The blood of different species 



49 



a slight tendency for the actual points obtained from blood to diverge 

 a little from the curve obtained from the haemoglobin solution. That 

 is just where the curve bends round at the top. To this we must add 

 that more recently the curve of the same human subject, Dr Camis him- 

 self, has been redetermined at Pisa by more exact methods than were 



TOO 



110 



90 



80 



70 



60 



50 



40 



30 



20 



10 



ffi 



L/L 



- o/ 



tr 



LU 



10 



20 



' 



30 



40 50 60 70 80 90 100 



FIG. 26. Ordinate = percentage saturation of haemoglobin. Abscissa = tension of 2 in 

 mm. of mercury. Dissociation curve of dog's blood at 40 mm. tension C0 . Dotted 

 line = dissociation curve of human blood (see Fig. 25). = determination for dog's 

 blood. o = ditto for haemoglobin solution with salts of dog's red blood corpuscle. 

 The area drawn round the point is the experimental error in each case. Temperature 

 37_38 C. 



then at our disposal. These methods enabled us to work accurately 

 at pressures of oxygen under 10 mm. The Pisa curve seems to differ 

 from that given above to a trivial though appreciable extent. The 

 difference, however, is quite negligible as compared with the dif- 

 ferences which exist between his blood and that of a dog, and it is 



B. R. F. 4 



