﻿Chemistry 
  and 
  Physics. 
  447 
  

  

  four 
  positions 
  in 
  one 
  complete 
  revolution 
  will 
  be 
  found 
  at 
  each 
  

   of 
  which 
  almost 
  complete 
  silence 
  prevails, 
  that 
  is 
  whenever 
  one 
  of 
  

   the 
  four 
  directions 
  of 
  destructive 
  interference 
  coincides 
  with 
  the 
  

   vertical 
  axis 
  of 
  the 
  resonance 
  column. 
  If, 
  while 
  the 
  fork 
  is 
  held 
  

   fixed 
  in 
  one 
  of 
  these 
  positions, 
  a 
  tube 
  made 
  of 
  card-board 
  or 
  of 
  

   glass 
  be 
  pushed 
  over 
  one 
  of 
  the 
  prongs 
  (especial 
  care 
  being 
  taken 
  

   to 
  avoid 
  touching 
  the 
  prongs) 
  then 
  the 
  sound 
  becomes 
  much 
  

   louder, 
  since 
  the 
  waves 
  from 
  the 
  covered 
  prong 
  are 
  screened 
  off 
  

   and 
  interference 
  is 
  precluded. 
  By 
  alternately 
  screening 
  and 
  

   unscreening 
  the 
  prong 
  with 
  the 
  pasteboard 
  cylinder 
  the 
  sound 
  

   can 
  be 
  raised 
  and 
  lowered 
  at 
  will. 
  I 
  think 
  that 
  this 
  is 
  one 
  of 
  the 
  

   most 
  striking 
  demonstration 
  experiments 
  in 
  interference." 
  — 
  

   Phys. 
  Ztschr., 
  p. 
  11, 
  Jan., 
  1920. 
  h. 
  s. 
  u. 
  

  

  8. 
  Poiseuille's 
  Law 
  for 
  Compound 
  Tubes. 
  — 
  The 
  question 
  as 
  

   to 
  whether 
  Poiseuille's 
  law 
  holds 
  for 
  complex 
  systems 
  of 
  tubes 
  

   having 
  many 
  branch 
  points 
  is 
  undoubtedly 
  one 
  of 
  importance 
  in 
  

   the 
  study 
  of 
  the 
  blood 
  circulation 
  of 
  vertebrates 
  and 
  it 
  may 
  not 
  

   be 
  without 
  interest 
  to 
  experimental 
  physicists. 
  This 
  problem 
  has 
  

   been 
  successfully 
  attacked 
  by 
  J. 
  Schleier 
  with 
  the 
  apparatus 
  

   described 
  in 
  the 
  next 
  paragraph. 
  

  

  A. 
  trunk-tube 
  of 
  0-32 
  mm. 
  internal 
  diameter 
  was 
  blown 
  larger 
  

   at 
  one 
  end 
  and 
  divided 
  into 
  two 
  branches. 
  By 
  means 
  of 
  short 
  

   sections 
  of 
  rubber 
  tubing 
  each 
  of 
  these 
  branches 
  was 
  connected 
  

   with 
  a 
  single 
  tube 
  that 
  was 
  in 
  turn 
  forked 
  into 
  two 
  branches. 
  

   This 
  process 
  of 
  doubling 
  and 
  connecting 
  was 
  repeated 
  until 
  the 
  

   outer 
  circle 
  of 
  tubes 
  comprised 
  64 
  separate 
  segments, 
  the 
  total 
  

   number 
  of 
  tubes 
  being 
  126. 
  All 
  of 
  the 
  tubes 
  at 
  the 
  end 
  of 
  the 
  

   system 
  were 
  united 
  in 
  a 
  single 
  outlet 
  tube. 
  The 
  lengths 
  of 
  the 
  

   tubes 
  decreased 
  from 
  the 
  intake 
  end 
  of 
  the 
  system 
  to 
  the 
  middle 
  

   and 
  then 
  increased 
  to 
  the 
  outlet 
  end. 
  The 
  least 
  and 
  greatest 
  

   lengths 
  were 
  2-5 
  cm. 
  and 
  15 
  cm. 
  respectively. 
  The 
  tubes 
  next 
  to 
  

   the 
  outlet 
  pipe 
  had 
  the 
  largest 
  diameter, 
  namely 
  051 
  mm. 
  The 
  

   end 
  of 
  the 
  system 
  was 
  also 
  provided 
  with 
  a 
  set 
  of 
  vertical 
  tubes 
  

   by 
  means 
  of 
  which 
  the 
  lateral 
  pressure 
  could 
  be 
  determined. 
  

   Each 
  tube 
  was 
  separately 
  calibrated 
  both 
  as 
  to 
  circularity 
  and 
  

   uniformity 
  of 
  bore 
  by 
  the 
  usual 
  mercury 
  thread 
  method. 
  Care- 
  

   fully 
  distilled 
  and 
  repeatedly 
  filtered 
  water 
  was 
  used 
  as 
  the 
  

   liquid. 
  Compressed 
  air, 
  at 
  a 
  pressure 
  of 
  130 
  cm. 
  of 
  water, 
  was 
  

   employed 
  to 
  drive 
  the 
  liquid 
  through 
  the 
  system. 
  

  

  The 
  loss 
  of 
  "head" 
  was 
  calculated, 
  by 
  the 
  aid 
  of 
  Poiseuille's 
  

   formula, 
  from 
  the 
  viscosity 
  of 
  water 
  at 
  the 
  prevailing 
  tempera- 
  

   ture, 
  the 
  dimensions 
  of 
  the 
  tubes, 
  and 
  the 
  mass 
  of 
  water 
  that 
  

   flowed 
  out 
  of 
  the 
  system 
  per 
  second. 
  The 
  mean 
  of 
  a 
  number 
  of 
  

   runs 
  was 
  found 
  to 
  give 
  a 
  calculated 
  value 
  slightly 
  less 
  than 
  7 
  per 
  

   cent 
  lower 
  than 
  the 
  loss 
  of 
  head 
  actually 
  observed. 
  A 
  discus- 
  

   sion 
  of 
  the 
  unavoidable 
  additive 
  errors 
  arising 
  from 
  the 
  length- 
  

   ening 
  of 
  the 
  system 
  by 
  the 
  rubber 
  connecting 
  tubes, 
  from 
  the 
  

   irregularities 
  at 
  the 
  branch 
  points, 
  from 
  the 
  curvature 
  of 
  the 
  

   stream 
  lines 
  at 
  these 
  points, 
  and 
  from 
  the 
  calibrations 
  of 
  the 
  

  

  