220 ANNUAL OF SCIENTIFIC DISCOVERT. 



teach how, for a pendulum of given length and weight, to measure the 

 speed of a vehicle on which it was carried, by observing its deviation 

 in front of the vertical. 



Such is the principle which I have desired to subject to experiment, 

 in order to test the possibility and the sensibility of a measure so con- 

 trived. This may be done by an experiment easily reproduced, and 

 which although deprived of elegance is none the less decisive. If we 

 get into a railway-carriage and resting our arm against a vertical sup- 

 port, hold a pendulum suspended in connection with the graduated arc 

 of a circle, in conformity with the conditions indicated above, and if, 

 with a proper amount of attention, we preserve this instrument as 

 much as possible from the effect of lateral shocks and vertical move- 

 ments, we may observe the following result. We shall hardly have 

 given the bob a slight retrograde impulse, than under the influence of 

 traction we shall see it almost immediately spring forward through an 

 angle which, for an ordinary speed of 28 miles an hour will soon 

 increase to about 33 degrees. In falling back on the contrary the 

 backward deviation is hardly 5 or 6 degrees, and the same motion con- 

 tinuing so long as the velocity lasts, there is thus a very considerable 

 and permanent inequality between the two branches of the oscillation. 

 This is in some degree the characteristic part of the phenomenon. If, 

 in this state of things, the velocity of traction should diminish, the 

 direct deviation diminishes nearly as rapidly ; and I have never failed, 

 in experimenting in this manner, to become aware of my approach to 

 a station, without taking my eyes off the pendulum. The inequality 

 between two branches of the oscillation is also sensible for inferior 

 velocities, and for less than 28 miles an hour, and we still see the pen- 

 dulum advance 10 degrees before the vertical. The rate of ships 

 being ordinarily comprised between 6 and 18 miles per hour, the vari- 

 ations of amplitude corresponding to such velocities, have a very 

 appreciable sensibility. It is clear that there must be certain dimen- 

 sions which will belong to the maximum of effect ; but as to that, 

 experiment does not always bear out suppositions of a complicated 

 and difficult nature, and it is experiment that will have to pronounce. 



If we wish to arrive at true precision in the kind of measure of 

 which I have been speaking, the great and real difficulty lies evident- 

 ly in the construction of apparatus capable of protecting the point of 

 suspension of the pendulum, whether from jerkings and shocks on a 

 railroad, or, above all, from the pitching and rolling of a ship. It is 

 upon this point that my researches are being directed. Already I have 

 made some attempts, and it was to try to combine these several dispo- 

 sitions that I seek to improve our prospects, as regards time, by this 

 communication. Perhaps it will never be possible to obtain an appar- 

 atus of continuous action, nor an apparatus capable of working inde- 

 pendently of the manual address of the operator. But many very 

 useful instruments are precisely in the same condition with regard to 

 continuity. 



