Audacity does not do that in one step.
He should use SHIFT + A to stop, which sets the cursor at the stop position.
Audacity would record in a new track, starting from the time point that was clicked. When played or exported, the professor’s recording would be mixed into the silences in the student rendition.
It would be more convenient to select the silent area then he would record for that length of time in the new track and recording would stop automatically.
Or if he does talk for too long, click at the end of the student’s phrase, CTRL + I (as in “eye”) to create a split, F5, then drag the student’s track rightwards until it is after the end of the professor’s feedback.
The professor would end up with a lot of tracks, and should probably do CTRL + A and Tracks > Mix and Render from time to time to reduce the number of tracks.
A better variant would be to create a new single track for the professor’s recording. Click in that track at the time point to record the response then SHIFT + R. That will always record in that track and not create a new one. If the professor goes on too long, do all the splits and drags as the final step.
Or you could insert pauses beforehand with something like Inserting pauses to aid transcription. or Extend Silences
Whichever way you look at it, Audacity is not ideally suitable for this. We need something like Missing features - Audacity Support.
It does not affect this discussion, but the current version is 2.0.6 Audacity ® | Download for Windows.
Gale