an area, but the dog dopen4a very lar1ely upon one single sne 
to  eoate Comee, althouh se..l      other faotorwi=t be 
conidre, which determine to some extent the ditfferene be- 
teen a       4og     a poor dog. Fortunately this remarkakbl 
charaterisati  to fin  gae has been develope  to suoh a high 
degree In the bid dog and over such a I    period that It 
sees probable that the relative ability to fin bira will be 
lees variable ia a bird dog than in ma , an It appears that 
the bird do will provide a relatively constant and very useful 
instrmet in measuring game bird abundance* 
          It seems Uioal to a4voeee this mnthod as a 
standard'ieL means of obtaining qantitiative information on 
the ae of upland e          birds, which oan be expressed in 
omparable terms* In order that the samplig nay be rpresenta- 
tive of the variouas types, it is advisable that a oovar map be 
prepared in advance of th  ens's work. Fortunatly N orthera 
WiohI~an is rapidly beingi provided such maps by the Land 
Eonomio Srvey, an    It is hoped that eventually Suuthern Yichian 
will be oever-mappe. When thtse become available the problem 
of sampling and fil interpretation of the results of a census 
for uplan game birds will be greatly faecilitated.