def base26_dublet_for_bits_56_to_64(a):
    b0 = a[7]
    b1 = a[8] & 0x01
    h = b0 | b1 << 8
    return d26[h]