﻿72 
  Scientific 
  Intelligence. 
  

  

  filled 
  in, 
  and 
  thus 
  obliterated, 
  by 
  the 
  combined 
  effects 
  of 
  diffusion 
  

   of 
  light 
  and 
  of 
  molecular 
  agitation. 
  This 
  is 
  the 
  first 
  satisfactory- 
  

   explanation 
  of 
  the 
  observed 
  fact 
  that 
  the 
  Fraunhofer 
  lines 
  are 
  

   entirely 
  absent 
  from 
  the 
  continuous 
  spectrum 
  of 
  the 
  solar 
  corona. 
  

   —Jour. 
  d. 
  Phys., 
  7, 
  89, 
  1917. 
  h. 
  s. 
  u. 
  

  

  7. 
  Direct 
  -Reading 
  Density 
  Balance 
  for 
  Solids. 
  — 
  When 
  great 
  

   precision 
  is 
  not 
  required, 
  the 
  density 
  of 
  a 
  given 
  liquid 
  may 
  be 
  

   obtained 
  at 
  once 
  and 
  without 
  calculation 
  from 
  the 
  scale 
  reading 
  

   of 
  a 
  suitable 
  hydrometer. 
  On 
  the 
  other 
  hand, 
  the 
  evaluation 
  of 
  

   the 
  density 
  of 
  a 
  solid 
  by 
  the 
  ordinary 
  method 
  of 
  hydrostatic 
  

   weighing 
  is 
  a 
  much 
  slower 
  process 
  since 
  it 
  involves 
  the 
  experi- 
  

   mental 
  determination 
  of 
  two 
  numbers 
  and 
  the 
  calculation 
  of 
  the 
  

   ratio 
  of 
  these 
  numbers. 
  The 
  disadvantages 
  of 
  the 
  latter 
  method 
  

   are 
  entirely 
  avoided 
  by 
  employing 
  a 
  balance 
  of 
  the 
  form 
  devised 
  

   by 
  Jules 
  Gasnault. 
  The 
  sphere 
  of 
  applicability 
  of 
  this 
  balance 
  

   is 
  the 
  precise 
  analogue 
  of 
  that 
  of 
  the 
  hydrometer 
  for, 
  with 
  a 
  

   fixed 
  calibration, 
  it 
  gives 
  direct-readings 
  for 
  solids 
  having 
  densi- 
  

   ties 
  lying 
  between 
  certain 
  limits. 
  Just 
  as 
  a 
  set 
  of 
  hydrometers 
  

   is 
  required 
  for 
  liquids 
  having 
  a 
  wide 
  range 
  of 
  densities 
  so 
  also 
  

   in 
  the 
  case 
  of 
  the 
  new 
  balance 
  different 
  scales 
  and 
  other 
  modifica- 
  

   tions 
  would 
  have 
  to 
  be 
  introduced 
  to 
  cover 
  extended 
  intervals 
  of 
  

   density. 
  

  

  The 
  balance 
  in 
  question 
  and 
  the 
  manner 
  of 
  using 
  it 
  may 
  be 
  

   briefly 
  described 
  as 
  follows: 
  An 
  assay 
  balance 
  has 
  the 
  right- 
  

   hand 
  portion, 
  say, 
  of 
  the 
  beam 
  prolonged 
  by 
  a 
  rod 
  of 
  some 
  light 
  

   metal 
  such 
  as 
  aluminium. 
  A 
  portion 
  of 
  this 
  extension 
  to 
  the 
  

   right 
  of 
  the 
  right-hand 
  scale-pan 
  is 
  graduated 
  in 
  such 
  a 
  way 
  as 
  

   to 
  indicate 
  densities 
  directly. 
  The 
  specimen 
  of 
  solid 
  (which 
  

   must 
  sink 
  in 
  the 
  liquid 
  employed) 
  is 
  first 
  placed 
  in 
  the 
  right- 
  

   hand 
  pan 
  in 
  air. 
  It 
  is 
  counterpoised 
  by 
  an 
  equivalent 
  tare 
  

   placed 
  in 
  the 
  left-hand 
  pan. 
  The 
  tare 
  is 
  conveniently 
  made 
  up 
  

   from 
  a 
  cheap 
  set 
  of 
  weights, 
  the 
  smallest 
  weight 
  having 
  a 
  mass 
  

   of 
  one 
  gram. 
  The 
  fine 
  adjustment 
  is 
  effected 
  by 
  sliding 
  a 
  rider 
  

   along 
  the 
  left-hand 
  segment 
  of 
  the 
  balance 
  beam. 
  The 
  moment 
  

   fx 
  of 
  all 
  the 
  vertical 
  forces 
  due 
  to 
  the 
  parts 
  of 
  the 
  tare 
  is 
  then 
  

   (approximately) 
  equal 
  to 
  V.D.a, 
  where 
  V 
  — 
  volume 
  of 
  solid, 
  

   D 
  = 
  density 
  of 
  specimen, 
  and 
  a 
  = 
  horizontal 
  distance 
  from 
  the 
  

   (•••Hi 
  i-al 
  knife 
  edges 
  to 
  the 
  knife 
  edges 
  supporting 
  the 
  right 
  scale- 
  

   pan. 
  The 
  solid 
  under 
  investigation 
  is 
  next 
  suspended 
  by 
  a 
  fine 
  

   wire 
  of 
  negligible 
  mass 
  from 
  the 
  prolongation 
  of 
  the 
  beam. 
  

   This 
  wire 
  lias 
  an 
  open 
  loop 
  at 
  the 
  upper 
  end 
  so 
  that 
  it 
  can 
  be 
  

   slid 
  along 
  the 
  beam 
  extension. 
  The 
  solid 
  is 
  immersed 
  in 
  an 
  

   appropriate, 
  inactive 
  liquid 
  contained 
  in 
  a 
  rectangular 
  trough 
  

   the 
  long 
  axis 
  of 
  which 
  is 
  roughly 
  parallel 
  to 
  that 
  of 
  the 
  beam. 
  

   The 
  adjustment 
  consists, 
  therefore, 
  in 
  displacing 
  the 
  loop 
  along 
  

   the 
  beam 
  cxlonsion 
  until 
  the 
  beam 
  is 
  again 
  horizontal 
  (as 
  it 
  was 
  

   bolh 
  before, 
  and 
  after 
  the 
  specimen 
  and 
  tare 
  were 
  counterpoised). 
  

   The 
  moments 
  now 
  satisfy 
  the 
  condition 
  p 
  = 
  V 
  (D 
  — 
  d) 
  o, 
  where 
  

  

  