90 
Dr.  W.  E.  Sumpner  on  the 
tests  were  33  in  number,  and  are  indicated  on  the  curve  by- 
delta  signs.  About  40  tests  were  taken  with  alternating 
currents,  half  of  them  with  a  non-inductive  resistance  in 
series  with  a  moving  coil,  and  the  rest  with  a  condenser 
substituted  for  this  resistance.  These  tests  are  separately 
marked  on  the  curve  with  rings  and  crosses.  The  observations 
are  not  all  plotted,  as  several  of  the  points  were  too  close  to 
be  distinguishable,  but  a  fair  selection  is  given. 
A  few  typical  observations  with  direct  currents  are  given 
in  Table  I.  and  will  be  sufficient  to  show  the  conditions  of 
the  test.  The  currents  A2,  A3  are  in  amperes,  x  is  the 
observed  deflexion  of  the  pointer  in  degrees,  and  <f>  is 
calculated  from  formula  (8). 
Table  I. 
A2... 
+2-2 
+  1-1 
+  •42 
-•5 
-2-8 
-5-2 
-6-2 
-8-9 
-9  3 
-9-5 
A3... 
2-7 
7-0 
7-2 
7*45 
7-6 
10-9 
7-2 
8-2 
4-3 
0-5 
a:   ... 
30 
40 
50 
605 
70 
80 
100 
1197 
139 
161 
0  ... 
176-4 
157-3 
153-0 
146-7 
12S-9 
121-7 
97-4 
85-9 
57-3 
326 
The  value  of  <£  so  calculated  corresponds  in  the  alternating 
tests   with   the    phase    difference    between    the    current   A1 
through  the  first  fixed  coil,   and  the  moving  coil  current. 
For  both  the  direct  and  alternating  current  tests,  the  deflexion 
x  of  the  instrument  could  be  read  to,  and  appeared  reliable 
to,  one-tenth  of  a  degree.     The  value  of  <£  was  determined 
with  almost  equal  accuracy  for  the  direct  current  tests,  the 
conditions  for  which  were  simple  and  could  easily  be  kept 
constant,  while  the  value  of  $  depended  merely  on  the  ratio 
of  the  readings  of  two  excellent  instruments  of  the  permanent 
magnet  type.     But  <f>  could   not  be   determined  with  such 
accuracy  in  the  alternating-current  tests,  for  several  reasons. 
The  testing    conditions    with   running    machinery    involved 
cannot    easily   be    kept   very   constant  ;    alternating-current 
instruments  are  not  so  reliable  as  those  for  direct  currents  ; 
while  <f>  has  to   be   determined  from  the  readings  of  three 
instruments  instead  of   from  those  of  two.     The  alternator 
used  was  provided  with  six  terminals  arranged  for  the  supply 
of    two-phase    and   three-phase    current.      The   three-phase 
terminals  were  connected  to  the  positive  ends  of  the  fixed 
coils   through   ammeters   and  banks  of   lamps,  the  negative 
ends  being  connected  together  to  form  the  neutral  point. 
The  currents  were  adjusted  to  approximate  equality,  but  only 
