.--A survey of the imitative abilities of birds reveals a spectrum ranging from species such as the White-crowned Sparrow (Zonotrichla leucophrys) and Song Sparrow (Melospiza melodla), which may copy only the songs of conspecifics, to versatile mimics in the families Sturnidae, Miraidac, and Pslttacidae, which may reproduce a variety of alien vocalizations and sounds (reviews in Nottebohm 1972, Baptista 1972). There are few records of song mimesis by estrildid finches (review by Immelmann 1969a). Some forms appear to be better mimics than others. Two estrildid species, namely the African Silverbill (Lonchura cantans) and the Zebra Finch (Taeniopygla guttara), cross-fostered under Bengalese Finches (Lonchura striata), learned the songs of their foster-father (Immelmann 1967, 1969a). On the other hand Nicolai (1965) and Giltringer (1970; see also Giltringer and Achermann 1972) list a number of estrildid species that produced only species-specific vocalizations when reared by Bengalese Finch foster-parents. I know of no record of a Gouldian Finch (Chloebla gouldiae) mimicking the song of another species as de- scribed here. A single Gouldian Finch was hatched and raised to maturity by a pair of Bengalese Finches along with three of their own young. In neighboring cages were an assortment of estrildids and fringillids, including a pair of Red Avadavats (Amandrva amandava) that the GouldJan Finch could hear and see. I separated the GouldJan Finch from the foster-parents when it could fend for itself at a little over 1 month in age. There- after it was kept in a large cage containing other estrildid species including adult Bengalese, a pair of GouldJan Finches, and a pair of Red Avadavats. The young Gouldian Finch eventually sang a song containing no elements of the songs of its foster-parents, nor of its own species, but only of the Red Avadavat. Hall (1962) described the songs of the three estrildid species under discussion. The song of the GouldJan Finch is typical of most estrildids in that it is soft and hardly discernible by the human ear beyond a few yards (Hall 1962, Immelmann 1965). Song spectrograms in Hall (1962) and Thorpe (1961: 113) reveal the complex struc- ture of the Gouldian Finch song. This involves "the simultaneous production of two or more notes each harmonically distinct" in which "a series of low-pitched clicks provides a more or less unbroken accompaniment to at least two other independent sets of higher notes, with a high-pitched descending whine cutting right across a second series of slightly lower notes" (Hall, 1962: 49). Thorpe (1961: 112) suggested that three vocal mechanisms may be operating simultaneously to produce this complex song. The song of the Red Avadavat is easily distinguished from the above by "the loud- ness and unusually simple harmonic structure of the individual notes and by their organisation into a well-defined...musical song-phrase" (Hall 1962: 44). The notes are delivered on a descending scale (Goodwin 1960, Harrison 1962a, Figure 1A). My captive wild-caught male usually gave a series of very soft "tit" calls (Figure 1C) before singing. Its song usually ended with a toy-trumpet-like "geh" (Figure 1A) that sounded very similar to the contact call of the Zebra Finch. A pair of Zebra Finches in a neighboring cage would call excitedly in response to the "geh" of the Red Avadavat each time it completed its song. 2.5 3 5 4 I recorded imitation songs of the Red Avadavat by the cross-fostered Gouldian Finch when it was 149 days old (Figure lB). These songs were always whispered, as in the songs of conspeciflcs. Except for element b being absent in 13 spectrograms, the imitated songs were quite stereotyped in that the same elements appeared in the same sequence within a total of 15 song spectrograms examined. The Gouldian Finch was never heard giving the soft "tit" (Figure 1C) calls with which the Red Avadavat tutor introduced its songs. The "geh" (Figure 1A) element is also absent in the mimicked songs (Figure lB). In other respects the mimicked songs bear many of the features present in the songs of its Red Avadavat tutor, namely in: (1) "the simple harmonic structure of the individual notes" (Hall 1962), (2) the notes delivered in descending pitch, and (3) the morphology of notes a to i (Figure lB). Whereas the notes in the song of its tutor (Figure 1A) are discrete, some of the whistles in the songs mimicked by the Gouldian Finch seem to fade away (Figure lB). They are frequency and amplitude modulated, and appear to be poor copies of the tutor's notes (see unlabelled notes in Figure lB). Notes d and e in the mimicked songs (Figure lB) also differ slightly in morphology from notes d and e in the tutor songs (Figure 1A). Some im- provisation appears to have been involved in the construction of these two notes. The male Red Avadavat tutor was mated and therefore sang only occasionally (see Harrison 1962b). Possibly a Gouldian Finch exposed to more Red Avadavat songs than was this one might copy the tutor songs more accurately. The duration of five of the tutor's songs without the terminal "geh" (Figure 1A) ranged from 2.83 to 2.89 sec. ( = 2.86 sec.; SD = 0.026 sec.). The duration of 12 of the mimicked songs ranged from 2.84 to 3.05 sec. (.: 3.00 sec.; SD = 0.089 sec.). These figures indicate that the latter's songs were also stereotyped in duration and were a good approximation of the length of the tutor songs. In normal Gouldian songs, which measure much longer the beginning and end of each motif is never clearly marked (Hall 1962). Immelmann (1967, 1969a) postulated an %motional bond" between tutor and tutored as a prerequisite to song learning in the three estrildid species he studied. Addi- tionally the cross-fostered individuals were irrevocably imprinted on the foster species (Immelmann 1969b, 1970). The cross-fostered Gouldian Finch was observed in a cage with other estrildid species from November 1971 to late March 1972. It did not appear to be imprinted on its foster species, namely the Bengalese Finch, nor on the Red Avadavat whose song it had learned. The cross-fostered Gouldian Finch always roosted near the two adult Gouldian Finches at night and followed them during the day. No social bond was apparent between this GouldJan Finch and its alien tutor, but that Gouldian Finches raised by Bengalese may be sexually imprinted on them is well-known (Immelmann 1965). Probably the fact that this cross-fostered Gouldian Finch was separated from its foster-parents immediately when weaned prevented its being imprinted (see Immelmann 1965). Estrildid finches also may differ individually in disposition to being imprinted (see e.g. Goodwin 1971). No GouldJan or Bengalese Finch song elements appeared in this GouldJan Finch's mimicked themes (Figure lB) possibly because my adult male Gouldian and Bengalese Finches did not sing enough to effect any learning of their songs. Nicolai (1965) reports having seen many Gouldian Finches raised under Bengalese Finches singing only Gouldian Finch songs. Nicolai's (1965) observations raise a number of questions regarding song develop- ment in Gouldian Finches. Firstly could it be that imitation is not important in normal song development of young Gouldians? If so the song mimicry, as I have reported for this species, is "facultative," as appears to be the case with the Oregon Junco (Junco oreganus) and the Song Sparrow (Melospiza melodia) (see review by Nottebohm 1972). In view of the complexity of normal GouldJan song, exposure to songs of adult conspecifics may possibly be prerequisite to normal song development. A spectrographic study of songs of Gouldian Finches raised under Bengalese Finches may reveal certain aberrancies in syllabic or temporal structure. Secondly, could it be that GouldJan Finches may imitate certain sounds (e.g. in conspecific song or in Red Avadavat song) and not others (e.g. in Bengalese song)? Interestingly, the notes in the songs of Bengalese (see spectrograms in Immelmann 1969a) and the "geh" ele- ment in the tutor song (Figure 1A) neither of which did my Gouldian Finch learn (Figure lB) are structurally very similar. Only a detailed laboratory study of song development by sound spectrography may answer these questions more adequately. I thank Ned K. Johnson and Gerhard Thielcke who read various versions of the manuscript and offered critical comments. LITERATURE CITED BAPTISTA, L. F. 1972. Wild House Finch sings White-crowned Sparrow song. Z. Tierpsychol. 30: 266-270. GOODWn% D. 1960. Observations on Avadavats and Golden-breasted Waxbills. Avicult. Mag. 66: 174-199. GooDwn% D. 1971. Imprinting, or otherwise, in some cross-fostered Red-cheeked and Blue-headed Cordon-bleus (Uraeginthus bengalus and U. cyanocephalus). Avicult. Mag. 77: 26-31. GiiTTnmER, H.R. 1970. Zur Evolution yon Verhaltensweisen und LautSusserungen bei Prachtfinken (Estrildidae). Z. Tierpsychol. 27: 1011-1075. GiiTTIGER, H. R., A*D J. ACERrA. 1972. Die Gesangsentwicklung des Kleinel- sterchens (Stern*estes cucullata). J. Ornithol. 113: 37-48. I-I, M.F. 1962. Evolutionary aspects of estrildid song. Syrup. Zool. Soc. Lond. 8: 37-55. I-IAmISON, C. J.O. 1962a. The affinities of the Red Avadavat, Amandava amandava. Bull. Brit. Ornithol. Club 82: 126-132. I-lmusoN, C. J. O. 1962b. Solitary song and its inhibition in some Estrildidae. J. Ornithol. 103: 369-379. IrrErANN, K. 1965. Australian Finches in bush and aviary. Sydney, New South Wales, Angus and Robertson Ltd. IrrErANN, K. 1967. Zur ontogenetischen Gesangsentwicklung bei Prachtfinken. Verh. Deutsch. Zool. Ges. GiSttingen 1966, Zool. Anz. 30 suppl: 320-332. IrrELrANN, K. 1969a. Song development in the Zebra Finch and other estrildid finches. Pp. 61-74 in Bird vocalizations (R. A. Hinde, Ed.). Cambridge, England, Cambridge Univ. Press. IrrFrAm, K. 1969b. Uber den Einfluss frfihkindlicher Erfahrungen auf die geschlechtliche Objectfixierung bei Estrildiden. Z. Tierpsychol. 26: 677-691. IrrELIANN, K. 1970. Der Zebrarink. Die Neue Brehm Bficherei. Wittenberg Lutherstadt, Ziemsen Verlag. NmOLAI, J. 1965. Vogelhaltung-Vogelpflege. Stuttgart, Kosmos Bucher. NOTTEBO, F. 1972. The origins of vocal learning. Amer. Naturalist 106: 116-140. TORPE, W.H. 1961. Bird-song.' Cambridge Monogr. Exp. Biol., No. 12. Lugs F. BAPTISTA, Museum o! Vertebrate Zoology and Department o] Zoology, Uni- versity o! Cali]ornia, Berkeley, Cali]ornia 94720. Present address: Max-Planck- Institut fiir Verhaltensphysiologie, Vogelwarte Radol]zell, 7761 M6ggingen, den Am Schlossberg, West Germany. Accepted 24 Oct. 72.