4 Newcomb and Diitton — Speed of 



about !N\ 60° E. It will now be seen that the planes of oscil- 

 lation of the first three clocks made wide angles with the di- 

 rection of motion of the first maximum, while the plane of the 

 fourth clock was almost exactly parallel with that direction and 

 perpendicular to the direction of motion of the second max- 

 imum. The fourth clock, then, may easily have escaped arrest 



until the second maximum while the other three would have 

 little chance of escaping the first maximum, even if they did 

 not stop during the lighter preliminary tremors. That the sec- 

 ond and third clocks stopped during the first maximum is ren- 

 dered probable by the way in which their pendulums were 

 caught. It would require a considerable acceleration in a di- 

 rection perpendicular to their planes of oscillation and at times 

 when the pendulums were near the extremities of their arcs of 

 vibration in order to throw their bobs far enough backward to 

 catch in the manner they did. These two clocks are relied 

 upon as giving the time of the first maximum. The chances 

 are, however, that the pendulums were not caught in this par- 

 ticular way during the first three or four oscillations, but went 

 staggering along for a very few beats until finally caught. An 

 interval of three or four seconds was probably occupied in the 

 rapid swelling of the quake from the preliminary milder phase 

 into the full power of the maximum. If we assume for the 

 beginning of the first maximum an instant of time about three 

 or four seconds earlier than that indicated by the two railroad 

 clocks, i. e. 9:51:12, our actual error, it is believed will not 

 exceed four seconds. The clock of James Allan & Co. prob- 

 ably stopped at a slightly earlier phase. If it may be assumed 

 to have been six or eight seconds slow, its stopping would have 

 been easily possible at that phase ; for many less sensitive 

 clocks throughout the country were arrested by tremors no 

 more forcible than. those in Charleston at the particular phase 

 thus indicated. We shall reach the same result, 9:51:12, if 

 we throw out the fourth clock as relating to the second maxi- 

 mum and (giving the weight 2 to both the second and third 

 clocks and the weight 1 to the first) take tlie mean readings of 

 the three. The whole tenor of the evidence from other clocks 

 in Charleston points strongly to a time a few seconds later than 

 9:51 for the first maximum. 



It is plainly necessary to select some phase of the earthquake 

 in Charleston or at the centrum as the beginning, with which 

 the beginning in all other places must be compared. It must 



