This description really does benefit from a visual aid, found here. (Plate 22 is on the left.) Note that the blade of our fencer on the left is to the outside of his opponent's blade, but to the inside of his arm. This is important later.
Plate 22, "A third wounding another third."
To break down the actions that get us to the image:
Both fencers start in Third, on the inside. Distance isn't specified, but I'm thinking misura larga. To be clear, in this plate, Fencer A initiates the action and B responds, but A will still wound B.
- Fencer A begins with a feint, straight in.
- Fencer B moves to parry, dropping the hilt to catch A's sword.
- In the tempo of step 2, A performs a cavazione to the outside, thrusting his sword through the angle created by B parrying and dropping his guard.
- B will be unable to push A's sword out because:
- A's forte will be right up against B's blade, as in the picture, and
- A's blade will be locked against B's arm. (Note that this could be a little problematic for the SCA, where if B is pushing hard enough outwards before A strikes him, A's blade could be locked into place, not having struck anything. It's an edge case, but technically possible.)
Examining this, and comparing it to Plate 21, we can see a lot of similarities. (Which makes sense, if we're seeing the fundamental actions of the system here.) Contrasting it to the second variation of Plate 21 (with the second response that he might use, returning to the outside and wounding in Third over the opponent's blade) , only assuming that 21's Fencer A strikes his opponent, shows us an almost identical ending.
Things of note (which are, unsurprisingly, the same as in the previous plate) include the victor moving in the tempo of their opponent's movement. Mezzo tempo actions are huge for Fabris, and it makes sense that he starts us off with them. Also, we can see a total lack of blade contact until the forte is right up against the opponent's sword - if even that much. Related to that, we start in Third. If there's no reason to leave it, we can stay in Third, regardless of whether we're on the inside, outside, or whatever. There's no blade contact, no need to really oppose the opponent's blade, so no need to move into another guard. Straightforward. Finally, we're making an attack without first gaining the opponent's blade; this is usually a bad idea, but in this case we're reacting to a specific response that the feint flushed out, and taking that tempo away. Since our opponent can't do two things at once, we can strike them while they're parrying, so we should be safe. (I feel that it's still far better to gain the blade and go from there, though.)
Another way the action could happen is like this:
Both fencers are again in Third, on the inside.
- Fencer A moves to find B's blade.
- B performs a cavazione and steps in.
- A takes that tempo and thrusts during the cavazione.
Fabris states that this illustrates that the motion of a cavazione is slower than staying in the middle and moving on a straight line. While normally this might not be the best idea, if the opponent isn't attacking, but merely disengaging around to the other side and stepping in, attacking does make a lot of sense according to Fabris.