¹ Revista de fil. esp., xix, 117-150, 229-260.

² Williams, From Latin to Portuguese, §33.4.

³ Gamillscheg, op. cit., 255.

⁴ Williams, op. cit., §74.1.

⁵ Diccionario etymologico da lingua portuguesa.

⁶ REW ³, No. 6441.

⁷ Possibly a misprint, since the second edition has firmar.

⁸ Archiv. f. d. Studium d. neuern Sprachen, cxxvii, 159 f.

⁹ Zur Kenntniss des Altlogudoresischen 64. REW ^2,3 No. 3332.

¹⁰ The word is not in Lewis and Short's dictionary, but Diez (Et. Wb. der Rom. Sprachen 371) quotes from Isidore the gloss Fistilor sibilo.

¹¹ Die Gegenstandskulturen Sanabrias und seiner Nachbar gemeinden, 76.

¹² Op. cit.

¹³ Op. cit., §92.9, p. 97.

¹⁴ Rom. Gram., 2.611, §92.9.

¹⁵ Lang., xviii, 29.

¹⁶ Cf. Spitzer, Hisp. Rev., x, 344, who accepts the semantic shift ‘snarl’ > ‘gnaw’; questions the form ringari erroneously written by me for ClL ringi; questions the alleged Old Port, relhar quoted by me in reporting Prof. Holmes's theory deriving Port. rilhar from Lat. religare; and prefers to derive Port, rilhar from the already posited ∗ringulare ‘snarl.’ The Johns Hopkins professor here predicts that etymologists will soon be limiting their etymological bases to the types already recorded in REW ³. Cf. Rev. de fil. esp. 13.116, where Spitzer attempted to derive Sp. sesgar ‘to cut across’ from ∗sessicare ‘to set,’ instead of the ∗sesecare proposed by me and adopted in REW ³. I now recognize that Ulrich's etymon ∗sexicare (:∗sexus for sectus ‘cut’) is more probably correct, in view of Hispanic participles in -x- for -ct-: ∗dejexare > Sp. dejar, ∗coaxare > Sp. quejar (Hisp. Rev., vi, 351; ix, 309 f.), as well as Lat. fixus fictus, fluctus fiuxus, frixus frictus, etc. (Hisp. Rev. vi, 351). ∗Sexicare was first proposed in Z. f. rom. Phil., iv, 383.