APPENDIX I. ON METHODS 



ESTIMATION OF THE OXYGEN CAPACITY OF BLOOD BY THE 

 FERRICYANIDE METHOD OF ESTIMATING OXYGEN IN BLOOD 



THE simplest measurement which it is possible to make of the oxygen in blood is 

 that of the total oxygen capacity of the haemoglobin. For this measurement it is 

 necessary to be careful about three things quite apart from apparatus. 



(1) The blood must be fresh so that it does not reduce itself as the result of 

 bacterial action. The length of time which may elapse from the time at which the 

 blood is drawn till the estimation is made depends upon the cleanliness with which 

 the fluid is collected and the temperature at which it is kept. In ordinary weather 

 blood taken by pricking the finger should be estimated within 24 hours. Blood may 

 be kept for some two or three days by placing it in a stoppered tube which is kept 

 in ice in a Dewar's flask. It cannot be kept indefinitely even at C. 



(2) The blood must be thoroughly shaken with air in order that it may be 

 fully saturated with oxygen. At room temperature the degree of saturation in air 

 is not measurably less than 100 % 



(3) The blood must be shaken just before it is used in order that no sedimenta- 

 tion may take place. 



The theory of the ferri cyanide method is obscure, the blood in faintly alkaline 

 solution gives off accurately the quantity of oxygen which could be abstracted from 

 the haemoglobin with a blood gas pump. Nevertheless the substance formed by the 

 reaction is methaemoglobin, a body which is credited with containing the same 

 quantity of oxygen as oxyhaemoglobin, though in a stable form. 



The following equation is given by Haldane for the reaction. The alkali in the 

 formula is sodium bicarbonate. The general nature of the formula is the same no 

 doubt if ammonia or other alkali be substituted. 



| + 4Na 3 (FcCv 6 )+4NaHC0 3 = O, + Hbf + 4Na 4 (FcCy 6 

 X) V) 



Haldane's demonstration of the accuracy of the method was made with an 

 ordinary Dupre's apparatus for the estimation of urea. 50 c.c. of blood and 100 c.c. 

 of dilute alkali were mixed in the bottle of the apparatus ; into the small tube in the 

 interior of the bottle were placed 20 c.c. of a saturated solution of ferricyanide of 

 potassium. The blood and the alkali were thoroughly mixed in order to allow of 

 complete laking of the corpuscles. The rest of the operation was conducted just 



