Reversal of normal moult sequence in the Kenneth Williamson In a short paper introducing the British Trust for Ornithology's Moult Enquiry (Williamson i960) I referred to the curiously aberrant wing-moult of the Spotted Flycatcher striata, in which the sequence of change of the primary feathers is ascendant (working from the outermost inwards), not descendant as in all other Palearctic . I pointed out that in other flycatchers (the genera , Xantbopygia, Empidornis, Bradornis and Dioptrorais have been examined), including the closely related M. gambagae of Africa which has often been placed as a subspecies of striata, wing-moult follows the orthodox sequence: this begins with the innermost primary (which I shall call pi, numbering the primaries descendantly) and works outwards to p9 and the minute pio, and also progresses inwards from the outermost secondary (si) to s6, the three tertials (often called SS7-9) moulting as a separate group after the start of the primary moult. Diesselhorst (1961) also drew attention to this reversal of the normal sequence in M. striata, adding that the moult of secondaries and tail-feathers is also abnormal. The latter, on the material available to him, showed a centripetal moult (progressing from the outermost feathers inwards to the central pair), while the secondaries and tertials appeared to be renewed from a single moult centre at s<5. A fuller study, based on 83 skins, was made by Stresemann (1963). She dissented from Diesselhorst on the moult of the secondaries, showing that the change begins with the inner group of three feathers usually called the tertials, ss6-4 then following descendantly and ssi-3 ascendantly. She wrote: 'This means a reversal of the order found in all other Muscicapinae, where three secondaries (9 to 7) [the tertials] are moulted in descending and six secondaries (1 to 6) in ascending sequence. The total reversal of moult-progression may originally have been caused by a single mutation. That mutation may have affected some central regulator which determines the sequence in the replacement of primaries as well as of secondaries and rectrices' (English summary, page m). Study of specimens in the collection of the British Museum (Natural History) suggests that this may be an oversimplification. The moult is a good deal more irregular in the Spotted Flycatcher than in most passerines studied, and this is especially true of the secondaries, as shown below. It would be tedious to list the exact stage of each 5° Moult in Spotted Flycatcher 51 of the British Museum examined, but the results can be summar­ ised as follows: (i) Moult of the primary coverts does not appear to be linked with the corresponding remiges as it is in other passerines. (2) Although the three outermost long primaries (PP7-9) are the first to moult, ascendantly, the short outermost pio often seems to moult later, as indeed it would in a normal sequence. (}) The alula or bastard wing is in moult from the beginning of December. It is moulting concurrently with pp7-io in most of the skins, this being the stage at which it would normally begin in a descending moult. (4) As stated by Diesselhorst, the tail-moult is centripetal and the usual pattern is for all the feathers to be moulting with the growth most advanced towards the outermost. Tail-moult does not appear to be in phase with any particular stage of wing-moult. (5) The tertials precede the secondaries as in other passerines. Six skins from south and east Africa with pp7-io and s6 growing have new tertials. (6) The moult of secondaries often begins descendantly from s6, as in other passerines, but perhaps as frequently it begins ascendantly from si. The progression seems to be from both ends of the series irrespective of whether si or s6 started first, so that S3 or S4 is the last to finish. Thus there does not appear to be a 'single centre' of moult at s6 (as in Diesselhorst), or a simple reversal of the normal passerine sequence (as in Stresemann), but an irregularity in the replacement of the true secondary feathers (ssi-6) which obscures any fixed pattern, though if such exists it is in all likelihood a nearly simultaneous start at either end of the series.

REFERENCES DIBSSBLHORST, G. 1961. 'Ascendente Handschwingen-Mauser bei Mtisckapa striata*. J. Orn., 102: 360-366. STRESEMANN, V. 1963. 'Zur Richtungsumkehr der Schwingen- und Schwanzmauset von Muscicapa striata'. J. Orn., 104: IOI-III. WILLIAMSON, K. i960. 'Moult as a study in field-'. Migration, 1: 171-175.

Kenneth Williamson, British Trust for Ornithology, Beech Grove, Tring, Hertfordshire