A second method was to appeal to the earth’s rotation on its axis. By ascribing a value of 24 hours to the complete rotation of our planet, we could divide its surface with equally distanced meridians, and then assert that the passage of the same star from one meridian to the next would always take the same time. Suppose, then, for argument’s sake, that the earth’s surface had been divided by 24 half meridians, all equally spaced. Then if a star passed a certain meridian at at a certain time , it would pass the contiguous meridian one hour later. In order to synchronise two clocks situated on these two meridians and , all that would be necessary would be to set the first clock at “zero” hour as the star passed its meridian , then set the second clock at the hour “one” as the star passed over the second meridian . The clocks would then be synchronised, so that an event occurring by the first clock as its hands marked a time , would be simultaneous with an event occurring by the second clock as its hands also marked . But this method again was open to certain objections which it is unnecessary to recall since we have discussed them in a previous chapter. They deal with the retarding influence of the tides and with the earth’s “breathing.” To be sure, these objections are of a hair-splitting nature, but they are theoretically valid just the same since they do not reduce to mere errors of observation. A third method of synchronisation considered by classical science was obtained by appealing to propagations, which would serve to establish communications between the spatially separated clocks. The propagation of sound waves suggested itself in this capacity. According to this method we should operate as follows: At noon, sharp, as marked by our chronometer at the centre, we should fire a gun. All the observers standing by their clocks on the circumference would then be instructed to set their clocks, the instant they heard the report, at the hour “noon plus the time required by the sound to progress from centre to circumference.” Given the velocity of sound in air, and given the distance from centre to circumference, this lapse of time could be computed, and the clocks synchronised. But here again we should have to make sure that no wind was blowing; also, even in the case of a stagnant atmosphere, we should have to assume that the speed of sound was constant and remained the same in all directions, i.e., that its propagation was isotropic. We knew that this was the case to a first approximation, but no highly refined experiment had ever established the fact. [Pg 182]