414  Mr.  H.  Morris- Airey  on  the 
needle  if  this  exceeds  4  or  5  per  second.  But  the  range  may 
be  extended  by  opening  out  the  plates ;  for  as  already  shown 
the  frequency  of  the  needle  for  constant  difference  of  potential 
is  inversely  as  the  distance  apart  of  the  plates,  provided  that 
the  minimum  distance  used  exceeds  three  times  the  length 
of  the  needle. 
The  constant  k  is  most  simply  defined  as  the  volts  across 
the  plate  required  to  produce  a  frequency  of  oscillation  of 
1  per  sec.  for  a  distance  apart  of  the  plates  equal  to  1  cm. 
Then  if 
d—  distance  apart  of  plates, 
N  =  frequency, 
V  =  voltage  between  plates, 
we  have 
V=MxN (Ill) 
An  instance  of  this  application  has  been  given  already 
Another  may  be  added  : — 
The  difference  of  potential  yielded  by  a  Wimshurst  machine 
was  measured  by  charging  a  Leyden  jar,  the  outer  coating 
being  connected  to  one  pole,  the  inner  to  the  other  pole,  the 
poles  being  kept  well  apart.  The  jar  was  connected  with 
the  pair  of  parallel  plates,  which  were  at  a  distance  of  10'5  cms. 
apart.  The  frequency  of  the  needle  in  the  field  between 
them  was  2'04  per  sec.  The  needle  used  had  a  constant 
(reduced  to  d=l  cm.)  of  1300  volts.  Applying  formula  III, 
we  have 
7=1300  x  10-5x2-04 
=27800  volts. 
The  machine,  it  may  be  added,  was  working  badly,  giving 
a  spark  between  balls  of  2  cms.  and  5  cms.  diameter  respec- 
tively, Leyden  jars  being  in,  of  something  less  than  1  cm. 
length. 
These  experiments  were  made  in  the  laboratories  of  the 
Birkbeck  College.  I  would  record  my  thanks  to  the 
Principal  for  facilities  for  carrying  them  out. 
XXXIY.    On  the  Resolcinq  Power  of  Spectroscopes. 
By  H.  Moeeis-Airey,  2LSc* 
THE  object   of  this  note  is  to  call  attention  to  a  factor 
which    appears  to  have  been  neglected  in  the    consi- 
deration of  the  resolving  power  of  spectroscopes. 
It  has  been  shown  by  Kayleigh  that,  using  an  indefinitely 
narrow   slit,  the  resolving  power  of  a   grating-spectroscope 
*  Communicated  by  tho  Author. 
