88 Experiments upon Gttnpowder, 



Here the agreement of the actual and computed re- 

 coils is as remarkable as that of the actual and com- 

 puted velocities in the foregoing table. 



By the Figure 17 may be seen at one view the result 

 of all these experiments and computations. The num- 

 bers upon the line A B (as in the Fig. 16) represent the 

 weights of the bullets, while the lines drawn from those 

 numbers perpendicular to A B on each side, and ending 

 at the curves w, n^ are as the velocities of the bullets in 

 the several experiments ; the line A B being the axis of 

 the curves, the lengths taken from A to the different 

 numbers towards B (= k) the abscissas, and the perpen- 

 diculars (=jy) the corresponding ordinates. The ordi- 

 nates to the curve hn are as the velocities computed 



from the theorem, ; r= y, and the ordinates to 



the curve p^ n (which is the logarithmic curve, as it is 



-p = jy) shew the velocities computed upon Mr. 



Robins's principles. The curve gf is drawn from the 



theorem — — :, -{- ^ = % ; and the actual recoil is 



\^a -\- x" 



marked upon the ordinates to this curve by large round 

 dots, which in all the experiments except the 86th and 

 87th very nearly coincide with the curve. 



In the Fig. 18, the numbers upon the line A B, taken 

 from A, denote the different charges of powder used in 

 the course of the experiments, while the ordinates to the 

 curve c ^ express the velocities of the bullets with the 

 vent at o. The lines drawn perpendicular from the line 

 A B to the line d"/ represent the recoil with the several 

 charges of powder and a leaden bullet, and the portion 

 of those lines that is comprehended between the line 

 A B and the line g h denotes the recoil when the given 

 charge was fired without any bullet. 



