88 HEMOGLOBIN 



equilibrium is therefore maintained between two forces acting on 

 the gases of the atmosphere, one force is that of gravity, which 

 attracts the air to the centre of the earth, the other is that of diffusion, 

 which tends to dissipate the air uniformly over the universe. Now 

 the mathematical expression of the velocity of diffusion contains 

 the molecular weight of the gas, that of the influence of gravity on 

 it does not. An equilibrium between the two may therefore be 

 represented by an equation in which the molecular weight appears 

 on one side, and one side only ; therefore if all the other quantities 

 in the equation are known, the molecular weight of the gas may 

 be calculated. Nor was Svedberg's the first important application 

 of this principle. By invoking its aid Perrin performed his classical 

 research on the sedimentation of gamboge particles into layers of 

 increasing density from an originally uniform solution. 



Haemoglobin molecules will not sediment, some force much more 

 powerful than gravity being required to divert their path in a uniform 

 direction. At very high speeds of rotation, however, centrifugal 

 force was made by Svedberg to effect this object. 



The theory then put simply is as follows: If a solution of haemo- 

 globin is centrifuged with sufficient rapidity it will cease to be homo- 

 geneous and the molecules of haemoglobin will separate out as particles 

 would do in an ordinary solution. Normally the solution is kept 

 homogeneous because the forces of diffusion overpower aU others, 

 and the rate at which the haemoglobin separates when centrifuged 

 depends upon the degree to which centrifugal force can overcome 

 diffusion. Now, of these two forces which can be brought into 

 equilibrium, one, the rate of sedimentation of the haemoglobin, does 

 not depend upon the molecular weight, the other, the rate of diffusion, 

 does. Therefore balancing the one against the other we get an 

 equation which contains the molecular weight on the one side and 

 not on the other. Clearly if we can measure aU the other things 

 in the equation we can calculate the molecular weight^. Fortunately 

 such things as do not cancel out can be measured. They are : 



( 1 ) The concentrations (Cj and Cg) of the solution at two points, these 

 points being at known distances (x^ and X2) from the centre of rotation, 



(2) The absolute temperature (T). 



(3) The speed of the centrifuge, giving the angular velocity (a>). 



' The formula for finding the molecular weight M is 



2RT {log, c^/c^) 



M = 



«« (1 - Vp) (a^i - Xj) (a^ + Xj)" 



