﻿No.im. 
  rHTLIPnXE 
  OnXITHOLOGY—WOBCESTER 
  AND 
  BOVRNS. 
  611 
  

  

  both 
  islands 
  are 
  fairly 
  well 
  known; 
  the 
  curve 
  rises 
  for 
  these 
  islands. 
  

   Marinduqne 
  marks 
  the 
  lowest 
  remaining- 
  point. 
  The 
  iskind 
  is 
  known 
  

   only 
  from 
  the 
  collections 
  of 
  the 
  Steere 
  Expedition, 
  and 
  in 
  making 
  them 
  

   we 
  had 
  to 
  tramp 
  miles 
  from 
  the 
  village 
  where 
  we 
  had 
  headquarters 
  in 
  

   order 
  to 
  get 
  into 
  scattering 
  forest 
  hardly 
  worthy 
  of 
  the 
  name. 
  

  

  It 
  will 
  be 
  noted 
  that 
  the 
  latter 
  part 
  of 
  our 
  curve, 
  where 
  we 
  are 
  deal- 
  

   ing" 
  with 
  large 
  islands, 
  each 
  of 
  which 
  has 
  some 
  virgin 
  forest 
  remaining:, 
  

   is 
  comi)aratively 
  regular. 
  

  

  Attention 
  should 
  be 
  called 
  to 
  another 
  important 
  fact. 
  By 
  reference 
  

   to 
  the 
  footings 
  of 
  Table 
  B 
  it 
  will 
  be 
  seen 
  that 
  not 
  only 
  does 
  the 
  number 
  

   of 
  genera 
  showing 
  two 
  or 
  more 
  species 
  reach 
  its 
  maximum 
  in 
  the 
  largest 
  

   islands, 
  but 
  the 
  number 
  of 
  species 
  into 
  which 
  genera 
  are 
  differentiated 
  

   reaches 
  its 
  maximum 
  as 
  well. 
  Luzon 
  leads 
  with 
  six 
  species 
  of 
  JJicaeuni 
  

   and 
  live 
  of 
  Cinnyris, 
  while 
  we 
  have 
  six 
  genera 
  with 
  four 
  species 
  each, 
  

   and 
  eleven 
  with 
  three. 
  Mindanao 
  follows, 
  having 
  one 
  genus 
  with 
  five 
  

   species 
  and 
  eight 
  with 
  three. 
  I 
  know 
  of 
  no 
  simple 
  means 
  by 
  which 
  

   this 
  factor 
  could 
  be 
  introduced 
  into 
  the 
  curve, 
  Dut 
  its 
  significance 
  should 
  

   not 
  be 
  lost 
  sight 
  of. 
  

  

  It 
  seems 
  to 
  me 
  that 
  the 
  facts 
  above 
  stated 
  justify 
  the 
  conclusion 
  that 
  

   in 
  the. 
  Philippines 
  the 
  larger 
  the 
  island 
  and 
  the 
  greater 
  the 
  diversity 
  of 
  

   its 
  surface, 
  the 
  larger 
  tlie 
  percentage 
  of 
  genera 
  represented 
  by 
  more 
  

   tlian 
  one 
  species, 
  and 
  the 
  larger 
  the 
  average 
  number 
  of 
  species 
  into 
  

   which 
  they 
  are 
  differentiated. 
  

  

  It 
  may 
  be 
  objected 
  that 
  we 
  also 
  find 
  the 
  largest 
  number 
  of 
  genera 
  

   distributed 
  in 
  accordance 
  with 
  Steere's 
  law, 
  i. 
  e., 
  with 
  but 
  one 
  species 
  

   in 
  a 
  place, 
  in 
  the 
  largest 
  islands. 
  This 
  is 
  true, 
  Mindanao 
  leading 
  with 
  

   32, 
  followed 
  by 
  Luzon 
  with 
  1*8, 
  Samar 
  with 
  27, 
  and 
  so 
  on. 
  It 
  should 
  be 
  

   remembered, 
  however, 
  that 
  we 
  are 
  dealing 
  here 
  with 
  relative, 
  not 
  abso- 
  

   lute 
  quantities, 
  and 
  the 
  real 
  question 
  is 
  not 
  whether 
  the 
  absolute 
  num- 
  

   ber 
  of 
  confirmations 
  of 
  Steere's 
  law 
  is 
  higher 
  for 
  these 
  islands, 
  but 
  

   whether 
  it 
  is 
  proportionately 
  higher. 
  It 
  can 
  be 
  readily 
  shown 
  to 
  be 
  

   proportionately 
  lower. 
  

  

  In 
  constructing 
  Curve 
  II, 
  to 
  illustrate 
  this 
  point, 
  I 
  have 
  used 
  the 
  same 
  

   abscissa 
  and 
  ordinate 
  as 
  for 
  Curve 
  I, 
  vertical 
  units 
  indicating 
  jjerceut- 
  

   ages, 
  and 
  horizontal 
  units 
  areas. 
  In 
  computing 
  percentages 
  for 
  each 
  

   island 
  I 
  have 
  taken 
  the 
  total 
  number 
  of 
  confirmations 
  of 
  Steere's 
  law 
  

   from 
  Table 
  A, 
  and 
  added 
  it 
  to 
  the 
  total 
  number 
  of 
  exceptions 
  shown 
  in 
  

   Table 
  B. 
  This 
  gives 
  the 
  total 
  number 
  of 
  genera 
  affording 
  evidence 
  in 
  

   each 
  case, 
  and 
  it 
  is 
  an 
  easy 
  mat,ter 
  to 
  ascertain 
  what 
  percentage 
  of 
  this 
  

   total 
  is 
  distributed 
  in 
  the 
  one 
  way, 
  and 
  what 
  in 
  the 
  other. 
  

  

  In 
  Curve 
  II, 
  as 
  in 
  the 
  curves 
  that 
  follow 
  it, 
  a 
  solid 
  line 
  is 
  used 
  to 
  

   represent 
  confirmations 
  of 
  Steere's 
  law, 
  and 
  a 
  broken 
  line 
  to 
  indicate 
  

   exceptions 
  to 
  it. 
  

  

  For 
  reasons 
  already 
  stated 
  the 
  curve 
  is 
  irregular 
  for 
  the 
  smaller 
  and 
  

   less 
  well 
  known 
  islands, 
  but 
  its 
  general 
  features 
  are 
  apparent 
  even 
  

   here. 
  The 
  solid 
  line 
  shows 
  a 
  constant 
  tendency 
  to 
  return 
  to 
  the 
  100 
  

   per 
  cent 
  mark, 
  the 
  broken 
  line 
  to 
  fall 
  to 
  zero. 
  

  

  