Generalising variety

After Schmidt section, specialize the m. Restrict to polarization which generates NS, maybe justify why m is always 2. Add \ell factors where it's missing. Comment how \ell^2 only affects whether \beta_{-} in rational?