The 12B4A is probably the closest tube (for operating characteristics) to a 1626, even though the RP, and gm and mu are somewhat different, they bias up nearly identically in the classic 12SL7/1626 Darling / Clementine circuit design.
FWIW, a 45 is essentially one half of a two-plate 2A3, ergo the 2A3 is an internally paralleled pair of 45s sharing the same jug
/ed B