^{Note 1 in page 243} George Tyler Northup, “La selva confusa de Don Pedro Calderon de la Barca,” Revue Hispanique, XXI (1909), pp. 168-338. (Not having a copy of the reprint at hand, I am obliged always to refer to the page numbers of the Revue.)

^{Note 2 in page 243} By fragmentary copy reference is here made, of course, to the so-called Osuna volume 133, which, together with its companions 131 and 132, was briefly described by Schack in his Nachträge (1854), pp. 41-43, and was used by Menéndez y Pelayo in the preparation of vols. IX and X (1899) of the Academy's edition of Lope de Vega, but has since disappeared along with the others. On these Osuna volumes and their relation to the “partes extravagantes” of Lope, see H. A. Rennert, “Notes on some Comedias of Lope de Vega,” MLR, I, pp. 96-110: cf. pp. 98-103.

^{Note 3 in page 243} He stated, however (p. 170), that he might have occasion to change his mind on this question if he were ever “allowed to consult” the missing Osuna volumes; but I have not learned that he has yet turned his attention to this matter again.

^{Note 4 in page 243} Op. cit., p. 172.

^{Note 5 in page 243} The latter somewhat hesitant: cf. Northup, p. 173, continued footnote. But in his “Bibliography of the Dramatic Works of Lope de Vega,” Revue Hispanique, XXXIII (1915), p. 241, Rennert appears definitely convinced that the play is Calderon's. He retains the title in his catalogue of Lope's comedias only because, apparently, he could not overlook Schack's statement that the play existed as of Lope in an “alter sehr seltener Druck.”

^{Note 6 in page 244} For an account of the finding of this volume, together with a brief description of it, see my article “Lope de Vega's Parte XXVII extravagante,” RR, XV, pp. 100-104.

^{Note 7 in page 244} So true is this that Menéndez y Pelayo (Academy ed. of Lope de Vega, XIII, p. cxviii) had to admit his inability at times to distinguish the style of Mira de Amescua from that of Lope de Vega.

^{Note 8 in page 244} A clue to such an undertaking is furnished by Northup himself, who in his introduction and some of his footnotes points to Tirso de Molina.

^{Note 9 in page 245} Restart (Saggi de bibliografia leatrale spagnuola, Geneva, 1927, p. 55, note 16) seems inclined to entertain a doubt as to the genuineness of the title-page and preliminaries of this volume, and consequently, I infer, as to the date. I have attempted to dispel this doubt in RR, XIX, p. 256.

^{Note 10 in page 245} It must be assumed that Schack's “Seite 1 bis Seite 146” is the equivalent of “fols. 1-146.” I am puzzled, however, by Schack's mention of two parts to Lanza por lanza de Luis Almanza. In the Barcelona copy there is nothing to indicate that there was a second part, nor, if such a play was ever printed, is there any room for it as an organic part of the volume. If it really was in the Osuna fragment in Schack's time (it was not there when Menéndez y Pelayo used the volume), it must have been a suelta inserted after the first part, since, as I point out in footnote 14, this part of the volume is so made as to be divisible into desglosadas.

^{Note 11 in page 245} Perhaps it is not so strange, after all. The gathering of the notes which comprise the Nachträge extended over a number of years, during which time Schack was primarily interested in other matters (cf. “Vorbemerkung,” p. III). It is probable that the two notes in which La selva confusa is mentioned were made at widely separated moments: hence the author's failure to investigate the relation of the MS to the printed edition.

^{Note 12 in page 246} Cotarelo y Mori, in Ensayo sobre la vida y obras de D. Pedro Calderón de la Barca, Parte primera, Madrid, 1924 (reprinted from BRAE, VIII-X), states (BRAE, IX, p. 36, footnote 1) that there is an anonymous suelta of the play dated 1791; but this statement, as I have learned by correspondence with Señor Cotarelo, is due to an erroneous interpretation of Northup's notice (op. cit., p. 169) of a suelta of Troya abrasada.

^{Note 13 in page 246} From sources readily at hand concerning Vallejo I do not find that the simple fact that this autor's company acted the play aids in the solution of the problem of authorship; but see the discussion of the relation between the MS and the edition of 1633 further on in this article.

^{Note 14 in page 246} In this connection an attempt must be made to explain the irregular structure of the volume and to formulate an opinion as to the reliability of the above mentioned statement in the anonymous dedicatoria. The table of contents printed on the same page as the dedicatoria is ample proof that in the Barcelona copy of the volume (notice that La Barrera's list, taken from Fajardo, gives an entirely different order) the plays are in the order in which it was finally decided by the printer, with or without the consent of the compiler (i.e., the anonymous signer of the dedicatoria), that the book should be issued. From a second and recent examination of the volume, however, I am convinced that the ninth play, El Médico de su honra, with its folios numbered 1-20 and its signatures A,B, C, was set up by the printer to be the first in the collection, followed by the present nos. 3-8 (fols. 21-146, signatures D-X). Why this play, constituting like each of nos. 3-8 a perfect desglosada, was transferred to its present position as no. 9, is more than I can explain. Its place at the beginning of the volume was taken by two others, Por la puente, Juana and Celoscon celos se curan, both probably printed by the same printer as nos. 3-9, or by one with whom he had some business connection, since they have the same uncommon head-band as nos. 4 and 9; neither of the two chosen could be inserted alone for the reason that they could not be separated, since the second begins on the verso of the folio on whose recto the first ends. (Did these two plays form the beginning of a parte never finished?) The last three plays in the volume are sueltas (taken from an existing stock or specially printed for this purpose?), the first two of them having the same head-band as nos. 1, 2, 4 and 9. (In the original plan it may have been the intention, inspired by the need to economize, to complete the volume with five such sueltas. Possibly the other two originally selected tentatively for this purpose were La madrasta más honrada and Los novios de Hornachuelos, listed by Schack as the second and third plays of the mysterious “tomo 133,” in which case the Osuna volume would seem to be, not a fragment of a once complete copy of the Parte XXVII “extravagante,” but material for a copy never completed.) In view, then, of the irregularities which characterize nos. 1, 2, 10, 11 and 12, it seems permissible to doubt whether they are part of the material gathered by the compiler in manuscript. But for that part of the volume which has consecutive signatures, i.e., nos. 9 plus 3-8 (La selva confusa being one of them), the statement quoted above from the dedicatoria regarding the source of the texts may well be true, to the best of the compiler's knowledge; and it may be well to take this into consideration in connection with the question of ultimate authorship of El sastre del Campillo, up to now unanimously accorded to Belmonte, as well as of Allá darás, rayo and Julián Romero, both hitherto considered doubtful Lope plays.

It may not be amiss, while we are on the subject of this Parle XXVII, to draw attention to the fact that Fajardo, on whose sole authority the volume was mentioned by bibliographers for so long a time and with so much reserve, has been found to be right in vouching for this particular “extravagante,” save that his list of contents as reproduced by La Barrera and others does not correspond exactly with what we find in the perfect Barcelona copy. Since we now know that he was right in this case, there seems to be no longer any reason for doubting that the other two “partes extravagantes” for which he vouches, namely Parte XXVI and Parte XXVIII, also existed in his time. The existence of the Parte XXVI is attested, as we know, by Lope himself (cf. La Barrera, p. 682): it was probably a mere “tomo colecticio,” for which a title-page must have been printed more than once, since the copy which Fajardo recorded bore the date 1645, and hence was not the “edition” known to Lope. (Cf. the case of the Parle treinta de Comedias famosas de varios Autores, which was issued at least four times, as follows: Saragossa, 1636; Saragossa, 1638 [copy in the Mazarine Library, Paris]; Seville, 1638; Saragossa, 1639; there is reason to believe that in the last three of these “editions” only the title-page was new [cf. Restori, op. cit., p. 56, note 20].) For the Parte XXVIII (1639), it looks as if some Saragossa bookseller took over a certain number of copies of the Parte veinte y ocho de Comedias de varios Autores, Huesca, 1634 (on the irregular structure of this volume, see Restori, op. cit., pp. 15-17; also, RR, XIX, pp. 254-255) and replaced most of the non-Lope plays which it contained with Lope sueltas, giving them a new title-page but using the same serial number. (Notice that on the title-page of the Huesca volume the name of Lope de Vega does not appear. Perhaps the success of four recent volumes, namely Doze Com. de L. de V. C. [y otros ant.], Guesca, 1634, and the three most recent regular partes of Lope's series [XXI, 1635; XXII, 1635; XXIII, 1638], as well as some of the other regular and irregular Lope partes which preceded these, on the title-pages of all of which his magic name was prominent, seemed to the Saragossa bookseller a sufficient reason for converting the Parte XXVIII de varios into a Parte XXVIII de Lope de Vega, in order to do which he was obliged to change not only the title-page of the volume, but also the first pages of some of the plays themselves.) The last existing copies of these two “partes extravagantes” may have been broken up shortly after Fajardo's time, sacrificed to the demand for sueltas. In conclusion, it may be pointed out that the two Osuna volumes 131 and 132 are of no more importance in connection with the Parte XXVI “extravagante” than any other collection, bound or not, containing one or more sueltas of the plays mentioned by Fajardo as composing this parte, so that there is no reason for listing the entire contents of the two Osuna volumes in a bibliography of Lope de Vega.

^{Note 15 in page 248} Perhaps no name of author was even associated with the performance. Was the select audience at all interested in who wrote the play?

^{Note 16 in page 248} In another study I propose to show that this play is probably not an original creation of Calderon's.

^{Note 17 in page 248} Of course it may be objected that he could have written it some time before the other play; but such a supposition would necessitate admitting that he wrote La selva confusa at a still earlier age.—Significant as showing that Calderón can hardly have attracted much attention at court as a playwright up to within a few months at least before June, 1623, are the facts on which Cotarelo (BRAE, IX, p. 37, continued footnote) bases the following observation: “Entre los meses de octubre de 1622 y febrero de 1623 se representaron, según otros datos, en Palacio 45 comedias de varios autores que entonces escribían, como Lope, Tirso, Vélez, Alarcón, etc., y ninguna de ellas pertenece a Calderón, cosa extraña si ya entonces fuese don Pedro algo más que un modesto principiante.”

^{Note 18 in page 248} This list is in the “prólogo” to the Obelisco funebre, etc., of Gaspar Agustīn de Lara, Madrid, 1684 (cf. BAE, VII [Comedias . . . . de Calderon . . . ., I], pp. xli-xlii).

^{Note 19 in page 249} The first is in the “prólogo” to the Quarta parte, Madrid, 1672 and 1674 (cf. BRAE, V, p. 543); the second is in the “advertencia” of the Verdadera quinta parte, Madrid, 1682 (cf. BAE, VII, pp. xxv-xxvi).

^{Note 20 in page 249} Vera Tassis, in the last mentioned list, also includes eight plays “manuscritas” attributed to Calderon; but there is so little chance of Vera Tassis's having made an investigation among manuscripts (he does not claim to have searched for anything but printed plays) that the absence of La selva confusa from this list of only eight titles can not be taken to mean that he knew of the manuscript of this play and let it stand as of Calderon, all the more so as he did not announce any intention of printing it in one of the Partes VI to X.

^{Note 21 in page 249} Ed. Northup, vv. 503, 554, [766], [803], 994, 1017, 1122, 1177, 1186, 1300, 1305, [1328], 1544, 1593; also. later, [cclxxxvii]. (The bracketed figures correspond to lines which are not in the edition of 1633.)

^{Note 22 in page 249} The term in the singular is certainly not accurately applied as a designation for the scene of action of this the greater part of the play. The Diccionario de autoridades defines “selva” as: “Lugar lleno de árboles, malezas, y matas, que le hacen naturalmente frondoso.” But the scene of action here is a spot where Fadrique as a “jardinero” is working (“labrar la tierra”) in the Duke of Mantua's gardens “murados de agua y jazmines,” situated “entre el Rin y el Po” (cf. ed. Northup, vv. 628-635); but the woods (“selvas”) have been cleared off to make way for a “poblado” (cf. vv. 1367-1368). The spot, close to or part of an “aldea” (vv. 1369, 1457, etc.), is fairly accurately described in vv. 368-375, xxxiv-xlv, 376-379, and there is no mention of trees or thickets here.

^{Note 23 in page 249} Vv. 2360, 2512, 2586, 2658, 2790, 2792, dlxv, dciv (not in the edition of 1633), 2933, 3025. It occurs also in v. 385 (not in the edition of 1633), but here the original reading in the MS was “aldea” (cf. Northup's note, p. 228). For further discussion of this word “selva” in the text see footnote 54. (In v. 291, for “la selba” the edition of 1633 has “las selbas,” this latter being the correct use of the word in this opening scene.)

^{Note 24 in page 250} There is just as great a probability that La selva confusa is El desdichado as there is, for example, that Roma abrasada is the Nerón cruel mentioned by Lope in the Peregrino.—Northup himself (p. 204, footnote on v. 503) seems to have had a suspicion that La selva confusa might be El desdichado.

^{Note 25 in page 250} This has been noted particularly by Stiefel (Zeitschrift, XXXVI, p. 460), who makes use of it very briefly among his arguments in claiming that La Española de Florenica was written by Lope.

^{Note 26 in page 250} Schack's words are: “. . . . das Stück ist durchaus nicht in Lope's Styl geschrieben, wie schon die Anfangsverse zeigen,” etc. The translation quoted above is Rennert's (Bibliography, p. 241). Notice that by “durchaus” Schack by no means implied that he had formed his impression of the style from reading the play “throughout.” Notice also the significant adverb “schon.”

^{Note 27 in page 250} Rennert (loc. cit.) has already corrected Northup's impression that Schack copied these seven verses from the printed edition which he had once seen. Cf. footnote 55 of present article.

^{Note 28 in page 250} I. e., ed. Northup, vv. 6-11 and 22-v.

^{Note 29 in page 250} That he did not read the manuscript through seems to be further implied by his remark: “. . . . dennoch wird es sich der Mühe lohnen, das Manuscript näher zu untersuchen.”

^{Note 30 in page 251} This is the case with such familiar plays as La vida es sueño, El mayor monstruo los celos (after the music), etc.

^{Note 31 in page 251} 1. La amistad pagada, 23 verses (prior to 1604, since it was printed in the Parle I); 2. Los primeros mártires del Japón, 26 verses (assigned to 1617); 3. Amar, servir y esperar, 70 verses (“La fecha se halla, probablemente, entre 1618-1623.”—Rennert-Castro); 4. La vida de San Pedro Nolasco, 57 verses (1629); 5. La carbonera, 80 verses (“Es comedia de la vejez de Lope.”—Rennert-Castro). Of these, Schack does not mention nos. 2, 3 and 4 in his Geschichte; he had examined no. 1 closely enough to find out what its subject matter is (op. cit., II, p. 265); and he seems to have read no. 5 carefully, since he analyzed it (II, pp. 311-313), but could well have forgotten all about its versification long before writing his note on La selva confusa for the Nachträge.

^{Note 32 in page 252} Fadrique leaps from a cliff (“peña”) before our very eyes (vv. 314-317), having already informed us that below flows the river Po (vv. 298-301). The lie is invented by Carlos in vv. 325-332. The marriage arrangements are first told to us by Flora (vv. 428-439) and are again mentioned later on in a conversation between Filipo and Carlos (vv 805-806, 829-840).

^{Note 33 in page 252} It begins well down the page on fol. 13 r. and ends about the middle of fol. 14 r.

^{Note 34 in page 253} Cotarelo calls attention also to the absence of the words “Jhs. Maria Joseph” usually found in Calderon autographs. Did these three holy names constitute a sort of pious invocation used by the poet only when he sought inspiration for composing, their absence from the present MS thus indicating that he sat down merely to copy?—Northup further remarks that there is no reparto; but this was a matter for the autor to attend to.

^{Note 35 in page 253} A sufficient reason for believing that the MS is not a borrador is the one advanced by Cotarelo himself, as noted above, while on the other hand evidence that it is the product, in part at least, of the purely mechanical process of transferring a text from one MS to another is furnished by some of Northup's notes at the end of the first act. For example: 1. After v. 547 the scribe (as we shall call him for the present) wrote vv. 538-539 for the second time, then crossed them out on again finding his place in his original. 2. The first part of v. 616 is spoken by one character and the second by another. Having read these two halves as a single verse, the scribe first wrote them all on one line, but immediately after, presumably upon looking back at his original, he struck out the second half and wrote it on a new line. If he had been composing, he would have stopped naturally after the first half and passed to the next line immediately to write the name of the new speaker. 3. In copying v. 970 (with its verb “bais”) which he had probably first read in conjunction with the preceding verse, his eye at this moment caught the “pues no bais” six lines below, but he got no further than “pues” when he perceived his error. 4. In v. 1082 he made the same mechanical error as in v. 616.—The only evidence of groping during the process of composition is his writing and rewriting of v. 62 (not in ed. of 1633), vv. 790-795 (in the inserted scene), and v. 2659 (which he seems to have wanted to change, although he finally came back to his original, for the corresponding line of which the ed. of 1633 has “muestran” in place of “mostrar”). The numerous substitutions of one reading for another are an entirely different matter, as will be seen later.

^{Note 36 in page 255} Besides, I am working with notes taken in Barcelona in the summer of 1927 with N and P before me, and at the present time I could not feel sure that my table of variants would be complete.

^{Note 37 in page 255} Rather than a superfluous line this is perhaps the first verse of a pair which was complete in the original, this first verse having been accidentally omitted by Calderon, as well as by the printer of P.

^{Note 38 in page 255} Following is a complete table of lines and passages missing from P, by blocks as they occur in N: 6-11, 22-81, i-v, x-xvii, xviii-xxv, xxvi-xxix, 189-190, 262-265, 332, xxx-xxxni, 360-363, 372-375, xxxiv-xlv, 376-379, 384-387, xlvi-xlix, 388-395, 398, l-liii, liv-lxii, 596-603, 696-703, 712-715, 736-804, 869-880, lxvii-lxxviii, lxxix-lxxxviii, lxxxix-xcii, 1100-1103, xciii-xcvi, 1104-1107, xcvii-c, 1116-1119, ci-cviii, cix-cxviii, cxix-cxxviii, cxxix-cxxxviii, cxxix-cxlviii, 1285, 1322-1329, cxlix-clii, clxi-clxviii, 1342-1345, 1394-1401, 1430-1433, clxxvii-cxcii, cxciii-cxcvii, cxcviii-cci, 1634-1641, 1646-1649, ccx-ccxxi, ccxxii-ccxxv, ccxxvi-ccxxix, ccxxx-ccliii, ccliv-cclix, cclx-cclxni, cclxtv-cclxvn, cclxvui-cclxxvn, cclxxvni-cclxxxi, cclxxxii-cclxxxix, ccxc-ccxciii, ccxcvi-ccxcvii, 1936-1939, 1946-1947, ccxcviii-ccxcix, ccc-cccv, 2023, cccxi-cccxiv, cccxv-cccxviii, cccxix-cccxxxiv, cccxxxv-cccxliii, cccxliv-cccli, cdii-cdv, cdxiv-cdxxx, 2503-2506, cdxxxi-cdxxxiv, 2595-2598, cdxxxv-cdlii, cdliii-cdlviii, cdlix-cdlxiv, cdlxxi-cdlxxvi, cdlxxvii-cdxciv, cdxcv-dxviii, dxix-dxxiv, dxxv-dxxxvi, dxlv-dlxii, dlxvii-dciv, 2908-2911, 2975-2977.

^{Note 39 in page 256} Two lines at the beginning, four after v. 992, and ten after v. 1091.

^{Note 40 in page 256} This line is: de hablar, y de discurrir.“

^{Note 41 in page 256} This count of 17 distinct passages can not be checked up from the table in footnote 38 above for the reason that between vv. ccli and cdii there are five, and between cdv and cdxiv there are two, separate passages.

^{Note 42 in page 256} So far as I am aware, there are no traces of another series of markings for excisions, nor of the erasure of the above mentioned new lines furnished by P, in accordance with all of which P could be a reproduction of M.

^{Note 43 in page 257} If this had been the case, what chance is there that the MS, after a trip to Barcelona, would have survived?—Notice that, in consequence of the conclusion just stated, the theory which is being followed here is still safe.

^{Note 44 in page 258} It could be easily shown, from a study of the context in each case, that the following passages wanting in P are likewise essential to the full elaboration of the thought: 596-603; 712-715. But cases of exactly the opposite of this phenomenon will be noted in connection with remarks on Calderon's version of the second act.

^{Note 45 in page 258} And again the theory, insofar as the source of M is concerned, is safe.

^{Note 46 in page 258} Furthermore, M does not derive indirectly from P, for the same reason that the former does not derive directly from the latter. On the other hand, I see no basis for denying that P could have derived indirectly from M, nor is it necessary to do so, since even if it could be proved that such is the case, the essential part of the theory which I have adopted would not be injured.

^{Note 47 in page 259} From what has already been noted concerning the continuity of composition which is manifest in Calderon's transcript or version, it would seem as if such a MS of Vallejo's must have been something better than an actor's copy. It may have been Lope's original.

^{Note 48 in page 259} Incidentally, has it ever been discovered whether or not the printing of the formula: “Representola . . . .” along with the title in a given edition (practice generally discontinued within a few years after the date of this volume) was a condition imposed upon the printer by the autor from whom the MS of the play was purchased?

^{Note 49 in page 259} There is of course no proof for all this. The MS from which P was printed could have been, on the contrary, another copy, direct or indirect, of Lope's original, but one to which, in any case, the above mentioned additions by Calderon had been transferred.

^{Note 50 in page 260} Such appear to be, for example, the corrections noted by Northup in vv. 385 (change made in consequence of new title), 659, 694, 1480, 2321 (“nadie” made the verse incorrect), etc.

^{Note 51 in page 262} The point could be further illustrated by a study of vv. 111, 166, 205 (“ynocencia” [ = “stupidity:” cf. v. 1483] makes nonsense here, while “obediencia,” although used as an assonant only four lines back, is plainly correct; but did Calderon himself make the change?), 337, 426-427, 560 (?), 1380, etc.

^{Note 52 in page 262} Hence the absence from P of such long passages as vv. 22-v, as well as the new scene in the first act, etc. The last 140 lines could be written on blank pages at the end of the other MS or MSS, or on new pages attached to them. This concluding passage, in fact, may be the only lines that ever got into another copy of the play: I would not venture to affirm that anything else in P, for example, is Calderon's.

^{Note 53 in page 263} If such a spot away from Madrid can be located, the dating of Lope's play will probably be a simple matter.

^{Note 54 in page 263} “Confusa” expresses the very essence of a good enredo: in La Barrera's general list of titles and sub-titles there are 16 with “confuso” or “confusión.” —It has already been pointed out (footnote 23) that in v. 385 “selba” has been substituted for “aldea” (probably by Calderon, as an afterthought): the same substitution may have been made, but during the copying, in v. 2360. (For obvious reasons, this could not be done, without a rewriting of the passages, in vv. 1369 and 1457.) In vv. 2658 and 2790 “selba” can hardly have been the reading of the original, since the context shows that the confusion which is bothering the Duke is in the dramatic situation, not in the scene of action. The occurrence of the word in the plural offers no serious difficulty, since we know, in fact, from vv. 1364-1369 that a portion of forest land has been cleared away and converted into a retreat for the Duke and his daughter, so that “selvas” could apply to the surrounding region. (And notice that the original MS reading of v. 1367 was: “a este campo . . . .” for “a este efeto . . . . .,” but it is difficult to reconcile the first reading grammatically with the rest of the sentence.) Still and all, the reading “selbas” is not always satisfactory. In v. 1979 the causal “que” makes nonsense: I suspect that the original had something like “a esta ribera, y sabiendo . . . .” (cf. v. 1639). In P v. 2513 has “yerbas” for “salbas” (?): evidently the entire passage, with its “selbas” in v. 2512, is not as it was in the original. In v. 2792, where a lady masquerading as a “villana” is being upbraided, “aquesta aldea” would seem better than “aquestas selbas.” V. dlxv belongs to a quatrain which appears to have been added by Calderon, for if this is connected with the preceding quatrain, we have a mixed figure (bird, Daphne); but even if the mixed figure is a slip by Lope, we should expect some such word as “matas”' rather than “selbas” in connection with the legend of Daphne. Vv. 2933 and 3025 will be disposed of later.

^{Note 55 in page 264} Hence a proof that Schack copied the opening lines from the MS, not from P.

^{Note 56 in page 264} Evidence that this passage has been inserted lies in the fact that Filipo's terminating remark: “Es en efeto ymagen de la guerra,” is not followed by a development of the parallel to which he points in these words.

^{Note 57 in page 265} Cf. “Cubrilde. (Ponenle una capa.)” in v. 462, which is not in P, of course. Thus prudery in the young Calderon appears to have dictated the rewriting of this scene! Or did the exigences of a court representation have something to do with it?

^{Note 58 in page 265} But these lines were later reassigned to Flora in M.

^{Note 59 in page 266} For “y tu patria” the “corrected” MS has “sentado” (!).

^{Note 60 in page 267} This exclamation retained by Calderon becomes contradictory, since in his version Fadrique has been represented as being saved by fishermen, not through his own efforts.

^{Note 61 in page 267} In this connection it would be interesting to consult the MS in order to see whether this name, and also the “pescador primero” and “pescador segundo” at the end of the list, have been squeezed into place. If this is the case, it practically proves that the changes requiring the addition of these names to the list was an afterthought of Calderon's, and not part of a preconceived plan of completely recasting the play.

^{Note 62 in page 269} Jacinta's words: “El (sic in P) secreto te prometo.” in v. 885 seem to contain a prompt reply to Marcial's expression of fear in vv. 881-883 and hence should follow immediately.

^{Note 63 in page 269} The lines which follow this décima indicate that Fadrique is completely resigned to his lowly station for the present and is not looking for a change of “fortuna” as expressed in the décima.

^{Note 64 in page 269} These are: vv. cxcviii-ccL, ccii-ccix, cclxiv-cclxvii, cclxxviii-cclxxxi, cccxv-cccxviii, cccxix-cccxxxiv.

^{Note 65 in page 269} These passages are: 1. clxix-clxxvi (Vv. 1422-1423 seem to contain a prompt reply to the question in v. 1421, and hence there should be no intervening lines). 2. clxxvii-cxcii (Filipo's manifestation of a lingering affection for Jacinta seems to be a part of Calderon's preparation for the end of the play as he conceived it). 3. cxciii-cxcvi (The “si es Flora” of v. 1454 is the logical complement of v. 1453. The short verse cxcvii shows up the patch). 4. ccx-ccxxi (The passage has one unintelligible line, v. ccxiv, and one stupid ripio, v. ccxx. In vv. 1675-1676 Calderon seems to have introduced changes in order to take the speech of which they form a part away from Oton and give it to Celia: in P we find: “porque antes de agora dize/que no lo era, y contradize . . . .,” the two verbs in the present tense being probably errors for “dijo” and “contradijo”). 5. ccxxii-ccxxv (Among other defects the passage has an incorrect verse, namely the third). 6. ccxxvi-ccxxix (The first and fourth verses of this redondilla do not rhyme). 7. ccxxx-ccliii (After beginning a relación with the more or less conventional “Sabrás, pues, . . . .” and getting well on in his story, Fadrique repeats this formula in v. 1762!). 8. ccliv-cclix (The passage begins with a ripio.—The “Yo lo perdono.” of v. cclvii was taken from v. 1796, which in turn, in order that a repetition of exactly the same words might be avoided, was changed from “Yo lo perdono, Fadrique.” to “Pero yo te lo perdono.” This change, by the way, was foolish, since Fadrique two lines further on plainly indicates that the Duke has just called him by his name). 9. cclx-ccxxiii (This stifled laughter is decidedly incongruous). 10. cclxviii-cclxxvii (Trivial). 11. cclxxxii-cclxxxix (The account of what Fadrique did after his rescue from the river Po is superfluous.—The “no” of the last line of the passage, repeated in the next line of the text, makes awkward patching.). To these should be added, I believe, although they are in P, vv. 1750-1753, for it is difficult to imagine that Lope would extend the long aparte of which these four lines are the conclusion from his redondillas over into this romance.

^{Note 66 in page 270} In connection with this second act, what is the significance of the word “Lamas” which, as Northup notes, Calderon scribbled several times in the margin of this part of his MS? Is it for “la amas?” Was Calderon distracted during these days?

^{Note 67 in page 270} These passages are: 1. dxlvv-dlxii (V. 2801, containing a snappish reply to Flora's “los hombres buscas” in v. 2798, should follow v. 2800 immediately). 2. dlxiii-dlxvi (This quatrain and the one which precedes it, when taken together, present a mixed metaphor, as already noted in footnote 54). 3. dlxvii-dciv (The “Basta.” [followed by an aparte] of v. 2819, equivalent to: “Quit your nonsense !”, shows that in the original Filipo did not succeed in breaking away from the Duke [vv. 2811-2818] to go in pursuit of Jacinta: hence there should be no intervening lines.—Filipo in this passage addresses trees which do not exist in the original, calling on them to stop Jacinta.—The passage ends in a ripio which has no raison d'être except the new title of the play.).

^{Note 68 in page 270} Vv. 2139 and 2483 are ripios supplied by the “corrector” after excisions. V. 2522 has been substituted for “no como ella dice hermana,” which connected up perfectly with the next line before the excision was made. A satisfactory reading for vv. 2659-2660 is furnished by P: “Qué bien muestran las flores/que a Flora deven sus matices diosa/Venus,” etc. V. cdlxxxix is too long as it stands, and moreover should not have “señor” in it, being addressed to Flora: it could be emended, for example, as follows: “Déjeme vuexcelencia.” V. dxi is likewise too long, but can be emended by the omission of “y esta.”

^{Note 69 in page 271} These are all in P also, except vv. 2908-2911 and 2975-2977. It has already been explained (footnote 52) how this part of M could get into P.

^{Note 70 in page 271} From the fact that this is a romance in “u-a” it might be argued that Lope's title must also have been “La selva confusa.” Not necessarily. Lope did not use romances in “u-a” in the final scenes of Amar por burla, En los indicios la culpa, Fuente Ovejuna and El premio de la hermosura: I have not had access to La prisión sin culpa.

^{Note 71 in page 272} This is not necessarily of any importance in connection with the honor question in the later work of Calderon.

^{Note 72 in page 272} The nearest approach to it that I can find are a few cases of versos sueltos (El verdadero amante, El esclavo fingido, Los esclavos libres). Nor was it used by Calderon himself except in Amor, honor y poder; but cf. footnote 20.—Schack may well have glanced over these closing lines in the MS, finding in them a confirmation of his impression that the entire play is Calderon's.

^{Note 73 in page 272} It was composed not later than 1603, year in which the first Peregrino was ready for the press. That the play was written a few years earlier than this date may be partly inferred from the fact that it has no gracioso, for Marcial is only a tosco or villano. Moreover, the sonnet in the third act (after v. 2288) has the appearance of belonging to the Belisa series of poetic works, voicing as it seems to do Lope's remorse over one of his frequent lapses into infidelity to her. (“Jacinta” at the beginning of the second line is replaceable by “Belisa,” the first syllable of which, according to Northup's note, Calderon actually wrote before correcting himself: could the sonnet have been inserted with this name into Lope's MS of El desdichado?) If this is the case, then the play was not written later than 1595; but this whole matter calls for a closer investigation. Incidentally, in P the text of the sonnet, but not the punctuation, is the same as in N, except that it has “juzgo” for “vyo” in the 11th line and “muera, o viva” for “viba o muera” in the 13th.