138 A1ST EMPIRICAL STUDY OF G\ RATING BODIES. 



General Barnard, in his second article on the gyro- 

 scope, offers an explanation 1 which does not explain. 

 He merely says that these upward movements are due to 

 a component which, when the axis is above the hori- 

 zontal, sends the instrument up, and when the axis is 

 below the horizontal, sends it down. This is undoubt- 

 edly true, but not very satisfactory, for he does not ex- 

 plain how, or whence, the component comes. 



For lack, then, of other explanation, I offer one of my 

 own : 



As already stated in the list of things done by the 

 gyroscope, two conditions are necessary for the success 

 of this experiment. (1.) The point of support when the 

 axis is horizontal must be a little above the axis pro- 

 duced. (2.) When set in operation, the axis must be 

 very much inclined. I find an angle of 60 u or 70° gives 

 good results, 2 and that nothing is better to rest it upon 

 than a pane of glass. 



Fig. 21 will help us to trace out the curious combina- 

 tions that lead to this result ; a b is a bar of light wood 

 eight inches long, two and one -half inches wide, and 

 one-half .inch thick ; p is a cone of tin passing through 

 the bar, and ending in a small steel cap. The standard 

 ends in a steel wire four inches long, on the polished toj) 

 of which the steel cap rests ; e is a movable weight to 

 balance the gyroscope ; d is a cap of tin that slides 

 snugly over the bar ; 1, 2, 3 and 4 are small tubes 

 soldered to d into which the lug of the gyroscope fits 

 firmly. By means of these the gyroscope can be placed 

 at different angles to the bar, a b. The cap d can be 

 moved towards or from the center. 



If the wheel be set in rapid motion, and placed as in 



1 Am. Jour. Ed., vol. iv., page 532. 



2 Probably with careful adjustment of the point, and with a very small load in propor- 

 tion to the mass of the rim of the wheel, and with higher speed than I can give, the gyro- 

 scope would rise from a much lower angle. I cannot say what the limit is, except that the 

 axis must be above the horizontal. 



122 



