196 Chapter XIII 



of gas measured to the minimum quantity consistent with the preser- 

 vation of the measurement being a simple measurement of length. 

 This of course is accomplished by narrowing the bore of the burette, 

 and as it appears to be as easy to draw a capillary tube of uniform 

 bore as a more capacious cylinder, the real limit of the process 

 is arrived at where serious errors are introduced by unavoidable 

 inequalities in the surface of the glass. Moreover, gas in capillary 

 tubes is in many ways easy to manipulate ; if enclosed between two 

 fluid surfaces there is not the same tendency for the gas to escape. 

 You can lay the tube down, set it on its end, expel the bubble of 

 gas, withdraw it, manipulate it in all sorts of simple ways, and still 

 you have it, and have it under conditions in which it is, so to speak, 

 all surface. 



Whilst in this country Brodie and I were pushing these principles 

 to what we regarded as considerable lengths for the purpose of 

 determining the gaseous exchange of the frog's kidney and found to 

 our own surprise that we could analyse 1/10 c.c. of gas with sufficient 

 accuracy, Krogh was working out a technique along the same 

 general lines, by the side of which ours appears crude to the last 

 degree. Nothing could be more beautiful or fascinating than his 

 method of analysing almost microscopic bubbles of gas, that is to say, 

 bubbles of about 1/100 c.c. This technique reformed aerotonometric 

 methods. A bubble of such small dimensions rapidly attains equi- 

 librium with any liquid with which it is in contact. If placed in 

 a stream of circulating blood, it is tossed to and fro, its surface and 

 that of the blood in contact with it are constantly changing, and the 

 stream in which it is may be so small that each corpuscle is out of 

 the animal but a very short time. Therefore there is no opportunity 

 for self-reduction of the blood, and moreover there is no amount of 

 haemorrhage which signifies. Krogh's method at once received the 

 acclamation of physiologists, partly because of its own beauty and 

 partly on account of the consciousness of its limitations which its 

 inventor betrayed. In a series of papers which are at once a model 

 of certainty of touch and modesty in presentation, he has revised 

 the whole field of the relation of oxygen pressures in the blood and 

 the alveolar air, in so far as it can be revised by methods of this 

 kind. 



Krogh investigated the criticism levelled against aerotonometric 

 experiments generally, that their results yield too low an oxygen 

 tension owing to self-reduction of the blood. He found that self- 



