upon &c. " * 307 
prove, that the inflammation of gun -powder r i^ inftantaheous, 
or that the whole charge is in. all cafes inflamed, and 4:4 cori- 
46 verted into an elaftic fluid' before the bullet i's fenfibly* ftioved 
46 from its place fuch reafonings and conclufioiis may lead to 
very dangerous errors. 
It is undoubtedly true, that if the principles affumed by Mr. 
robins with refpect to the manner in which gumpowcfer takes- 
fire, and the relation of the elafticity of the generated fluid to 
its desifity, or the i^tenfity of its preflure upon the bullet as it 
expands in the barrel, were juft, and if the loft of force by the 
vent and windage was in all cafes in confide ruble, or if it was 
prevented, the deductions from the theory refpe&ing the velo- 
cities of bullets of different weights would always hold good. 
But if, on the contrary, it fhould be found upon making the 
experiments carefully, and in luch a manner hs intirely to pre- 
vent inaccuracies ariling from adventitious circumftances, that 
the velocities obferve a law different from that which the theory, 
fuppofes, we may fairly conclude, that the principles upon 
which the theory is founded are erroneous. 
Let us now fee how far thefe experiments' differ from the 
theory. Thofe numbered from 84 to 92 mclufive were made 
in fuch a manner that no part of the force of the powder was 
lbft by the vent, of by windage, as has already been mentioned, 
and all poffible attention was paid to every circumftance that 
could contribute to render them perfect and conclufive.- 
A particular account of them with the means ufed for form- 
ing the bullets, and making them fit for the bore, and the con- 
trivance for preventing the efcape of the elaftic fluid by the 
vent, &c. may be feen in the general table, p. 245, The foU 
lowing table fhews the refill t of them. 
’ \ 3 : : ‘ r ; Ni B s 
