﻿228 
  Scientific 
  Intelligence. 
  

  

  with 
  some 
  general 
  physical 
  quantity 
  such 
  as 
  atomic 
  weight. 
  As 
  

   is 
  well 
  known, 
  partial 
  success 
  was 
  first 
  attained 
  by 
  Kayser 
  and 
  

   Runge, 
  and 
  by 
  Rydberg, 
  who 
  showed 
  that 
  the 
  frequency 
  differ- 
  

   ences 
  v 
  between 
  the 
  components 
  of 
  a 
  doublet 
  series, 
  and 
  the 
  

   differences 
  v 
  lf 
  v 
  2 
  between 
  the 
  components 
  of 
  a 
  triplet 
  series, 
  vary 
  

   in 
  such 
  a 
  way 
  that, 
  for 
  the 
  same 
  column 
  of 
  the 
  Mendelejeff 
  table, 
  

   the 
  v's 
  are 
  roughly 
  proportional 
  to 
  the 
  squares 
  of 
  the 
  atomic 
  

   weights. 
  Better 
  agreement 
  between 
  the 
  quantities 
  in 
  question 
  

   was 
  found 
  later, 
  by 
  Runge 
  and 
  Precht, 
  to 
  be 
  effected 
  by 
  their 
  

   so-called 
  law 
  which 
  is, 
  that 
  the 
  logarithm 
  of 
  v 
  is 
  a 
  linear 
  func- 
  

   tion 
  of 
  the 
  logarithm 
  of 
  the 
  atomic 
  weight. 
  More 
  recently, 
  the 
  

   pioneer 
  work 
  on 
  characteristic 
  X-rays, 
  by 
  Moseley 
  and 
  others, 
  

   has 
  led 
  to 
  the 
  belief 
  that 
  atomic 
  numbers 
  are 
  more 
  fundamental 
  

   than 
  atomic 
  weights. 
  This 
  being 
  granted, 
  the 
  obvious 
  thing 
  to 
  

   do 
  is 
  to 
  try 
  to 
  find 
  a 
  connection 
  between 
  the 
  frequency 
  difference 
  

   v 
  and 
  the 
  atomic 
  number 
  N. 
  

  

  In 
  this 
  direction, 
  a 
  certain 
  measure 
  of 
  success 
  has 
  been 
  attained 
  

   in 
  a 
  recent 
  paper 
  by 
  H. 
  Bell. 
  As 
  suggested 
  by 
  Moseley 
  's 
  law, 
  

   ■\/v 
  = 
  m(N 
  — 
  JV 
  ), 
  Bell 
  plotted 
  the 
  square 
  root 
  of 
  the 
  frequency 
  

   as 
  ordinate 
  against 
  the 
  atomic 
  number 
  as 
  abscissa, 
  for 
  elements 
  

   of 
  the 
  same 
  Mendelejeff 
  family. 
  In 
  general, 
  the 
  points 
  thus 
  

   plotted 
  fall 
  fairly 
  closely 
  on 
  straight 
  lines. 
  For 
  example, 
  in 
  the 
  

   case 
  of 
  the 
  doublet 
  series, 
  Li, 
  Na, 
  K, 
  Rb, 
  Cs, 
  or 
  Al, 
  Ga, 
  In, 
  or 
  

   N, 
  As, 
  Sb, 
  Bi, 
  etc., 
  satisfy 
  the 
  linear 
  law. 
  Similarly, 
  for 
  the 
  

   triplet 
  frequency 
  intervals, 
  this 
  relation 
  is 
  found 
  to 
  hold 
  between 
  

   Mg, 
  Ca, 
  Sr, 
  Ba, 
  Ra, 
  etc. 
  In 
  two 
  instances, 
  the 
  diagrams 
  show 
  

   that 
  one 
  right 
  line 
  branches 
  from 
  another 
  at 
  a 
  certain 
  element. 
  

   For 
  illustration, 
  at 
  the 
  K 
  point 
  of 
  the 
  Li-Na-K-Rb-Cs 
  line 
  there 
  

   originates 
  a 
  line 
  passing 
  through 
  Cu 
  and 
  Ag. 
  The 
  paper 
  also 
  

   contains 
  the 
  numerical 
  data 
  obtained 
  by 
  substituting 
  in 
  both 
  the 
  

   formula 
  given 
  above 
  and 
  in 
  the 
  logarithmic 
  relation 
  proposed 
  by 
  

   Runge 
  and 
  Precht. 
  On 
  the 
  whole, 
  the 
  advantage 
  seems 
  to 
  favor 
  

   the 
  linear 
  equation. 
  For 
  lack 
  of 
  space, 
  the 
  author's 
  discussion 
  

   of 
  the 
  exceptional 
  cases 
  and 
  of 
  the 
  theoretical 
  aspect 
  of 
  the 
  

   problem 
  will 
  have 
  to 
  be 
  omitted. 
  — 
  Phil. 
  Mag., 
  36, 
  337, 
  1918. 
  

  

  H. 
  s. 
  u. 
  

  

  9. 
  Mirrors, 
  Prisms 
  and 
  Lenses; 
  by 
  James 
  P. 
  C. 
  Southall. 
  

   Pp. 
  xix, 
  579. 
  New 
  York, 
  1918 
  (The 
  Macmillan 
  Co.).— 
  This 
  

   volume 
  is 
  essentially 
  a 
  text-book 
  on 
  geometrical 
  optics. 
  

   Although, 
  of 
  necessity, 
  a 
  part 
  of 
  the 
  ground 
  covered 
  coincides 
  

   with 
  the 
  corresponding 
  portions 
  of 
  the 
  author's 
  earlier 
  volume 
  

   entitled 
  ''The 
  Principles 
  and 
  Methods 
  of 
  Geometrical 
  Optics," 
  

   nevertheless 
  the 
  two 
  books 
  are 
  not 
  coextensive, 
  and 
  they 
  were 
  

   written 
  with 
  entirely 
  different 
  objects 
  in 
  view. 
  Since 
  the 
  present 
  

   book 
  is 
  intended 
  primarily 
  for 
  students 
  whose 
  mathematical 
  

   attainments 
  are 
  relatively 
  small, 
  the 
  analytical 
  developments 
  

   have 
  been 
  made 
  as 
  simple 
  as 
  is 
  consistent 
  with 
  clearness 
  and 
  

   rigor. 
  

  

  