General description:

By July, primary moult is obvious as the outermost flight-feathers are dropped now and the birds look 'short-winged'. Most birds have the wing-coverts completely fresh and neat, as the gaps in the median and greater wing-covert rows have been closed. By the end of the month, many birds are still moulting lesser coverts in the upper row, which may be difficult to detect, covered by the lowest scapulars. Some birds still show old second generation feathers, which were included in the post-juvenile moult last autumn (see 1cy September & October). The inner median coverts which were moulted first in line in the complete moult two months ago, start to bleach already and loose the warm tones. The majority of 2cy michahellis is easily identified by (some) adult-like grey scapulars.

By July, 2cy michahellis in the last stage of the complete moult (second pre-basic moult) can be found commonly around the harbour of Etaples in NW France with 10's of birds in the direct surrounding of the fish-market and 100's of immature birds using the near landfill of Dannes to feed on.

Grey-based scapulars or barred scapulars:

The images on the right-hand side show 2cy michahellis, classified by their general appearance: the top birds are typical "barred individuals". The fresh wing-coverts and the fresh scapulars show a barred pattern. Just by this pattern, these birds look more immature than the birds in the bottom images.
The bottom images show 2cy michahellis which have grown in many fresh plain grey scapulars and some wing-coverts. Therefore they appear more adult-like, although the amount of replaced feathers is not necessarily larger than in the barred individuals of the top images. Its just the look: grey scapulars suggest maturity. Compare this barred individual with this more advanced looking individual; both 2cy michahellis from July.

Moult stage and pace of remiges and rectrices:

We have extensive surveys on primary moult stage of both 2cy michahellis and 2cy argenteus. The figures can be viewed in the tables and cross-tables below. We described 2cy birds in detail on six occasions at Etaples, Calais region.  Peter Adriaens kindly provided video footage to score primary moult stages in Croatian birds. Late July 2001, a visit was organized to Lignano, Italy, to score over 100 2cy michahellis on this location.

By 3^rd July, 2cy michahellis from Croatia have an average 3.0 primaries left (n =n 27), i.e. P8-P10 still juvenile. During the month, the outer primaries are shed as well, with primary moult score of 2.2 (n = 45) and 2.3 (n = 48) remaining old primaries at Etaples, NW France by 10^th July. By the end of July, 2cy michahellis has only one or two old juvenile primaries left, with one-third of the birds having dropped all primaries. 
Compared to argenteus, these michahellis figures show that the average gap in the primaries is larger in michahellis (the gap between new second generation primaries and retained old juvenile primaries). 

By early July, the average 2cy michahellis has P5 renewed and fully grown. By the end of the month this score has increased one primary, P6 fully grown by 28^th July. The most advanced birds by late July show an obvious short wing, with P10 dropped and P7 already fully grown. One bird has P8 fully grown. Note the relatively low SD value (standard deviation) in these primary moult scores, which implies a small range of variation. 

The majority of 2cy michahellis have started the moult in the rectrices and the secondaries. As can be seen in previous sections, michahellis may moult rectrices in the partial moult in 1cy, but the complete moult in the tail-feathers is normally postponed until July; when the inner four primaries are grown, the moult in both secondaries and rectrices will start.

Moult stage and pace of wing-coverts:

In several sections of 1cy michahellis it's mentioned that birds include wing-coverts in the post-juvenile moult (unlike argentatus / argenteus where this post-juvenile moult is confined to the scapulars). To distinguish between 2cy birds which have still suspended moult (after the post-juvenile moult) and birds which have really started to replace wing-coverts, we separate between old second generation feathers and fresh second generation feathers in the April, May and June sections. This is much easier said than done, especially by May and June, when the old second generation feathers (required as early as late-August) and the juvenile wing-coverts may be extremely similar and abraded.
Its interesting to see that the wing-coverts, which were included in the post-juvenile moult are left out of the first stage of the complete moult in early summer. They are replaced at a later stage, after July in some individuals. This is nicely demonstrated by a Swiss metal-ringed michahellis Sempach 885629. The exact figures of such "stepping" moult sequences are hard to obtain, as birds which do drop these earlier moulted feathers are not recognized of course. However, irregular moult sequences are not found in the tertials, as in most birds, replacement in the tertials follows the ordinary sequence, from tertial #1 downwards.

The complete moult in the wing-coverts may start by April, with the majority of 2cy michahellis starting moult by early May. Except the difference in timing of moult, these new second generation feathers may show a different pattern than second generation feathers from the partial moult. Fresh second generation feathers moulted from April and May onwards appear 'less immature': the base colour of the feathers is not buff-brown but buff-grey or feathers may be "grey-based" and the pattern may be less well-defined.

The images on the right-hand side and table 5 present figures and illustrations of active wing-covert moult by July. In most birds, progress in wing-covert moult is most obvious in the lesser coverts, which show moult gaps in almost all individuals. Moult in the median and lower lesser coverts is finished in most birds, and by early July, many birds still show growing central greater coverts. The lesser coverts may show complete missing rows, exposing the bases of the fresh lower lesser coverts. In a few images, old second generation feathers are still in place: here and of course in this Swiss individual. The phenomenon of leaving older second generation coverts in place, while moult progresses in the same tract, was clear by June, but now by July, most of these old second generation are dropped as well, and may be an explanation for single grey and more adult-like feathers (see e.g. here and here). See Gull Topography for explanations of feather tracts.

As can be seen in many images of May, June and early-July 2cy michahellis, the sequence of moult in the wing-coverts is sometimes hard to obtain correctly. Post-juvenile feathers are maintained over a longer period and result in a 'jumpy' moult pattern and only after these old post-juvenile feathers are replaced for new third generation feathers, the wing-covert rows appear neat and tidy. It's the end of July then...

Moult stage and pace of scapulars:

Moult to third generation scapulars may start as early as mid-January in 2cy michahellis. By June, all present birds show active moult in the scapulars, with largest numbers for the class 75%-99% third generation scapulars (n = 24). By the end of July, this moult in the scapulars is completed, with fresh feathers in the lowest scapular row. 
Previous sections of 2cy michahellis show the most commonly seen moult sequence in the scapulars, starting with the lower upper scapulars and finishing with the rear lowest scapulars. So not surprisingly, its mainly the lowest scapulars which may still be in active moult by July, as is illustrated by this bird.

Tables:

Table 1 & 2: 2cy Larus michahellis July and 2cy Larus argentatus argenteus July primary scores to compare.

Cross-Tables Larus michahellis 2cy:

Table 3 - 8: 2cy Larus michahellis primary moult scores, horizontal new primaries, vertical old primaries in Etaples and Lignano, July.