I'm starting to work through Fux's Gradus ad Parnassum. For the two-voice, first-species, Phrygian-mode, lower-voice exercise (p. 36 on the first Google result for "Fux Gradus"), Aloys says that he allows the voice-crossing
because otherwise [he] would have had to use direct motion up to this point, which would have resulted in less satisfactory voice leading.
What's he talking about?
What makes this composition
more acceptable than, say, this? (edited to correct some of the extraneous errors pointed out in the first answer)
To be clear, the upper voice in both is the cantus firmus, which I can't change. Also, both voices I think are assumed to be adjacent since we're only considering two parts for now.