400  Mr.  Mond  and  Dr.  Wilderman  on  a 
clip  (7),  provided  with  spring  and  clamping  screw,  serving 
to  hold  the  spindle  (f)  in  operative  engagement  with  the 
half-nut.  The  driving  is  effected  through  spur-wheels, 
supported  hy  the  bracket  (/3) ;  the  bevel-wheel,  as  shown  in 
detail  in  figures  II.,  III.,  and  IV.,  being  arranged  to  run  on 
ball  bearings.  The  arrangements  of  the  pen-carrying  system 
are  similar  to  that  described  with  reference  to  the  horizontal 
type  of  instrument,  the  electrical  connexions  being  also 
substantially  the  same. 
The  vertical  type  is  an  improvement  on  the  horizontal  type, 
though  usually  made  of  a  smaller  size,  since  it  allows 
continuous  reading  of  the  records  on  the  drum,  can  stand 
more  knocking  about,  is  easier  to  be  accurately  constructed, 
and  gives  also  better  results. 
When  the  whole  spiral  is  completed,  the  instrument  stops 
automatically  by  itself.  In  the  horizontal  type  we  raise 
then  the  spindle  (ff)  from  the  half-nut  (s),  and  push  it 
through  the  boss  of  pinion  (K)  back  to  its  former  position, 
stopping  first  the  governor  by  means  of  (7').  In  the  vertical 
type,  we  open  clip  (7),  release  the  screw  of  the  spindle  from 
the  half-nut,  and  push  it  up  to  its  former  position. 
Both  types  are  set  into  motion  or  stopped,  whenever 
desired,  by  stopping  or  releasing  the  governor  bv  means 
of7';   ^ 
While  costly  and  complicated  arrangements  for  calibration 
each  second  are  indispensable  for  instruments  with  revolving 
drums,  all  we  require  for  very  accurate  work  with  our 
instrument  where  the  movement  of  the  pen  is  very  regular  is 
to  calibrate  it,  using  an  ordinary  -J-  sec.  or  T]o  sec*  watch  and 
electric  key  at  longer  intervals,  say  each  minute  (at  the 
same  time  as  we  make  our  research  observations),  and  in  this 
case  we  do  not  even  require  to  exactly  adjust  the  absolute 
value  of  a  revolution,  having  only  to  use  a  factor.  The  same 
or  almost  the  same  error  in  calibration  falls  upon  a  minute 
instead  of  upon  a  second,  and  the  results  obtained  in  this 
mariner  are  much  more  accurate  and  satisfactory.  Since  no 
work  can  be  done  for  nothing,  any  extra  work  of  the  pen 
spent  on  calibration  is  not  conducive  to  the  accuracy  of  the 
results  obtained,  and  it  is  well  that  we  are  able  to  avoid 
frequent  calibration. 
Up  to  the  present  we  have  tried  a  great  number  of 
arrangements  for  calibration  some  of  them  very  complicated 
and  difficult.  The  great  accuracy  and  regularity  in  the 
movement  of  the  writing-pen,  as  well  as  the  great  length  of 
the  spiral,  make  our  vertical  instrument  a  very  handy  and 
easilv  manageable  recording  calibrator  for  other  instruments 
