The photolysis of benzobarrelene 7-A has been studied in considerable detail. Direct photolysis gives 7-C (benzocyclooctatetraene), but when acetone is used as a photosensitizer, the di- $\pi$-methane rearrangement product 7-B (benzosemibullvalene) is formed. A deuterium-labeling study gave the results shown. What mechanistic conclusions do you draw from these results? Is there a feasible mechanism that would have resulted in a different isotopic label distribution in 7-B?