36 
DAIRY  CATTLE  AND  THE  BUTTER  TEST: 
TWENTY  YEARS’  EXPERIENCE, 
By  Ernest  Mathews, 
With  a Report  on  some  Investigations  of  the  Factors  which 
Influence  Churnahility^ 
BY 
W.  F.  CooPEB,  B.A.  (Cantab.),  F.C.S.,  L.  E.  Robinson,  A.R.C.Sc.  (London), 
and  W.  H.  Nuttall,  F.I.C.,  F.C.S., 
from  the  Cooper  Laboratory  for  Economic  Research,  Watford,  Herts. 
I PROPOSE  in  the  following  pages  to  review  shortly  the  butter- 
test  and  milk-test  trials  with  which  I have  been  associated, 
more  or  less,  during  the  past  twenty  years,  and  in  doing  so  I 
shall  endeavour  to  show  that  they  have  been  productive  of 
some  good,  but  in  making  this  last  statement  I would  have  it 
clearly  understood  that  to  the  late  Mr.  John  Frederick  Hall, 
of  Sharcombe,  Wells,  Somerset — the  originator  of  the  butter- 
test  trials — whose  views  on  the  subject  were  far  in  advance  of 
his  time,  the  credit  for  any  good  that  can  be  traced  to  them 
entirely  belongs. 
Another  fact  possibly  known  only  to  me  has  impelled  me 
to  rush  into  print,  that  is,  that  on  those  trials  which  I alone 
have  conducted,  sums  of  money  in  prizes  and  attendant 
expenses  amounting  to  at  least  9,000L  have  been  entrusted  to 
me  to  look  after.  I am  sure,  therefore,  it  will  be  readily 
understood  that  I feel  constrained  to  give  an  account  of  my 
stewardship,  and  to  try  and  show  that  these  large  sums  of 
money  have  not  failed  to  further  the  cause  to  which  they  have 
been  given. 
It  is  not  necessary  to  describe  the  way  in  which  butter 
tests  or  milk-yield  trials  are  carried  out,  as  the  conditions, 
report,  and  tables  are  published  after  each  competition,  those 
for  the  past  year  at  the  R.A.S.E.  Show  at  Gloucester  being 
found  on  page  221.  I would,  however,  refer  specially  to  the 
figures  under  the  heading  “ Butter  Ratio  ” and  to  the  averages 
of  the  various  breeds,  as  from  these  the  most  valuable  informa- 
tion is  to  be  obtained. 
The  butter  ratio  figures  are  got  at  by  dividing  the  weight 
of  milk  by  the  weight  of  butter  obtained,  the  quotient  giving 
the  number  of  pounds  of  milk  used  to  make  one  pound  of 
butter.  If  the  quotient  be  divided  by  10'3  the  weight  of  milk 
will  be  shown  in  gallons. 
The  first  butter-test  trials  were  held  at  the  London  Dairy 
Show  in  1886,  but  until  1889  centrifugal  cream  separators 
were  not  used,  the  cream  being  scalded  on  the  Devonshire 
system,  and  the  butters  made  by  hand. 
