QUANTITY AND COMPOSITION OF BLOOD 967 



In applying this method in cases of disease it is important not to give 

 too large a dose of carbonic oxide gas. In a normal individual 30 per 

 cent, of the hemoglobin may be combined with carbon monoxide 

 before any oxygen hunger is felt, and it is possible to saturate half the 

 haemoglobin with this gas, though with considerable discomfort to the 

 individual. In cases such as heart disease, where the patient is at 

 the very margin of his resources, even 30 per cent, diminution of the 

 oxygen capacity of the blood may have serious results, and the carbon 

 monoxide inspired must be therefore kept at the lowest limit at which 

 it is possible to carry out a reliable determination of the relative 

 carbonic oxide saturation of the blood sample. 



The total blood volume probably varies appreciably with altera- 

 tions in the conditions of the animal, and may be found different on 

 two succeeding days. It is certainly influenced by the height of the 

 blood pressure as well as by the oxygen tension in the air breathed, and 

 therefore alters with the altitude. Some of these variations we shall 

 have to consider more fully in a later section. Any lowering of blood 

 pressure causes an absorption of fluid from the tissues into the blood, 

 so that the latter becomes more dilute. The blood content during 

 the last stages of bleeding may contain little more than 50 or 60 per 

 cent, of the haemoglobin which was present in the first samples of blood, 

 pointing to a corresponding dilution of the blood during these few 

 minutes by means of tissue lymph. By this means, i.e. the absorption 

 of fluid from tissues, the volume of circulating blood after a limited 

 haemorrhage is rapidly brought up to normal, so that there is a circula- 

 tion of a fluid impoverished in corpuscles. The latter are made up 

 in the course of a few weeks as a result of increased activity in the 

 bone-marrow. 



RELATIVE AMOUNT OF PLASMA AND CORPUSCLES 

 The relative amount of corpuscles in a given sample of blood is 

 most easily determined by Blix's method. The blood is mixed with 

 a definite amount of 2J potassium bichromate, and the mixture is put 

 into small graduated capillary tubes, which are then placed in a cen- 

 trifuge revolving about 10,000 times per minute. The corpuscles 

 rapidly accumulate in an almost solid mass at the bottom of the tube, 

 and their volume can be directly read off. It is often possible by work- 

 ing quickly to receive blood into such graduated capillary tubes and 

 to centrifuge it rapidly before it has had time to coagulate. The 

 corpuscles are hurried down to the bottom of the tube within two or 

 three minutes and their volume can be in this way directly determined. 

 An indirect method for the same purpose was devised by Hoppe- 

 Seyler. The total proteins of defibrinated blood are determined and 

 compared with the total proteins of the washed corpuscles and of 



