The MIDI clock is MIDI's successor to DIN sync. (It doesn't appear to have an official name. The MIDI spec contains six system realtime messages, four of which work together to synchronise "clock-based MIDI equipment". I believe most people collectively call these four messages MIDI clock out of convenience.)
It's very similar to its predecessor, consisting of a steady stream of clock pulses sent at a rate of 24 pulses per quarter note (as you might expect, given that Roland co-wrote the spec); a start signal; a continue signal; and a stop signal. As MIDI, unlike DIN sync, sends discrete events as data packets along a single connection, it has some tradeoffs: the timing accuracy is slightly reduced (lagging by about 320 to 352 microseconds), due to the bottleneck of the rate MIDI packets are all sent at; but you save a lot of wires, because the MIDI sequencer's single MIDI output cable can send both the clock events and many overlapping note events (along with other things).
While the MIDI clock itself merely replicates the functionality of DIN sync, it can be combined with Song Position Pointer events, allowing the MIDI sequencer or multitrack recorder to tell the other equipment exactly where in the song it's up to. This makes it much more convenient than DIN sync, as you can now fast forward to halfway through a song and record just that part of a sequenced track, rather than having to record it from the beginning in order for the sequencer to know where it is. Not bad for the early 1980s!
- "The MIDI 1.0 Specification" Electronics & Music Maker, May 1984