1896.] on Chronographs and their Application to Gun Ballistics. 185 



primary does the secondary occur, miglit I suppose be tested by means 

 of the revolving mirror — but this would not have been entirely satis- 

 factory, inasmuch as it would not have tested the actual record on the 

 drum. So I devised the following, which, though apparently very simple, 

 requires care to get good results. On the rim of the drum I insert a 

 piece of ivory. Fitted to the bed-plate is a hinged piece of brass 

 whose far end presses against the rim of the drum. The circuit from 

 the primary wire of an induction coil runs through the brass piece 

 and the drum, except when it is interrupted by the ivory. A sharp 

 break here occurs, which leaves its record on the drum by means of 

 the steel pins and secondary current, as before described. If the 

 drum is revolved slowly, the spark will give the true position at 

 which the ivory breaks the circuit. If, now, there is any retardation 

 or delay in the record of the spark, it will be shown on the drum 

 when it is rotated rapidly — the record lagging behind that obtained 

 by the slow break. Knowing the speed of the drum, the time of 

 retardation can be obtained. 



Fig. 5. 



I have made several of these experiments. On the table are some 

 of the records . To test the variability of the records, it suffices to 

 move the recording points along the drum at each break, when the 

 records should be in a straight line. These specimens will show you 

 how accurate they are. Great care must be exercised to turn the 

 rim perfectly true and smooth ; also the brass piece rubbing against it 

 must be often smoothed up. 



The measurement of these retardations is a delicate matter, as we 

 are dealing with a retardation of only 10 millionths of a second. 

 I think that the improvements I have carried out in these instru- 



