰ For full documentation see ThLL 8.622-3, where the material is presented with the usual excellency.

⁸¹ See J. B. Hofmann, Lateinische Umgangssprache² 141-2 (Heidelberg, 1936), on the stylistic trend underlying this usage.

⁸² A fraction of the material offered by ThLL may be quoted here : Cum uxore et filiola et mellito Cicerone (Cicero); mellitus erat passer - mellitos oculos tuos - surripui tibi ..., mellite Iuventi, saviolum (Catullus); Fronto ... mellitissime, meus amor (M. Aurelius); mi mellite, mi morite (Apuleius); nutrix mellitissima fecit alumno suo - marito mellitissimo - patri mellitissimo (CIL) ; patri dulcissimo, mellito - anima dulcís ... anima mellita (Inscr. Christ.).

⁸³ To the quotations from Aurelius, Apuleius, Sedulius, Dracontius, Venantius Fortunatus, should have been added the phrase labra mellita, extracted from Ennodius.

⁸⁴ There exists an extensive literature on the suffix involved; to the sources quoted in note 15 to my essay Relics of mergus, mergulus, and mūcurō in Ibero-Romance (AJPh. 67.151-67), add the pertinent chapters in the monographs of Kuhn (Upper Aragon) and Rohlfs (Gascogne).

⁸⁵ V. de Vit, Totius Latinitatis Onomasticon 4.450.

⁸⁶ On the application of this criterion, see the article Hispanic Reflexes of Latin morsicāre, Phil. Quart. 24.233-54.

⁸⁷ This would satisfy the postulate of Meillet, as stated in the preface to his and Ernout's Dictionnaire étymologique de la langue latine: ‘Or, en grammaire comparée, toute preuve s'exprime par la formule: “la concordance ne saurait être fortuite”. Un rapprochement perd donc en valeur probante tout ce qu'il gagne en facilité. C'est ce que les étymologistes perdent parfois de vue.‘ This reproach, addressed to Indo-Europeanists, bears repetition in a Hispanic study.

⁸⁸ The same image emerges in the following lines of Berceo: Los dichos que di:.:a melados semeiavan, | com̃o los que de voca de Gregorio manavan (Vida de Santo Domingo de Silos, ed. J. D. Fitz-Gerald, quatrain 121cd).

⁸⁹ Passing mention may be made of some marginal uses of the word under study. Ptg. belida, Gal. velida is the popular word for albūgō (beside learned albugem); see L. Freire, Grande e Novíssimo Dicionário da Língua Portuguesa 1011; M. Lugris Freire, Gramática do idioma galego 153 (La Coruña, 1922). The phrase ter belidas nos olhos is tantamount to ‘pretending not to see’, according to H. Michaëlis, Neues Wörterbuch der portugiesischen und deutschen Sprache¹³ 1.119 (Leipzig, 1923). Is it correct to speak of similarity with the case of la niña del ojo ‘pupil of the eye’, known to be modeled on an Arabic archetype? In cant, vellido means ‘velvet’; see R. Salillas, El delincuente español : el lenguaje 309 (Madrid, 1896); L. Besses, Diccionario de argot español 168 (Barcelona, n. d.); cf. Juan Hidalgo, Vocabulario de germanía, apud G. Mayáns y Siscar, Orígenes de la lengua española 2.273 (Madrid, 1737); also Poesías germanescas, ed. John M. Hill, p. 108 (Bloomington, 1945): bellido ‘terciopelo’. This meaning the word acquired in imitation of velludo ‘velvet’, as used, for instance, by Góngora ; see the vocabulary of Alemany y Selfa (986). Again in cant, vellida, also spelled bellida, serves, along with bellosa and vedilla, to paraphrase ‘freçada’, i.e. ‘woollen blanket’ (RFE 7.371-2); see the aforementioned sources and R. J. Slabý, Wörterbuch der spanischen und deutschen Sprache 713 (Leipzig, 1932). Here again, the word seems to have been attracted into the orbit of villus, -ī Góngora used bellido for ‘beautiful’ (Obras 2.387: Casado el otro se halla | con la del cuerpo bellido ‘the other finds himself married | to the one [maiden] with the beautiful body‘) but his use of vellido (Alemany y Selfa, Vocabulario 985) is not altogether clear and would require special study, like that of Don Agustín Moreto (see note 19). Evidently by that time ancient vellido was no longer readily understood and thus lent itself to wilful distortion in cant and to fanciful transformation in poetic style.

⁹⁰ The type Mellid(e) is peculiar to Galician-Portuguese. See J. J. Nunes, Nomes de Pessoas na Toponímia Portuguesa, Bol. Acad. Sc. 13.1257-74. A different interpretation of Mellida is proposed by J. M. Piel, Os Nomes Germânicos na Toponímia Portuguesa, BF 5.44, 280.

⁹¹ Villōsus and (infrequent) villūtus are attested derivatives; see Ernout-Meillet, Dictionnaire étymologique² 1108. Vellado occurs in Vida de Santo Domingo, quatrains 23, 669; velloso in Apolonio, quatrain 518; Libro de Buen Amor, quatrains 448, 1013, 1017, 1201, 1486; Carlos Maynes, chapter 1; Enperador Ottas, chapter 2; Grail Fragments, fols. 290ro, 290vo. On vellosero, used as an adjective and as a proper name, see Oelschläger, Medieval Spanish Word-List 213a (Santoña, a.d. 1210). Velludo (Cancionero de Baena 477, many writers of the Golden Age) tended to develop the meaning ‘velvety’, in contradistinction to velloso ‘hairy’.

⁹² It is important to remember that Menéndez Pidal upholds his translation of barva vellida as ‘barba hermosa’ in the editions of the Poema de Mio Cid of the years 1913 and 1923 (p. 143).

⁹³ The connection of melindre with Hispanic m(i)el ‘honey’ has been recognized as early as S. de Covarrubias Orozco, Tesoro de la lengua castellana o española², vol. 2, fol. 107ro: ‘Género de frutilla de sartén hecha con miel: comida delicada y tenida por golosina. De allí vino a significar èsse nombre el regalo con que suelen hablar algunas damas, a las quales por esta razón llaman melindrosas.’ An attempt to link melindre to mollis ‘soft’ is mentioned by P. F. Monlau, who prefers, with the Diccionario de Autoridades 4.534, to explain it as an inversion of ύδρόμ∈λι. Diez has not concerned himself with melindre, to judge by the index to his dictionary prepared by J. U. Jarník. New theories were advanced by the Norwegian scholar Storm (Roman. 5.181) : melindre was supposed either to reflect melitinus (an adaptation of , signifying ‘mixed with honey’ in Plinius Valerianus, putative author of the treatise De re medica assigned to the 4th century; the word was applied, perhaps on account of the color, to a variety of stones by Pliny, Isidor of Seville, and others, ThLL 8.619) or else mellītulus; if the latter conjecture was correct, melindre was explicable either as the outgrowth of *melilde > *melinde or as a parallel to amygdala > almendra ‘almond’, the shift to d being due to l. It should be borne in mind that 70 years ago the conjectural reconstruction of bases had a vogue, while knowledge of the actual phonological modifications was in its incipient stage. The second theory of Storm was acclaimed by F. A. Coelho (in the supplement to his Dicionário Manual Etimológico da Língua Portuguesa), G. Körting (throughout the three editions of his Lateinisch-Romanisches Wörterbuch, 1891-1907), the compilers of the Academy Dictionary, Alemany Bolufer, A. de Pagés, and yet others. C. de Figueiredo, Novo Dicionário da Língua Portuguesa⁴ 2.155, was unwilling to concede more than general relationship between Lat. mel and Ptg. melindre. Cajador y Frauca, La lengua de Cervantes 2.720, suggested distortion of a fantastic base [*]melindera. Meyer-Lübke, who lacked accurate knowledge of the range of meanings of the word (inadvertently, he may have substituted ‘süsslich’ for Körting's ‘süssliches Wesen'), classed it provisionally with the word family of mel, but explicitly stated the need for further study of the suffix involved (REW³ §5469: ‘Sp. melindre ist lautlich schwierig, das Suffix bedarf noch der Erklärung'). A. Nascentes, Dicionário Etimológico da Língua Portuguesa 500 (Rio de Janeiro, 1932), had no opinion of his own. Spitzer's recent derivation from OFr. melide (AIL 3.19-21) is not convincing. Mellīulus is attested in Apuleius and St. Jerome.

⁹⁴ Menéndez Fidai, Manual de gramática histórica⁶ 160: ‘En voces semicultas, tanto t'l como d'l trasponen sus dos elementos, trocándose en ld: capitulu > cabildo, titulu > tilde, rotulu > rolde, modvlu > molde; *anethulu (por anethum) > aneldo, eneldo; spatula > espalda; foliātile (esto es, pānis foliātilis) > ant. hojalde, mod. hojaldre. Estas voces son tardías y semicultas, a juzgar también por la -e final, en vez de -o, que ofrecen algunas, y por la vocal acentuada de las cuatro primeras.’ On OSp. pildora, OPtg. pilora ‘pill’ < pilula, see C. Michaëlis de Vasconcelos, RL 13.402. On bul(d)a, cel(d)a, libel(d)o, péndola beside péñola, see A. Castro, Glosarios latino-españoles de la edad media, p. lxix; the forms with -d- are traceable to the 14th century.

⁹⁵ The best-known case perhaps is mudre beside mugre ‘dirt’ in the Vocabulario universal of A. de Palencia (1490). García de Diego, RFE 7.139, compares Sp. mugre, mudre to Ptg. Indro, Gal. ludre in an effort to characterize them as back formations from lutulentus; his theory is not convincing, in view of lutum > Sp. lodo ‘dirt’. Mugre is more likely to reflect mūgil, -ilis, which fundamentally means ‘adhesive, glutinous, sticky substance’. Ludro and ludre should be dissociated from mugre, mudre and classed as offshoots of Lat. glūten. For more details on -dre, -gre, see excursus B. OSp. mudre ‘dirt’ is explained as a cross of mūcor and putris by García de Diego, Contrib. 124.

⁹⁶ Olingre has recently been identified by G. Sachs in El Libro de los Caballos: tratado de albeitería del siglo XIII 75, 98, 140 (Madrid, 1936) ; follín and hollín ‘soot’ occur in the same text (67, 102). On the history of fullīgō, see García de Diego, Divergentes latinos, RFE 5.135-8, and especially A. Steiger, Sobre algunas voces que significan ‘hollín’ en las lenguas románicas, Homenaje ofrecido a Menéndez Pidal 2.35-48. Steiger localizes la holingre in Burgos, los olingres in Santander. Comparison with Rum. funingine, ORum. furi(n)dzine, Mac. Rum. fulidzine, furi(n)dzine enables him to claim that the attraction of -īgine into the orbit of sanguine, inguine may go back to the 4th century.

⁹⁷ C. Michaëlis' plan to write a study on the complex development of these suffixes has not materialized (RL 13.3S8). For brief statements, see Cornu, Grundr. rom. Phil.² 1.991, and my note The Suffix -āgō in Astur-Leonese-Galician Dialects, Lang. 19.256-8, AJPh. 65.378. There must have existed several channels of transmission. Compare Sp. plantel with Cal. chantage < plantāgine; the former shows an arrested development of the radical and an advanced form of the suffix, the latter exhibits the opposite distribution. Farrāgine has left a host of reflexes in Old and Modern Portuguese: ferragem, ferrãe, farrãe, ferrém, farrém, farrã, ferrã, farrão, ferral, beside Gal. ferraya, Mir. ferranha (RL 13.317-8); Old Spanish documents contain ferragine, ferreine, ferrenne, ferrenio, ferrein, ferrén (Oelschläger, Med. Span. Word-List 93). Sariāgine yielded OPtg. sartén, sartán; Sp. sartén; Arag. sartana, Basque zarthagiña (RL 13.319; K. Voretzsch Festschrift 82). Cartilāgine produced OPtg. cacragem beside cachagem, with r disappearing before the sibilant as in Sp. sacho, macho, viesso. To the previously classified gordén, tristén type add podrén ‘something that is rotting’ (Canellada, El bable de Cabranes 302). The interpretation of podrén as a cross of putrāmen and *putrīmen (García de Diego, Contr. 135-6) is inadequate. Earlier interpretations of andén (listed, for instance, by Monlau) have been superseded by Horning's derivation from *andāgine (ZRPh. 29.540-2), accepted by Hanssen (Gramática histórica 138) and Meyer-Lübke (REW³ §447), yet the recent publication of an astronomical treatise of Enrique de Villena demonstrates that by the year 1400 andén stood for ‘orbit of a star’ (RFE 27.22, 23, 24, 25) and thus perpetuates Lat. ambāgō, -inis, related to ambāgēs ‘path, lane, orbit’; ambāgine was perceptibly remodeled in Spanish after *ambitāre > andar.

The variant -ūgine has not survived as a separate morpheme outside of the Western zone. Witness Gal. fcrrugem < ferrūgine beside Algarve ferruja (Corominas, Indianoromanica 14; reprint from RFH 6), OJudeo-Sp. ferrucha (Gaspar Remiro, BRAE 4.635), enferoǧado (Coplas de Yócef, ed. González-Llubera, line 147), Sp. herrín, herrumbre, also herrimbre (Herrera, Agricultura; see Rodríguez Marín, Dos mil quinientas voces 203). On Western forms like herrume, jarrume, jerrumble, see RFE 23.231-5. Ferrojar, ferrumbre occur in Canc, de Baena, see Voc. Med. Cast. For further data see Krüger, El dialecto de San Ciprián 86; also Mezcla de dialectos, Horn. MPid. 2.134.

The suffix -īgine shows the most complicated development. It is customary to regard robīgine > robín as the typical product, yet this may be due to over-simplification of the facts. In Catalan, -īgine yielded -itja; see L. Spitzer, Lexikalisches aus dem Katalanischen und den übrigen iberoromanischen Sprachen 47-8 (Geneva, 1921). In Portuguese, one more and one less advanced variants are on record: fuligem beside esfulinhar ‘basculhar, varrer; limpar teias de aranha; limpar, tirar a fuligem das chaminés’ (Steiger); cf. farragem ~ farrãe. Lentīgō survives only in Gal. lentegoso ‘pecoso’.

The three allied suffixes have been ceaselessly influenced by each other and by cognate terminations. Association with robīgō may have caused aurūgō to give rise to the variant from aurīgō, which was later confused with aerūgō, producing Sp. orín ‘rust’, according to J. Brüch, Miscellanea di linguistica dedicata a Hugo Schuchardt 39 (Geneva, 1922) ; see also RL 13.356; Manual⁶ 70, 231; García de Diego, Contr. 18; Corominas, AIL 1.173-5. Steiger explains Ptg. fulugem, Gal. fuluge, f(e)luge ‘soot’ as due to the influence of ferrūgine > ferruge and knows (through J. Jud) that traces of this interference are perceptible in Northern Italy. He discovers that the variant holtén ‘soot’ (Logroño) reflects -āgine, not, as García de Diego had surmised, -ǐgine. Fullīgine was contaminated by -īmen in Central Raeto-Romance (Steiger supports Ascoli, AGI 1.369), perhaps also in Valencian; by -īdine, possibly also by -īgia in Central Ladino and in Lower Engadine (HMP 2.41). Not yet definitively classified is the group xorguina, enjorguinar, jorguín, holguín, conceivably Basque; see Larramendi; Diez; Baist, ZRPh. 5.244, 7.122 and RF 3.516; Meyer-Lübke, ZRPh 8.225; García de Diego, RFE 5.135-8; Steiger, HMP 2.42; Rohlfs, Festschrift Voretzsch 68; Castro, Glosarios 254. Wavering between -ūdine and ūgine appears in OProv. encluge ‘anvil’; perpetuated in modern patois, see Bloch-Wartburg, DELF 1.253. Fr. enclume is not clear; VLat. *incūdō, -inis has been supposedly influenced by inclūdere (Ascoli, AGI 3.399); has yielded subsequently dissimilated *encnume (Niedermann, AR 5.445-6); has suffered agglutination with, and subsequent transposition of, the definite article (Niedermann, Mélanges Meillet 107) ; has merged with Gall. *klădo ‘ to split’ (Gamillscheg, EWFS 357); has been remodeled after clou < clāve (Dauzat, DELF 277); is due to confusion between excūdō and exclūdō (Spitzer, AIL 2.16-7). Basque ingude, ingure perpetuates incūde, while unkide shows a violent distortion (Voretzsch Festschrift 76) ; Sp. yunque seems to go back to *idunke, arrived at after the assibilation of [k]; the -l- of enclume reappears in Ast. yuncla recorded by J. M. Canellada. In some dialects, yunque is feminine (see Krüger, El dialecto de San Ciprián 128) ; most words in -ín, -én have become masculine, in America even sartén; see C. E. Kany, American-Spanish Syntax 6 (Chicago, 1945).

⁹⁸ A systematic study of suffix-change in Ibero-Romance, comparable to the book of G. Cohn on Old French shifts (1891), remains to be written. I assemble here a few references: -ex ~ -ica: Rohlfs, Voretzsch Festschrift 71; -ōriu ~ āriu: Menéndez Pidal, Manual⁶ 231; -an ~ ano: Manual⁶ 232; -in ~ ino: Michaëlis, RL 13.401-2; Sánchez Sevilla, RFE 15.165-6; Menéndez Pidal, Manual⁶ 232; -ol ~ -or in Ptg. (RL 3.162); -al ~ -ar: Wagner, VKR 3.87-92; -itūdine ~ūmine: Cornu, Rom. 7.365, 9.96; Port. Gr.² §§31, 195; Machaëlis, RL 13.331; García de Diego, Elem. gram. hist. cast. 191 (G. Sachs discovered legūmine > legunde in Old Leonese); -āmine, -īmine, -ūmine: C. Michaëlis, RL 13.314 (strāmen > Ptg. estrume), Steiger, BRAE 10.49-51 (līmināre > l-umbral); on quejambre beside quejumbre, Lang. 21.177; -īgine ~ -ūgine: in addition to the abovementioned cases, cf. robīgine > Ptg. rabuge(m)_t Can. rabuja, and calīgine > OGal. calume, OSp. columbre, Ptg. carugem, caruja > Amer.-Sp. garúa, beside Ptg. caligeiro, caigeria (carefully analyzed by J. Corominas, Indianoromanica 14); -az ~ -aço ~ -aco: RL 13.336 and Manual⁶ 231; -iz ~ -iço ~ ico: Manual⁶ 231; -uz ~ uzo: Baist, RF 34.468-9; -age < āticu and -age < -āgine in Portuguese: Lang. 18.54; -ǐnus ~ īnus: AJPh. 65.372-8; -īgine ~ -īna, in an effort to bring out the gender: Herzog, Festschrift Mussafia 487-501; pre-IE ăru ~ -ălu ~ -ănu: C. Michaëlis, RL 13.296 and BH 9.194-6; Menéndez Pidal, Manual⁶ 228-9 (also in Festschrift Mussafia and Orígenes del español). Of course, not all explanations so far proposed are equally convincing; peldaño need not be a cross of *pedīle, pedāneu (Manual⁶ 185-8), if we agree that l may have arisen spontaneously before a voiced dental stop as in galdrufa, see Asín Palacios, BRAE 7.356-8. Marginal to our problem are cases of assimilation of a meaningless termination to a suffix (Ar. alfanīqa > OPtg. alfenina ‘saccus magnus’, RL 13.232-3) and of confusion between meaningless terminations (RL 13.233, 237, 257, 261-3, 394-5).

⁹⁹ Mellīgō was patterned after mollīgō according to M. Leumann in Stolz-Schmalz, Lateinische Grammatik⁵ 241; accepted in Walde-Hofmann. The derivative mellīginōsus occurs in Pseudo-Dioscorides. The passages in Pliny are quoted in extenso ThLL 8.620.

¹⁰⁰ See Lang. 20.108.

¹⁰¹ The word family of melindre is small: Murc. melindrillo ‘bocadillo, alimento que los trabajadores del campo suelen tomar a media mañana (García Soriano; misinterpreted in the AcademyDictionary) ; melindrear and melindrizar ‘to act the prude', the former attested in Polo de Medina and Fr. Damián Cornejo, the latter in Polo de Medina and Tirso de Molina, see Dicc. Aut. 4.534; C. Fontecha, Glosario de voces comentadas 235 (Madrid, 1941); melindrero ‘prudish, dainty, fastidious’ occurs in La Pícara Justina (Dicc. Aut.); its synonym melindroso in Mariana, Cervantes, La Pícara Justina, El Inca Garcilaso, Luis de Ulloa, Jovellanos. For melindrosamente, see the quotation from Fr. Agustín de Jesús María (1648) in Mir y Noguera, Rebusco 497. Melindreria is known to me only from dictionaries. The variant form melindro, quoted from Fr. Damián Cornejo by Mir y Noguera, Rebusco de voces castizas 496, may be due to an early misprint.

Melindre in its culinary acceptation is but occasionally found in literature: Gracián, El Criticón 2.113 (ed. Romera Navarro); Pragmática de tasas de 1680 (Dicc. Aut.); Pardo Bazán, in Pagés 3.752. In its figurative sense, it is amply documented from the late 16th century (Cervantes, Quevedo, Tirso de Molina, Fr. D. Cornejo), see Cejador y Frauca, La lengua de Cervantes 2.720; Dicc. Aut. ; Pagés. Alemany y Selfa, Vocabulario de las obras de don Luis de Góngora 632-3 (Madrid, 1930) claims that this poet used melindre as ‘finical, too particular person’; this meaning is dubious; on the quality of the vocabulary, see D. Alonso, RFE 18.44-55. On the phraseology of melindre (hacer melindres, no andar con melindre, hablar con melindre, de regalo y de melindre, sin melindres, a lo melindroso), see Cejador y Frauca, Fraseología o estilística castellana 2.89 (Madrid, 1921-4) ; J. Mir y Noguera, Frases de los autores clásicos españoles 492 (Madrid, 1899), and the dictionary of Slabý. Melindre occurs in the refranero; see Correas, Vocabulario de refranes 308 (Madrid, 1924). Notice Lope de Vega's play Los melindres de Belisa (1617), rewritten by C. M. Trigueros (18th e.) under the title La melindrosa o los esclavos supuestos. Data on present range and frequency are found in M. A. Buchanan, A Graded Spanish Word Book² 160 (Toronto, 1929). Absent from medieval Latin-Spanish glossaries, from Nebrija, P. de Alcalá, C. de las Casas; listed in Sánchez de la Ballesta, Pereivale-Minsheu, Oudin, Vittori, Franciosini, de La Porte, Sobrino, Pineda.

In Portuguese, melindre also designates a kind of sweets (as early as R. Bluteau, Vocabulario Português e Latino 5.410-1); figuratively, it stands for ‘delicacy, tactfulness, scrupulousness’; also, it is used in reference to certain species of flowers of the balsamina family, as in Filipino Spanish; see the dictionaries of Fr. Domingos Vieira, C. de Figueiredo,L. Freire, F. J. Santamaría, W. E. Retana. Melindroso is fully as important as in Castilian and has even evolved some additional significations, clustering around the central meaning of ‘delicate’ ('weak, lacking vigor, easily impressionable, innocent of malicious intent, precarious, risky, perilous'). There exist several derivatives: melindrar ‘to offend, to hurt, to scandalize’ ; melindrante ‘ offensive’ ; melindrável ‘delicate, touchy ‘ ; melindrabilidade, melindrice, melindrismo ‘sensitivity'. Brazilian usage is explained and documented by C. Teschauer, Novo Dicionário Nacional² 607 (Pôrto Alegre, 1928). The important variant melingre in the Minhoto dialect is listed by C. de Figueiredo, Novo Dicionário⁴ 2.155, with the comment: ‘entreter o tempo fingindo que se trabalha ou se come, e gastá-lo afinal em trejeitos ou momices’. Melindre, unknown to Old Catalan, has penetrated into Valencian; see J. Escrig y Martínez and C. Llombart, Diccionario valenciano-castellano 854 (Valencia, 1887), a not very reliable book. On the source of the medial -l-, see note 56.

¹⁰² Mel, mellis has survived all over the Iberian Peninsula; earliest occurrence in Spanish, according to Oelschläger: Himnos, quatrain 9; earliest occurrence in Portuguese: a.d. 88, see N. P. Sacks, The Latinity of Dated Documents in the Portuguese Territory 167 (Philadelphia, 1941) ; Old Catalan material in Aguiló i Fuster 5.90. This stem is consistently avoided in the terminology of apiculture, as follows from the careful study of W. Brinkmann, Bienenstock und Bienenstand in den romanischen Ländern (Hamburg, 1938), nor does it enter into the designation of the insect itself, in contrast to the situation in Greek and in certain varieties of French (mouche à miel ‘bee’, studied by J. Gilliéron in 1918). On contamination of mel with Greek μλι, see note 56; compare the conflict between native mālum ‘apple’ and imported mēlum < μλοv, briefly described by Ernout-Meillet. On the merger of malum and mel, see García de Diego, Contr. 113, s.v. malum cotonium.

¹⁰³ See Fr. Joaquim de Santa Rosa de Viterbo, Elucidário² 1.182 (Lisbon, 1865), s.v. apeiro; Diccionari Aguiló 5.90; Oelschläger, Medieval Spanish Word-List 130, with reference to a Mozarabic document of the year 1214.

¹⁰⁴ Figueiredo, Novo Dicionário⁴ 2.155; Acevedo y Huelves, El bable de occidente 148; Cuveiro Pinol, Diccionario gallego 204. The forms are meleiro in Astur-Galician, melieiro in Portuguese.

¹⁰⁵ For Old Spanish, see Steiger, BRAE 10.185, 279 (El Corbacho); for modern dialects, see A. Cabrera and J. Alemany [Bolufer], Voces extremeñas recogidas del habla vulgar de Alburquerque y comarcas, BRAE 4.95: ‘dícese de la bellota que contrae una enfermedad por la cual se desprende del cascabillo y muere’. The form of the suffix -ioso is striking but not abnormal; it is difficult to decide without special inquiry whether it is due to epenthesis of [j], not uncommon in Western dialects, or to contamination, or to the conflict between -l- and -ll-.

¹⁰⁶ See A. de Rato y Hevia, Vocabulario de las palabras y frases bables 82 (Madrid, 1891). Mellicus lives on in Ast. esmielgar ‘quitar la miel de las colmenas’ (Canellada); in Ptg. melgueira ‘cortiço com favos de mel ; dinheiro que se junta às ocultas ; lucro inesperado’, with the more emphatic variant remelgueira; in Ptg. melgaço ‘louro’; and in the important Spanish word family remilgarse ‘to be affectedly nice’ (Moreto, see A. de Pagés, Dicc. 4.781a), remilgado ‘squeamish’ (Tirso de Molina, see ibid.), remilgo ‘exaggerated prudery’ (Flora Osete, ibid.), as distinct from remelar ‘to sweeten’ (see J. Mir y Noguera, Rebusco de voces castizas 628 [Madrid, 1907]). In Asturias, arremelgar has come to signify ‘threaten’ (Canellada). This branch must be very old; melliculus, -a, -um ‘honeysweet’ and melliculum, -ī ‘honey-pel ‘ occur in Plautus. There are also some dubious formations tentatively connected with the word famity of mel. Jud's reconstruction of a type *melix ‘larch-tree’ (cross of larix and mel) for Piedmont and Savoy is explicitly rejected REW³ §5481a. Spitzer's explanation of Sp. melescar ‘recoger restos utilizables’ as a blend of brisca and miel (AIL 2.176) has been replaced by Corominas' derivation from mariscar (AIL 2.180-1). Desmelgar ‘desmelar’ has been identified in El Bierzo by García Rey.

¹⁰⁷ Körting, Lateinisch-Romanisches Wörterbuch³ §6070 rejects *melligera as a base of Sp. mangia, Ptg. mangra. Diez thought that these words basically meant ‘honeydew’. The actual use in Old Portuguese is clearly stated in Elucidário² 2.74-5: ‘Orvalho da névoa que tolhe e não deixa medrar as searas. ... Em algumas províncias fóra deste reino se livram deste dano, meneando suavemente o trigo, como faz o vento. Para isto dous homens tomam pelas pontas urna corda, e caminhando com ella estendida na altura das espigas, vão estas sacudindo de si a perniciosa mangra.’ Add to this definition A. Cabrera's information on the use of mangria in Spanish Extremadura (BRAE 4.94): ‘Enfermedad de las sandías, que les cubre las hojas de una costra negruzca y pegajosa e impide que el fruto madure en las debidas condiciones'. G. Baist proposed the base *mannula (ZRPh. 5.562), later macula (ZRPh. 7.121) ; the latter explanation was accepted by Mcyer-Lübke, see REW³ s.v. macula. Added weight is given to this theory by the occurrence of mandra ‘aprisco, majada’ in Covarrubias and Franciosini (2.490), recently pointed out by R. Menéndez Pidal, La Etimología de Madrid y la antigua Carpetania, RBAM 14.3-4; mandra perpetuates macula ‘mesh’, cf. redil. As will be shown elsewhere, Sp. majada is an outgrowth of macula, not of magālia.

¹⁰⁸ Not related to our word family is Arapenese-Gascon mèlic, melico, meligo, which continues umbilicus ‘navel’; see the word lists of B. Coll y Altabás, L. V. López Puyoles and Valenzucla la Rosa, V. Ferraz y Gastán, A. Kuhn, G. Rohlfs, P. Arnal Cavero (1944).

¹⁰⁹ A. Álvarez de Villasandino in the Cancionero de Baena (Madrid, 1851), fol. 53ro: las brevas melares. I owe the reference to Cejador y Frauca, Vocabulario medieval castellano 272 (Madrid, 1929). Compare cerezos soldares, quoted by Rodríguez Marín, Dos mil quin. voces 354.

¹¹⁰ A. de Pagés, Gran diccionario 3.750 : ‘Hazer las abejas la miel ; en los ingenios de azúcar, dar la segunda cuchura al zumo de la caña’; J. Lamano y Beneite, El dialecto vulgar salmantino 534-5 (Salamanca, 1915): ‘marcarse la bellota por excesivo calor, destilando un zumo [la mela] muy parecido a la miel’. Mella, -ae ‘honey water’ is, incidentally, recorded in Latin; there is no way of proving that it is organically connected with Sal. mela (the character of the medial consonant is no valid criterion).

¹¹¹ See Berceo, Vida de Santo Domingo de Silos, quatrain 121.

¹¹² See J. Calcaño, El castellano en Venezuela: estudio crítico 508 (Caracas, 1897), and T. Sibaudet, Vocabulario y refranero criollo 243 (Buenos Aires, 1943), who provides the following definition: ‘Caballo cuyo pelo es todo blanco plateado con reflejos de color miel, ojos negros y piel negra o manchada’. The designation of the colors of horses has been widely debated by students of Spanish; see G. Bertoni, I nomi spagnuoli dei colori del cavallo nel MS di Leida CLXX (231 Scal.), HMPid. 1.151-4; D. Granada, Los nombres de los colores del caballo en América, BRAE 7.628-32, 8-58-63, 187-98, 349-60; A. Alonso, Preferencias mentales en el habla del gaucho, in El problema de la lengua en América 143-80 (Madrid, 1935).—Directly developed from the noun, rather than from the verb, are both the adjective melado and the substantive melada ‘slice of toasted bread, soaked in honey’, also found in Catalan (Aguiló i Fuster).

¹¹³ F. J. Santamaría, Diccionario general de americanismos 2.271 (Mexico, 1942), lists Chil. mellingue ‘se dice de la zorra que tiene las patas de color castaño claro’; on the suffix involved, see J. U. Hubschmied, Romanisch -inco, -anco, in Mélanges A. Duraffour 211-70 (Paris-Zurich, 1939). Is Sal. melengín ‘endeble, enteco’ (Lamano, Dial. vulg. 534) an offshoot of this formation, with the semantic development known from melindre?

¹¹⁴ Melcocha is listed in the oldest dictionaries, including that of Nebrija. It occurs in J. Rufo (Pagés, Gr. dicc. 3.751), Saavedra Fajardo (C. Fontecha, Glosario 235). The derivation is discussed by Hanssen, Gramática histórica 164. Melcochero ‘gingerbread baker’ is also very old; see Castro, Glosarios E 1485; Nebrija's dictionary; a passage from J. Rufo in Pagés. Less well known is amelcochar, recorded by J. M. Dihigo, Léxico cubano 200 (Habana, 1928) : ‘Dar a un dulce el punto espeso de la melcocha; fingir agrado, afectar complacencia; enamorarse’.

¹¹⁵ Meloxa, meloja is the graph found in early dictionaries. Cuban meluza is explained by Santamaría, Dicc. gen. amer. 2.271 : ‘Humedad pegajosa o residuo de la miel o de azúcar’. A quotation, without comment, of meluça from Fr. P. Beltrán's unpublished La caridad guzmana is offered by Rodríguez Marín, Dos mil quinientas voces 243-4.

¹¹⁶ Fr. Domingos Vieira, Grande Dicionário Português 4.197: ‘Iguarias em que entra mel branco, a maneira de murcelas, feitas de amêndoas pisadas, açúcar em ponto, pão de rala, etc.’ On Sp. mellizas, see Dicc. Aut. 4.534: ‘Cierto género de salchichones hechos con miel de donde tomaron el nombre.’ Only through accurate information on the manufacturing of this tidbit can one determine, whether mellizos ‘twins’ exercised a measure of influence on the formation. Mellicus and (ge)mellicus resembled each other to the point of confusion ; melguizo ‘twin’ is used in an extensive territory; see J. Borao, Diccionario de voces aragonesas² 266 (Saragossa, 1908) ; J. Puyol y Alonso, Vocablos usados en León, RH 15.6.

¹¹⁷ In addition to the well known parallel cadāver > Sp. calavera ‘skull’ notice the wavering between the proper names Medica and Melica, registered by Ernout-Meillet, Dict. étym.² 603, s.v. mēlicae, -ārum.