70 DISCUSSION 



Davies: No, we don't lose all the activity by any means, and what we 

 lose may only reflect the poor dispersion of the freeze-dried product on 

 resuspension. With regard to ultrasonics, it seems possible that you 

 could lose a great deal of activity without detecting the loss by your 

 test. In my test a one-tube difference in titre is a 50 per cent change in 

 activity, and from successive ultrasonic treatrnxnts, sampling after each, 

 I can plot a curve for loss of activity. However, the ultrasonic effect 

 maybe only a feature of the test system and is not necessarily an intrinsic 

 loss. 



As for the freeze-drying : Dr. Kandutsch, you can freeze-dry your 

 preparations and they retain activity, don't they ? 



Kandutsch: Yes. 



Davies: And if you extract with solvents the activity is also retained ? 



Kandutsch: This is not so clear-cut. With some solvents under some 

 conditions, activity is retained. I haven't investigated enough solvents 

 or conditions to make a definite statement on this. 



Davies: And your materials, Prof. Medawar? With some solvents 

 you retain some activity ? 



Medawar: Yes, after extraction with hpid solvents we get this residue 

 activity which we suspect is not H-2 activity at all. 



Davies: On extracting our material with ether at — 1 5° we can get back 

 a residue that retains some activity ; it is a lot less active, but then it is 

 also a lot less dispersible. When we extract with chloroform-methanol, 

 also in the cold, activity is lost. 



Amos: Do you get precipitin activity with isoantibodies with this 

 material ? 



Davies: I thought it would be most unlikely, and we haven't 

 tried. It will precipitate with rabbit antisera but I take this to be due to 

 association of H-2 with the species-specific "antigen 3". 



Amos: It is usually said that mouse antibodies do not precipitate, 

 but they can with the right sort of antigen. Mouse anti-mouse thyroid 

 will give good precipitin reactions. 



Another point concerning absorption of anti-E, you can't relate the 

 anti-EK to anti-E. This must be non-specific absorption, because the 

 anti-K should still have given you agglutination. You had two ex- 

 amples, one where you thought this might be absorption with F and one 

 withE. In both of these you had an antibody with a double specificity; 



