In parallel powering the + side of the supply is fed to the microphone via one modulation (signal) lead, and the current returns via the other modulation lead; the shield isn't involved. This contrasts with phantom powering, in which the two modulation leads carry the DC voltage in common mode, while the shield is the "ground" for powering purposes.
The "k" in the model designation stands for "Kudelski" (= Nagra). Sennheiser and Nagra had introduced parallel (or "T") powering for condenser microphones, but the DIN committee then standardized the opposite signal polarity from what those two companies were using. When Schoeps introduced the CMC ("Colette") series in 1973/74, they offered a standard-compliant CMC 4-- amplifier model (without the "k"), and later added the "k" model which was powered, and had its audio signal polarity arranged, the same as the Sennheiser shotguns of the time.
Parallel powering was never prevalent in recording studios, so for the most part, people didn't have to worry about polarity conflicts among multiple microphones. Powering multiple T-powered microphones from a central supply is problematic in terms of crosstalk as well; few people mourn its passing.
Filtering of the DC supply is critical for parallel powering, since any residue of hum or other noise will be directly overlaid onto the audio signal (= differential mode). Also, the DC must be blocked on the output side of the power supply, or else it will flow across the inputs of the preamp, recorder or mixer to which the microphone is connected--but I'm fairly certain that it is blocked by electrolytic capacitors in the BZ 41; it's easy enough to verify that.
If your supply is unmodified, the "n" in its name would indicate a 10 dB resistive pad at the output, which you could presumably bypass today. When the CMC 4-- was new, it was still considered a very high-output microphone, and preamp overload in general was more often a problem then than it is today.