> When I performed a simple DFT on data to find a period, I > freezed up: the Fourier spectrum showed two clearly evident > periods with values 258 days (same as GCVS and Kiss et al.) > and also 131 days. Quick look at averaged lightcurve showed > that these is caused by well pronounced hump on ascending > branch, which recently developed into double maximum. Ondrej, V Boo isn't on my list for obvious reasons, but there are similar specimens in the southern sky. For example R Cen and R Nor have distinct double periods but are classed as M. Recent speculation about T Cen suggests it may be M with two periods, one of which fades in and out of detection. L2 Pup (SR) ranges between episodes of Mira-like regularity and complete incoherence. S Car (M) often exhibits lumps on its lightcurve like an SR. And so on.... The distinction between SR and M is crumbling, as we find more and more stars unexpectedly "changing sides" or "misbehaving" like the above examples. I reckon that the classes M and SR should be viewed as two (temporary) aspects of the same phenomenon – if they are not already. As to where the L class fits in I can only speculate. But to continue the example of L2 Pup, if you only had the light curve for the 1998-1999 seasons to examine, you would probably have concluded that this star was a low-amplitude irregular. To quote a Bhuddist friend of mine: "all things are impermanent". So keep watching. Cheers, Fraser Farrell