READJUSTMENTS OF REGULATION . 31 
otherwise needed. Excess of CO,, on the other hand, 
facilitates the dissociation, so that the giving off of 
CO, to the blood in the body capillaries helps to make 
the curve steeper and so facilitates the oxygen supply 
to the tissues. 
The curve is not at all of the shape which would be 
expected on purely chemical grounds from what is 
known of other substances which dissociate in a similar 
manner. It was discovered by Barcroft and his pupils 
that the inorganic salts present along with the haemo- 
globin in the red corpuscles determine this peculiar 
form. When the haemoglobin is freed from these 
salts its dissociation curve has the form which would 
have been expected on chemical grounds—namely, that 
of a rectangular hyperbola. With this form of curve 
the oxyhaemoglobin would be wholly unsuited for 
performing the work which it actually performs in 
the body. The action of the salts is almost certainly 
connected with their power of causing the haemoglo- 
bin molecules to become aggregated into groups. Bar- 
croft also found that it is in virtue of its action as 
an acid when in solution that CO, affects the dissocia- 
tion curve. Alkalies shift the curve to the left, while 
acids shift it to the right; and the changing position 
of the curve is an extraordinarily delicate index of 
small changes in the reaction of the blood. 
Both the plasma and the corpuscles of blood contain 
substances which enter into chemical combination with 
CO,; and these combinations dissociate with fall in 
the pressure of CO,, and re-form with rise, just as 
oxyhaemoglobin dissociates and re-forms. The whole 
