^{page 249 note 1} My thanks are due to Professor G. B. A. Fletcher and to Mr. C. Garton for reading and correcting an earlier draft of this paper, and to Dr. J. C. Coulson, Lecturer in Zoology at Durham University, for making some very valuable ornithological comments.

^{page 249 note 2} e.g. the description by one of Aristotle's pupils the of in H.A. 9. 14. 6i6^a14, to which Pliny (or rather, his source) adds the diagnostic feature of the white feathers close to the neck, H.N. 10. 47. Cf. Thompson, D'Arcy W., A Glossary of Greek Birds² (London, 1936), pp.46 ff.;Google ScholarFowler, W. Warde, ‘The Birds of Virgil’, in A year with the Birds³, (1889), pp. 239 ff.Google Scholar

^{page 249 note 3} Cf. Charles Darwin's remark upon reading Ogle's Aristotle (op. cit. in the following note): ‘Linnaeus and Cuvier have been my two gods, though in very different ways, but they were mere schoolboys to old Aristotle’ (Life and Letters iii. 252).

^{page 249 note 4} Cf. Singer, C., Greek Biology and Greek Medicine (Oxford, 1922), p. 24;Google ScholarOgle, W., introduction to his Aristotle on the Parts of Animals (London, 1882).Google Scholar

^{page 249 note 5} In H.A. 9.49B. 632^b15, a pupil of Aristotle assumes that the blackbird with brown plumage is a seasonal variation of the blackbird with black plumage; in this error he was faithfully followed by Aelian, , N.A. 12. 28,Google Scholar and Pliny, , H.N. 10. 28Google Scholar; Dionysius, Av. 1. 27, explains the two forms as separate species. It may surprise us that the Peripatetic school was unable to understand sexual differentiation in so common a bird, yet the error is easily explained as confusion between the brown adult female and the not dissimilar juvenile male, which does change in its first year from brown to black plumage.

^{page 250 note 1} Fowler, W. Warde, op. cit., pp. 244 f.Google Scholar

^{page 250 note 2} See Kroll, W., R.-E., s.v. Plinius d. Äl- tere, 309 f.,Google Scholar and cf. Dannemann, F., Plinius und seine Naturgeschichte (Jena, 1921), pp. 31 ff.Google Scholar

^{page 250 note 3} Boraston, J. Maclair, J.H.S. xxxi (1911), 216 ff.Google Scholar, is correct in his general appraisal of the Homeric poems, though often wrong in his identifications.

^{page 250 note 4} As Fowler, W. Warde, op. cit., points out; contrast his p. 215,Google Scholar ‘there is hardly to be found, in the whole of Virgil's poems, a single allusion to the habits of birds … which is untrue to fact’, with his admission that Virgil shares the limitations of the countrymen of his own day, p. 216. See further Royds, T. F., The Beasts, Birds, and Bees of Virgil (Oxford, 1918), pp. 34 ff.Google Scholar; and Jermyn, L. A. S., Greece and Rome xx (1951), 26 ff.CrossRefGoogle Scholar

^{page 250 note 5} Thompson, D’Arcy W., C.R. xxxii (1918), 95.Google Scholar

^{page 250 note 6} Arist. H.A. 6. 1. 559^a4. Unfortunately, manuscript corruption prevents our knowing for certain what that name was.

^{page 250 note 7} Alexander of Myndos, in Athenaeus 2. 65B.

^{page 250 note 8} [Arist.] H.A. 9. 32. 618^b25.

^{page 250 note 9} Op. cit. in n. 5, 92 ff.

^{page 250 note 10} These examples are merely colloquial and regional variants. It hardly needs saying that such colloquial inexactness ought to be anathema to exact scholarship: yet how often does one come across, even today, such slipshod and inaccurate translations as, for instance, ‘mountain-ash’ or (worse still) ‘rowan’ for ornus in Latin: the manna ash (Fraxinus ornus)! This error is probably inspired in most cases by Lewis and Short's interpretation as ‘the wild mountain-ash’ (after Riddle, White-Riddle: Littleton in 1715 correctly followed Columella in saying only ‘a wild ash with broad leaves’). And errors in the translation of hyacinthus and uiola are not much fewer.

^{page 251 note 1} See below, part II of this paper.

^{page 251 note 2} Bannerman, D. A. and Lodge, C. E., Birds of the British Isles, xi (1962), pp. 38 ff.Google Scholar

^{page 251 note 3} Fisher, James, The Fulmar (London. 1952), chapters 7–11;Google ScholarFisher, and Lockley, R. M., Sea Birds (London, 1954), p. 105.Google Scholar

^{page 251 note 4} Fisher, , British Birds xlvi (1953), 153 ff.;Google ScholarStresemann, E. and Nowak, E., J. f. Ornithologie xcix (1958), no. 3.Google Scholar Though Fishei accepts Mayr's suggestion that the expansion was due to a genetical cause, there are other equally plausible explanations of this change. Dr. Coulson writes in a letter to me (15 October 1963), ‘The short answer is that biologists do not know why the spread has occurred’.

^{page 251 note 5} The International Commission of Zoological Nomenclature, with a view to settling a long-standing dispute, recently decided (Opinion 401, published 1956) to use Gavia (after Forster, 1788) in place of Colymbus (Linnaeus, 1758) as the generic name for the divers, a group of birds rarely seen in most of the Mediterranean area, and unlikely to have been called by this name in the ancient world. Disparity between the meaning attached to words such as gauia by modern biologists and the ancient usage is nowadays too common to arouse comment, and perhaps to some extent this disparity is inevitable. Yet the objections to the modern use of Gavia as the generic name for the divers are not merely pedantic. This is a case where the ancient use of the word can be established with reasonable certainty as different from the modern scientific one. Secondly, gauia forms the etymological source for the noun used for ‘gull’ in many of the Romance languages today: e.g. gabbiano (It.), gaivão (Port.), gaviota (Sp.). The word gavia itself was still used in fairly recent times as a dialect word for ‘gull’ in Castile (Coro-minas, Diccionario critico etimologico de la lengua castellana, s.v. gaviota), and gavina occurs even today around Naples and in Catalan; cf. F. Cetti's note on the use of the latter form in eighteenth-century Sardinia (Gli uccelli di Sardegna[Sassari, 1776], p. 291). And modern biologists cannot in this case excuse themselves with a plea of ignorance, since their predecessors in the eighteenth and nineteenth centuries correctly applied the word Gavia to the gull (e.g., Gmelin, S. G., Reise durch Russland zur Untersuchung der drey Natur-Reiche [St. Petersburg, 1770] i. 152)Google Scholar, as the report of the Commission itself acknowledges.

^{page 252 note 1} The Thesaurus cites additionally three occurrences in writers of the fifth century A.D. Eucherius, who was Bishop of Lyons c. 440, says that in the pentateuch ‘larus gauiam significat’; (Instructiones ii. 157. 15: cf. Gloss. Lat. i. 262b, s.v. gauiam); this suggests that the Graecism larus was in popular use at the time of the Vulgate, but whether it was a late importation or whether it flourished unrecorded in the sermo uulgaris of classical times we do not know. Polemius Silvius {fl. c. 448) mentions gabia in his calendar (Laterculus, iii: Mon. Germ. Hist. ix. 543. 20); on the betacism, so typical of the period, see Parodi, E. G., Romania, xxvii (1898), 177 ff.;Google ScholarTerracini, B. A., Arch. Glott. It. xxvii (1935), 133 ff.;Google Scholar and Vossler-Schmeck, , Einfuhrung ins Vulgärlatein (Munich, 1954), p. 101.Google Scholar Finally Caelius Aurelianus mentions that cerebrum gauiae was wrongly prescribed by some physicians in cases of epilepsy (Chron. 1. 4. 119; cf. also i. 4. 129). The work of Soranus that Caelius Aurelianus was translating is now lost, so that we do not know the Greek name of the bird translated as gauia; Galen's denial of any efficacy as a remedy to the boiled or dried stomach of the (Kühn, xii. 336) may or may not be relevant in this connexion.

^{page 252 note 2} See p. 250.

^{page 252 note 3} Clearly the doubts which have perhaps justifiably been cast over the authenticity of this passage of Aristotle in modern times (see Thompson, D'A., Oxford Translation of:he History of Animals [London, 1910]Google Scholar, together with Aubert and Wimmer's edition [Leipzig, 1868], ad loc.) would not have Deen shared by the elder Pliny himself.

^{page 253 note 1} On the winter ‘hiding’ see Lones, T. E., Aristotle's Researches in Natural Science (London, 1912), p. 244.Google Scholar

^{page 253 note 2} Cf. Kroll, , op. cit., p. 310.Google Scholar Instances like the present one refute the theory of Detlefsen (Hermes, xxxvi [1901], 18)Google Scholar that Pliny followed Aristotle directly without any intermediary source.

^{page 253 note 3} Bannerman, and Lodge, , op. cit. viii. 11.Google Scholar This fact was first observed in more modern times by the Northumbrian naturalist William Turner, who wrote in 1544 (Allium praecipuarum … historia [Cologne, 1544], p. 90,Google Scholar = Turner on Birds, edited by Evans, A. H. [Cambridge, 1903], pp. 110f.)Google Scholar, ‘Et ego utrumque obseruaui, nam in rupibus marinis iuxta hostiũ Tinae fiuuij (the Tyne) mergos nidulantes uidi, & in Northfolcia cum ardeis in excelsis arboribus’. On the basis of this observation Turner identified Pliny's mergus here as the cormorant, being followed by Caius, J., De rariorum animalium alque stirpium historia (1570), fol. 22Google Scholar (printed in the Cambridge reprint of Turner), and by Philemon Hol land in his translation of Pliny (1601); Conrad Gesner, (Historiae Animalium, post humous revised 2nd edition of 1617 published at Frankfurt, iii. 108 fff.), quotes Turner's report but himself uses mergus rather differently: see below, p. 258, n. 3. Cf. also Thompson, D'A., C.R. xxxii (1918), 95, n. 1.Google Scholar Two other palaeoarctic marine and aquatic species nest at least rarely in trees: the mallard and (very occasionally) the shelduck, but neither of these is likely to have been termed mergus by a Roman. Keller, O., Die antike Tierwelt ii (Leipzig, 1913), 244 n.Google Scholar, discussing this passage of Pliny, refers to an alleged observation by the nineteenth-century naturalist Audubon that herring gulls ‘oft Baumwipfelzu ihrer Nistung auswählen’, and concludes from this that Pliny's tree-nesting mergus was a gull. But Keller's note, which has unfortunately influenced Steier's article in R.-E., s.v. Mowe, 2415, is very misleading. The gull that Audubon was referring to was not a Mediterranean variety but the American herring gull (Larus argentatus smithsonianus): see Audubon's Labrador Journal for 7 June 1833; and such tree-nesting is exceptional even for this American variety (Coues, , Key to North American Birds [Boston, 1887], p. 744)Google Scholar, being resorted to only when the normal ground nesting sites are disturbed (Bent, A. C., Life Histories of North American Gulls and Terns [U.S. National Museum Bulletin 113], pp. 103 ff.).Google Scholar As to the European herring gull, only one attempt by any member of this species to nest in a tree has ever been noted in Britain, and this was not in the end successful (Bannerman, and Lodge, , op. cit., xi. 293 f.).Google Scholar

^{page 253 note 4} Cf. Makatsch, W., Die Vogelwelt Mace-doniens (Leipzig, 1950), pp. 341 ff.Google Scholar

^{page 254 note 1} In this country cormorants and shags breed from the beginning of March, and correspondingly earlier in more southerly latitudes: a century ago Lindermayer (Die Vägel Griechenlands [Passau, 1860], pp. 166 f.Google Scholar) noted that in Greece both cormorants and shags were already breeding in January. Thus Pliny's observation that they nest at the beginning of spring is proved reasonably accurate: to the Romans spring began on the 23rd day after the sun's entry into Aquarius: i.e., February 7 (Varro, R.R. 1. 28), 8 (Pliny, H.N. 2. 122), or 9 (Ovid, Fast. 2. 150); cf. Unger, G. F., Jahrb. f. Philol. cxli (1890), 473 ff.Google Scholar

^{page 254 note 2} On the identification of see especially Thompson, D'A., Glossary, 1895 edition, pp. 17 f.Google Scholar (citing earlier literature and identifications, and suggesting tentatively that the term is applied to large gulls, such as black-backs and the herring gull: cf. his note in Companion to Greek Studies,³ [1916], p. 44), and Keller, O., op. cit., pp. 242ff.Google Scholar (suggesting that = any gull or similar sea bird, such as shearwater, tern, or petrel: he is followed by Steier, op. cit., 2414). In his later work, Thompson seems to me to stress the shearwater equation too strongly (C.R. xxxii [1918], 94 f.Google Scholar; Glossary, 1936 edition, pp. 27 ff.), though he has persuaded L.S.J., s.v. Cf. also p. 257, n. 3 of this paper.3 From this point on, to save a certain amount of awkward repetition, I shall use the term ‘cormorant’ for both the cormorant proper, Phalacrocorax carbo, and the shag, Phalacrocorax aristotelis (formerly Ph. graculus).

^{page 254 note 3} From this point on, to save a certain amount of awkward repetition, I shall use the term ‘cormorant’ for both the cor-morant proper, Phalacrocorax carbo, and shag, Phalacrocorax aristotelis (formerly Ph. graculus).

^{page 254 note 4} Sundevall, C. J., Die Thierarten des Aristoteles (Stockholm, 1863), p. 158.Google Scholar

^{page 255 note 1} The beginning of summer was generally equated with the rising of the Pleiades, 15–19 May (Arat. Phain. 264 ff.; Varro, R.R. 1. 28; Ovid, Fast. 5. 600), though the chronology is complicated by the fact that the Greeks sometimes used to include spring as well. For a full discussion see Aubert and Wimmer, i. 187 f., and Gomme, A. W., Commentary on Thucydides iii. 699ff.Google Scholar (the Greek practices), and Frazer's note on Ovid, loc. cit. (the Roman practice).

^{page 255 note 2} See below, my discussion of the Apuleius passage. On itself see especially D'A. Thompson, Glossary, s.v. Thompson (with Sundevall, , op. cit., pp. 157 f.Google Scholar; Keller, , op. cit., p. 244Google Scholar; and Steier, , op. cit. 2414Google Scholar) recognizes that the term may also be applied to terns, citing the description in Dionysius, Av. 2. 4, of one kind of that is ‘white and short like the ’ a description which suits the tern better than any other marine bird. Thompson wisely omits from his second edition the allegation that the common tern lays about the same time as the herring gull. In fact, all the terns except the Sandwich tern begin laying some three weeks later than the herring gull.

^{page 255 note 3} Dr. M. J. Cullen, in Bannerman and Lodge, xi. 80. It is perhaps worth recording that Gesner also (op. cit., pp. 527 ff.) noted the close similarity between gulls and terns. Certainly one of the smaller species of Mediterranean gull (Lams melanocephalus) has the same basic colour pattern as the majority of the terns: black on the head, white body, greyish upper wings. The size of this gull also closely corresponds to that of several of the terns.

^{page 255 note 4} Cf. Sundevall, p. 164; Aubert and Wimmer, pp. 88 f.; Thompson, Glossary, s.v.

^{page 256 note 1} So Thompson, loc. cit., but first edition only.

^{page 256 note 2} Dr. Coulson notes that the word ‘black bird’ is sometimes used as a name for shags and cormorants by inexperienced people in this country.

^{page 256 note 3} Aelian, N.A. 5. 48, includes and in his own list of birds hostile to each other, a list which derives ultimately from Aristotle. Consequently, either the reference to as a sea bird was found by Aelian (or Aelian's source, see Wellmann, M., Hermes xxvi [1891], 481 ff.Google Scholar) in the Aristotelian passage cited, in which case any alleged interpolation in Aristotle must have predated not only Aelian but also his source, i.e., have been due to an ancient scholar working in the Mediterranean area; or alternatively, the reference was interpolated into Aristotle from Aelian, in which case Aelian and his source must have found a reference to as a sea bird in some other ancient author. This rules out of court any reference to a non-Mediterranean sea bird for , but this is as far as we can go towards its identification. Gossen's suggested equation with the avocet (Qullen und Studien z. Gesch. d. Natural, u. d. Medizin, iv [1935 324Google Scholar) is worthless.

^{page 256 note 4} See further Sundevall, 163, and Thompson, Glossary, s.v.; Steier, 2415, notes that neither can be identified.

^{page 256 note 5} The identification as buzzard is accepted without question by Keller, , op. cit., 14 f.Google Scholar, and St.-Denis, Budé Pliny, H.N. 10, 158, n. 5 on § 204; Thompson, Glossary, s.v., is wisely rather more circumspect. Yet it may be worth pointing out, in view of the allegation about the hostility of the to toad and snake, that the latter creatures form a steady if small part of the buzzard's diet: cf. Rörig's surveys in the earliest years of this century, revealing that snakes and am phibians form between eight and nine per cent, of the bird's diet.

^{page 256 note 6} Cf. Keller, , op. cit. 242.Google Scholar The hostility here recorded by Pliny could be either that between predator or prey, or alternatively the aggressiveness that is typical of many marine birds at their nesting sites. For instance, shags and cormorants will go and nest in areas already occupied by other sea birds and drive them out.

^{page 256 note 7} Pintianus is reported to have misspelled gauiae when making the correction, but his equation of it with the Aristotelian is clear enough evidence that Mayhoff's and Detlefsen's attribution of this correction to him is justifiable. About the correction itself there can be no doubt in view of the frequency with which C and G are confused in capitals and uncials.

^{page 257 note 1} See Schlottmann, W., De auctoribus quibusdam in Plin. l. xviii (Rostock, 1893)Google Scholar; Kroll, in R.-E., xxi. 335.Google Scholar Aratus was known to Pliny probably only through a Roman intermediary's translation of the Aratea.

^{page 257 note 2} Jermyn, L. A. S., Greece and Rome xx (1951), 26 ff.CrossRefGoogle Scholar, and especially 33.

^{page 257 note 3} There is little point in attempting a precise identification of mergus and gauia here from the information given by Pliny about weather prognostication. For one thing, the data are too vague (cleaning of feathers and flight from the water would suit a good many species of sea birds); for another, the scientific value of such weather lore is virtually nil, as Armstrong, E. A., The Folklore of Birds (London, 1958), pp. 62f., 84 ff., 94 ff.Google Scholar, conclusively demonstrates. However, one further fact does emerge from Pliny's words here: since gulls and cormorants go to land in bad weather, while shearwaters remain at sea, it is clear that mergi in the present context cannot include shearwaters. I note this in view of what Thompson, D'A. writes in C.R. xxxii (1918), 95,Google Scholar although that scholar's general remark, that ‘in all probability the Shearwater was at least one of the birds which went by the name of Mergus in antiquity’, seems reasonable enough. Cf. p. 255, n. 2.

^{page 257 note 4} It is unlikely that the gauiae and mergi have anything in common with the pseudo-Theophrastean (§ 28), whose clamouring flight from the sea was inter preted as a sign of breeze or gale, depending on the loudness of the cry: may be a blanket term, but it does not include gulls and related species. Cf. Classica et Mediaevalia xxiii (1962), 234 n.Google Scholar

^{page 257 note 5} For Virgil, Aen. 5. 124 ff. (on which see below) is not a certain reference to the cormorant, even if a highly plausible one.

^{page 258 note 1} Epist. 3. 7 (p. 147. 22 Halm = p. 490. 22 Horn). Mr. A. M. Tynan, Curator of the Hancock Museum, Newcastle upon Tyne, confirms this interpretation. Cf. also Bannerman, and Lodge, , op. cit. viii. 5 f.Google Scholar

^{page 258 note 2} Exotericarum liber xv de subtilitate ad Hieronymum Cardanum (Paris, 1557)Google Scholar, ccxxxiii (on this work see Hall, V., Trans. Amer. Philos. Soc. xl [1950], 140 ff.)Google Scholar. The practice of taming cormorants for the purpose described is well known today in the Far East, where the native species is identical with that of the Mediterranean bird.

^{page 258 note 3} This is more or less the use to which Magnus, Albertus, De Animalibus, xxiii. 24Google Scholar (p. 2386 Venice edition of 1495 = ii, p. 1504 Stadler's Münster edition of 1916–20) puts the word: ‘mergus non tarn species auis est quam genus multas habens species’. The same definition was accepted by Gesner, in his Historiae Animalium ² iii. 108 ff.Google Scholar

^{page 258 note 4} Cf. Varro, L.L. 5. 78: mergus receives its name quod mergendo in aquam captat escam; and Ovid, Met. 11. 795 (which is discussed below).

^{page 258 note 5} Royds, , op. cit. (p. 250, n.4), pp. 37f.Google Scholar,firmly identifies the mergi of Virgil, Georg. 1. 361 f., as gulls (after Keightley), but Thompson, , loc. cit. (p. 257, n. 3)Google Scholar, sees here and in Claudian, Bell. Gildon. 492 f, references to the shearwater. Thompson's interpretation of the Virgil passage cannot be right, for reasons already advanced (see p. 257, n. 3); while the Keightley-Royds explanation forgets that cormorants as well as gulls leave the sea during bad weather.

^{page 259 note 1} So Thompson, , op. cit. on p. 257, n. 3, his p. 95, n. 1.Google Scholar

^{page 259 note 2} So Keller, , op. cit., p. 246.Google Scholar A further objection to the interpretation as 'skua’ may be advanced. Cicero, N.D. 2. 124, refers to a report which he has read (Keller suggests Pytheas as its author) about the platalea, and his description makes the equation of platalea with skua very plausible. Pliny, (H.N. 10. 115)Google Scholar copies this description, but calls the bird platea (if the manuscripts are to be trusted); consequendy, if he had meant to refer to the skua at 10. 130, only fifteen sections further on, why did he not call the bird platea there too ?

^{page 259 note 3} Cf. Bannerman and Lodge, xi. 289,301 f.

^{page 259 note 4} Gf. also w. 752 f., in which the name mergus itself occurs.

^{page 259 note 5} Most recently by Wilkinson, L. P., Ovid Recalled (Cambridge, 1955), pp. 169 ff.;Google Scholar cf. also Bardon, H. in Ovidiana (edited by Herescu, N. I., Paris, 1958), pp. 88f.Google Scholar This quality is closely related to what H. Fraenkel aptly terms the ‘clear concreteness’ of Ovid's sceneries (his Ovid [Berkeley and Los Angeles, 1945], pp. 85 f.).Google Scholar

^{page 259 note 6} Cf. Fraenkel, , op. cit., pp. 80 f.;Google ScholarWilkinson, , op. cit., pp. 160 ff.Google Scholar

^{page 259 note 7} Cf. the passage of Varro cited above, p. 258, n. 4. Wilkinson collects several other examples of Ovidian etymologizing, op. cit., p. 162.

^{page 260 note 1} So also are the legs of the grebe: which Keller, , op. cit., p. 241,Google Scholar and Steier, op. cit., 3417, fail to observe when they identify this Ovidian mergus as a Podiceps.

^{page 260 note 2} Since the middle of the last century commentators and translators have tended to fall into two classes with their treatment of mergis here. They either mistranslate the word as ‘gulls’, or play safe with an impre cise rendering, such as ‘sea-fowl’, ‘uccelli marini’, and the like. Earlier scholars were often nearer the mark: see Forbiger's note at Georg. 1. 361. Dryden's accuracy helps him to produce a good line: ‘The Cormorants above, lye basking in the Sun’; cf. John Martyn's translation of Georg. 1. 361 (but then Martyn was a Professor of Botany). Gavin Douglas also got it right with ‘skar-this’, which is the same word as the Scottish Gaelic sgarbh and the Scandinavian skarv.

^{page 260 note 3} On the similarity between the birds cited, see p. 255, n. 3.

^{page 260 note 4} Keller's, claim (op. cit., p. 244)Google Scholar that gauia and mergus are synonymous cannot be substantiated.

^{page 261 note 1} There are of course errors in Apuleius' description (thus no parrot has five toes, parrots do not dive on to rocks), yet it is accurate enough to make feasible the identification of the bird he describes as die Indian ring-necked parakeet (Psittacula krameri manillensis, formerly Palaeornis torquatus), even though Apuleius omits to mention the black part of the neck ring, and exaggerates the bird's powers as a mimic. The same species is described by Pliny, , H.N. 10. 117Google Scholar (perhaps a partial source of Apuleius here) and pictured on a Pompeii mosaic (Naples Museum, no. 9992) reproduced by Keller, , op. cit., plate 19.Google Scholar (I am indebted to Mr. J. J. Yealland, Curator of Birds, Zoological Society of London, for information here.)

^{page 261 note 2} According to Apuleius, the noctua produces cantum serum carmine intorto, the ulula produces cantum uespertinum carmine querulo, and the bubo produces icantum nocturnum carmine gemulo. Noctua is here the little owl (cf. Royds, , op. cit., pp. 45 ff.;Google ScholarKeller, , op. cit., pp. 39 ff.)Google Scholar, whose ‘low, wailing note’ is a characteristic accompaniment to the Mediterranean sunset (Tristram, , Natural History of the Bible, p. 195);Google Scholar the context here indicates mat serum has roughly this time reference. Carmine intorto is a little obscure, since intortus is applied both to the shrill, plaintive kiu of the little owl and to the more melodious song of the nightingale (Pliny, , H.N. 10. 81)Google Scholar. The context of the latter passage, however, suggests that intortus refers to continuous or legato sound as opposed to broken, staccato sound. The Apuleian ulula that sings in the evening (i.e., after sunset) carmine querulo seems to be the tawny owl (not the screech owl, as H. E. Butler translates) with its nocturnal tremulous song. The bubo (= eagle owl: see especially Keller, , op. cit., 36 f.Google Scholar, and Austin on Virgil, Aen. 4. 462: Butler is again wrong with ‘horned owl’) has a song that is well described as gemulo, and is characteristic of the late night: cf. Pliny, , H.N. 10. 34.Google Scholar Thus, however artificial Apuleius’ series may appear to be here, it is at least accurate so far as the owls are concerned.

^{page 261 note 3} Though contemporary popular usage seems the more likely source in view of what we know about the word's subsequent history: see p. 251, n. 5.

^{page 261 note 4} I say this as a warning against Steier's brusque dismissal of this passage as ornithologically useless, op. cit. 2417.

^{page 261 note 5} ‘Flies’, not ‘swims’, as Keller misinterprets (op. cit., p. 242: schwimmt…weg); pinnis natat is simply a typical Apuleian conceit derived from poetry (cf. Virgil, , Aen. 4. 245, 6. 16).Google Scholar

^{page 261 note 6} Thus editors and translators from Adlington on are wrong to translate the word here as ‘gull’, 'sea-mew’, and the like. Several of the terns dive into the sea and submerge completely, but such behaviour is most untvnical of epulis, although very occasionally a black-headed gull may be seen to dive beneath the surface (cf. Bannerman and Lodge, xi. 252). Dr. Coulson reminds me that the gannet occasionally penetrates to the Eastern Mediterranean (cf. Erhard, , Fauna der Kykladen [Leipzig, 1858], p. 61)Google Scholar, and that this bird's spectacular diving would impress any observer; but the equation of gauia and gannet here or elsewhere seems very unlikely in view of Pliny's statement that the gauia breeds in the Mediterranean area and generally lays three eggs.