Velocities  of  Cannon  Balls , &c.  7 1 
inverfe  ratio  of  the  fquare  roots  of  the  weights  of  the 
fhot.  But  the  accurate  ratio  cannot  certainly  be  drawn 
from  thefe  numbers,  on  account  of  the  doubtfulnefs  of 
fome  of  them,  as  was  before  obferved. 
It  is  very  remarkable,  that  in  the  experiments  of  this 
day,  the  mean  velocity  with  two  ounces  of  powder  is 
973,  whereas  it  was  no  more  than  626  in  the  former 
day  with  the  fame  quantity  of  powder,  notwithftanding 
the  balls  were  heavier  with  the  greater  velocity  in  the 
proportion  of  1 9 to  17  nearly.  This  remarkable  dif- 
ference mull  be  chiefly  owing  to  the  windage  in  the  firft 
courfe : and  from  hence  we  may  perceive  the  great  ad- 
vantage to  be  gained  by  the  ufe  of  balls  approaching  in 
proportion  nearer  to  the  diameter  of  the  bore  of  the  gun 
than  what  is  prefcribed  in  the  prefent  eftablifhment. 
Poflibly,  however,  fome  part  of  this  difference  might  be 
owing  to  fome  fmafl  inequality  in  the  powder,  as  that 
which  was  ufed  this  day  was  taken  from  the  bottom  of 
a barrel.  Perhaps  alfo  fome  part  of  the  effect  may  be 
owing  to  the  greater  degree  of  ramming  which  the  pow- 
der had  in  this  courfe. 
The  third  courfe  was  made  on  the  12th  of  June, 
1 7 7 5 , it  being  a clear,  dry,  and  calm  day.  The  powder 
in  the  experiments  of  this  day  was  rammed  in  the  fame 
degree 
