282 J. ALFRED GRIFFITHS. 



In other instruments, such as Osier's Anemometer and Moscrop's 

 Engine Recorder,* instead of the pen moving proportionately to 

 the increment of space, the pen takes up a position indicating the 

 velocity of motion, and only moves when the velocity changes ; 

 but in such instruments the inertia of the mechanical parts seriously 

 interferes with the perfection of the record. 



In the writer's instrument both these principles are involved. 

 The pen is moved by the prime variable, and the length of the 

 line measured perpendicular to the motion of paper measures the 

 time integral. One pair of axles in the train moving the pen 

 carry friction rollers, which can be periodically separated or 

 recoupled by tripping clockwork. Thus instead of the saw-teeth 

 on the diagram measuring equal intervals of space, they indicate 

 equal intervals of time, and consequently the strokes are of vary- 

 ing length, the length of each being an exact measure of the time 

 integral during one interval. If this lifting period be made 

 indefinitely small, and at the same time the vertical scale of the 

 pen's motion be increased in inverse proportion, the lines will be 

 very close together, and their envelope will be the differential 

 coefficient of the main function, that is, will exactly measure the 

 speed or velocity of the prime motion. In practice the period is 

 not indefinitely small, and the actual length of the line gives the 

 total space moved in the lifting interval, and also the average 

 velocity during that interval, and the interval may be made small 

 enough to give the differential curve with sufficient exactness for 

 most purposes. 



The mechanism designed for this purpose by the writer may be 

 almost indefinitely varied in detail according to Circumstances, 

 but requires three essential parts : — 



First, an axle or roller moving continuously with the prime 

 motion, and continuously tending to move the pen at right angles 

 to the motion of the paper. 



Second, a clock to measure the time interval and to advance- 

 the paper band with a regular motion, which may be continuous 

 or intermittent. 



* English Patents, 1321 of 1881 and 6965 of 1884. 



