In Section VIII of Relativity (On the Idea of Time in Physics), Dr. Einstein proposes a thought-experiment in which "Lightning has struck the rails on our railway embankment at two places A and B far distant from each other." He then goes on to show that the two lightning strikes may be regarded as simultaneous or asynchronous depending on one's frame of reference - hence opening the door to a new understanding of time as relative to the frame of reference of the observer. All very well and good. But can we not use this idea to construct an experiment? Substitute firecrackers in carriages on the train for lightning strikes on the embankment. One explosion occurs many cars "forward", the other many cars "backward." An observer on the train is centered between the explosions. An observer on the embankment would see the explosions as simultaneous, but... With respect to the on-train observer: Scorches on the carriage floors tell us that the firecrackers were equidistant from the observer. Light from the two flashes tells us that the events were not simultaneous, in accordance with the discussion in Section VIII. But despite the asynchronous flashes, the reports of the two firecrackers will arrive at the observer's position at the same time, telling the observer that the explosions were simultaneous. Right? I would expect that the comparison of simultaneity of any two events using light and sound would yield measurable discrepancies precisely because light travels independently from the frame of reference of the observer, but sound travels in the frame of reference of the atmosphere. This means that an experiment could be constructed which would provide strong validation of SR using the simultaneity discrepancy between light-based and sound-based observations. Does anyone know of any such experiments?