I;58  Mr.  swift’s  Account  of 
filently  with  the  points,  and  fo  entirely,  that  the  electro- 
meter falls  to  50.  With  balls  a quarter  of  an  inch  dia- 
meter, indeed,  it  paffes  off  with  a little  hilling  noife,  but 
this  feldom  amounts  to  a fhock:  but  with  balls  three- 
quarters  of  an  inch  diameter  an  explofion  conllantly  hap- 
pens, and  the  magazines  are  fired. 
To  put  this  matter  ftill  more  out  of  doubt,  I place  a 
ball  of  nine  inches  diameter  on  one  of  the  conductors,  and 
the  explofion  is  very  violent,  always  more  certain;  and 
yet  the  machine  does  not  difcharge  itl'elf,  for  the  electro- 
meter falls  not  more  than  20°. 
The  next  experiment  I make  with  the  water  conduc- 
tor is,  placing  the  houfes  a,  b,  c,  in  a negative  Hate,  by 
eoniieCting  them  with  the  culhion  of  the  machine,  or 
with  the  outfide  of  a battery : when  the  cloud  is  charged 
and  paffes  over  the  houfes,  with  points  at  the  upper  end 
of  their  conductors,  there  is  no  explofion;  the  points 
feem  to  draw  off  all  the  eleCtrical  matter  during  the  paf- 
fage  of  the  clouds  of  four  feet  and  a half  long:  but 
when,  in  this  pofition  of  the  houfes,  balls  of  three-quar- 
ters of  an  inch  diameter  are  placed  inltead  of  points, 
there  is  a fmall  explofion,  and  a confiderable  rejiduum  of 
the  matter  is  left  in  the  battery.  I then  change  the  in- 
fulated  water  for  wire  to  compleat  the  circle:  on  the 
paffage  of  the  clouds  over  the  houfes  there  is  a confi- 
4.  derable 
