﻿55 
  

  

  Tim 
  the 
  slightest 
  risk 
  of 
  h)singa 
  single 
  eg-g 
  from 
  leakage, 
  

   it 
  oc<^nirred 
  to 
  me 
  that 
  tlie 
  proper 
  thing 
  to 
  do 
  would 
  l>e 
  to 
  

   construct 
  a 
  box 
  in 
  wliich 
  the 
  water 
  passed 
  throngh 
  a 
  

   membrane 
  of 
  blotting 
  or 
  filtering 
  paper 
  into 
  the 
  hatching- 
  

   chamber 
  and 
  out 
  of 
  this 
  again 
  tlirongh 
  a 
  similar 
  mem- 
  

   brane 
  into 
  another 
  chamber 
  and 
  allowed 
  to 
  escape. 
  In 
  

   some 
  snch 
  way 
  it 
  will 
  be 
  possible 
  to 
  hat(di 
  the 
  minutest 
  

   eggs 
  and 
  w^atch 
  the 
  reprodnction, 
  divisicm, 
  etc., 
  of 
  the 
  

   very 
  minutest 
  monads. 
  The 
  idea 
  is 
  jjossibly 
  one 
  of 
  im- 
  

   mense 
  X3ractical 
  valne 
  in 
  oyster-cnlture 
  and 
  opens 
  np 
  a 
  

   field 
  of 
  work 
  in 
  embryology 
  in 
  thepnrsnit 
  of 
  which 
  many 
  

   conqnests 
  are 
  still 
  to 
  be 
  made. 
  

  

  Another 
  form 
  of 
  appai'atns 
  Avliich 
  suggested 
  itself 
  to 
  

   me 
  during 
  my 
  experiments 
  with 
  the 
  device 
  Jnst 
  described, 
  

   has 
  the 
  advantage 
  of 
  being 
  easily 
  02:)erated 
  by 
  some 
  sim- 
  

   ple 
  form 
  of 
  pump, 
  by 
  means 
  of 
  which 
  the 
  water 
  may 
  be 
  

   constantly 
  renewed 
  in 
  the 
  hatching 
  chamber. 
  A 
  metal 
  

   ring 
  with 
  a 
  bibulous 
  membrane 
  on 
  either 
  face, 
  made 
  

   water-tight 
  around 
  the 
  edges 
  of 
  the 
  ring, 
  so 
  as 
  to 
  enclose 
  

   a 
  space 
  into 
  which 
  a 
  short 
  piece 
  of 
  tnbing 
  may 
  be 
  screwed 
  

   at 
  one 
  side 
  and 
  connected 
  with 
  a 
  pnmp 
  by 
  a 
  rubbei* 
  tube. 
  

   In 
  this 
  way, 
  by 
  pumping, 
  water 
  would 
  be 
  drawn 
  into 
  the 
  

   space 
  included 
  by 
  the 
  ring 
  and 
  membranes, 
  provided 
  

   these 
  were 
  immersed 
  in 
  the 
  water 
  contained 
  in 
  the 
  batch- 
  

   ing 
  chamber. 
  Of 
  course, 
  the 
  mem])ranes 
  would 
  have 
  to 
  

   be 
  strengthened 
  both 
  inside 
  and 
  outside 
  by 
  a 
  single 
  

   thickness 
  of 
  stout 
  <'anvas, 
  the 
  same 
  as 
  I 
  used 
  in 
  milking 
  

   my 
  experimental 
  boxes, 
  so 
  as 
  to 
  have 
  the 
  bibulous 
  mem- 
  

   brane 
  between 
  the 
  sheets 
  of 
  canvas 
  to 
  avoid 
  rupturing 
  

   the 
  former, 
  by 
  pressure 
  or 
  accident. 
  The 
  diameter 
  of 
  

   the 
  ring 
  and 
  membranes 
  might 
  be 
  six 
  inches, 
  as 
  more 
  

   than 
  that 
  would 
  probably 
  render 
  the 
  membranes 
  liable 
  to 
  

   collapse 
  when 
  the 
  juimp 
  was 
  in 
  operation. 
  'I'he 
  amount 
  

   of 
  water 
  passing 
  through 
  the 
  apparatus 
  I 
  should 
  tliink 
  

   would 
  not 
  need 
  to 
  be 
  more 
  than 
  what 
  would 
  comj)letely 
  

   change 
  the 
  water 
  in 
  the 
  hatching 
  chamber 
  once 
  in 
  every 
  

   half 
  or 
  one 
  hour, 
  so 
  that 
  the 
  ]iump 
  uecessnry 
  would 
  be 
  

  

  