Charles, you have found a very interesting challenge to our guidelines. The would be no question if the author had used the word "per" instead of a slash, or if the actual numerator and denominator of the problem were used in the narrative ("3 mol CO2" instead of "3 moles of CO2", etc.). However, this is not the case. The slash launches us into Nemeth Code. The fraction, then, is "3 moles of CO2 [per or over] 1 mole of C3H8" which means the second transcription is the correct one. I have repeated it here for clarity.
(See the next post for the simbraille)