The ballast in one of the 4ft 40/36W fluorescent lamps in my garage says that it can be used with 1x 1200mm 40/36W, 2x 600mm 18/20W and 1x 1050mm 40W fluorescent tubes. The current rating on the ballast is 0.46A which is the same as every other 4ft 40W ballast except this one says it's compatible with with the 1 meter (or 1050mm) tubes.

I'm guessing you should just try this tube on a normal 40W ballast then!

@Eliot_240: But the lamps I had found are 970mm long so as Medved said the lamp current of those would be 0.556A wich is more than "standart" 36/40w 0,43A

Sorry, I didn't check the measurements of the lamp in this thread properly before posting! I don't know what the 1050mm long lamps are at all then.

The PL-L 55W & F54T5/HO are in 460-530mA range. Ballasts for these lamps are accommodating of wide range of lamp voltage and OCV doesn't really matter as long as it isn't too low as they're all programmed rapid start.  Filaments are heated first, then arc voltage is applied, so you don't really have to worry about lamp instant starting which was a problem with "rapid start" electronic type that applied arc voltage and filament voltage at the same time. 

These ballasts should be readily available. 

PL-C is a 320mA design, so a it'll work on a PL-C ballast, but will be under-driven. 

Generally saw metre tubes in fridge-freezer cabinets in supermarkets in the 70s and 80s both here in the UK and in Europe, but certainly never used for other purposes such as domestic, etc. Europe never switched from the standard imperial lengths because this would be costly for everyone involved from manufacturers having to explain this to stockists and users, stockists having to double their space to house two different sets of tube lengths; confused members of the public buying and returning "wrong" length tubes or worse still "adapting" standard luminaires to accommodate metric lengths irrespective of electrical compatibility.

This was an explanation I once received from a rep from an unnamed company whilst working in the sign industry at a time when signmakers began switching to metric sizes.

Back to lamp lengths, the only metric size was the one-metre as far as I can recall, but it just so happens that the standard five-foot is a metre and a half long!!!!

/Filaments are heated first, then arc voltage is applied, so you don't really have to worry about lamp instant starting which was a problem with "rapid start" electronic type that applied arc voltage and filament voltage at the same time. 

This is true only on some ballast designs (only those using PTC for preheat control and/or series choke with electronic starters).
All other do apply voltage across the tube during preheat phase (try to touch the "hot" side, it will glow), while this voltage is supposed to be too low to ignite the discharge in the designated tube. But if you connect tube with the striking voltage below the voltage during preheat, it would start prematurely (in fact as RS). On some dimmable ballast (mainly based on the the IR21592/3 chip) this would be a problem, as the state machine in the controller would not see the current pattern expected to signal the lamp ignited, so it would fall into FailureToStrike error mode (so shut down the lamp)

It is consequence of how these ballasts work: During preheat the lamp LC circuit is only operated further from the resonance, so there flow enough current to preheat the lamp (in the current mode preheat circuit) and/or the voltage on the inductor is quite high (in voltage mode preheat circuit) to heat up the filament, but the voltage is not sufficient (with margin) to strike the discharge. When it is supposed to ignite the lamp, the frequency start to sweep down towards resonance, so the voltage rise till the lamp ignite. And as the lamp ignite, it damp the LC tank, what reduce the voltage boost, so all voltage and currents (mainly trough the capacitor) in the circuit drop, what reduce the heating power to negligible level (so it could be stated, then it in fact stop heating filaments).

This was the first attempt at making a metric tube, introduced in the 1970s, with a length of 970mm so as to realise a 1000mm long fitting.  It was a commercial failure and really did not take off.  Part of the reason was that its efficacy was much lower than the standard 4' 40W tube and it needed a new higher current ballast which was rather large.

The second metric tube was the 1047mm 40W T12 (actually 42W) introduced in EU by either Philips or Osram, I am not sure who was first.  This was more logical because it was developed to suit the standard 600mm metric ceiling modules, which are slightly too short to accept all of the imperial length tubes except the 6' length.  However the 40W T12 1047mm tube was also slow to be accepted because its efficacy was about 10% less than a regular 4' 40W T12, and it needed a new and large ballast.

The problem was elegantly solved by Thorn who invented the 1047mm 40W T8 (actually 39W) in 1975.  The change of glass diameter increased tube voltage to almost exactly the same level as a 4' 40W T12 so that the same ballast could be used - and efficacy also matched that of the imperial version.  That was the first metric tube which really made sense and achieved much better success for a number of years.  More recently, the tube was re-named 38W (although it still in reality consumes 39W).  The same tube is also sold in a U-bent version 525mm long, for use in single 600x600mm ceiling modules.

Today these metric T12 and T8 tubes have almost completely disappeared, except for replacement purposes.  The long-standing problem of metrication was solved by the modern T5 long tubes, whose lengths are all about 5cm shorter than the imperial tubes to enable their compatibility with 600mm modular ceiling systems.

F40T12 is metric too. Length less pins is 1200mm rather than 48.0".

There is also a strange lamp that is designed to utilize the same ballast but shorter by exactly 4cm called 40W/1160mmT12/RS