152 Discussion 



observed that rats offered gradually increasing quantities of potassium in 

 their diet ate and absorbed the relatively very large quantities needed to 

 produce a lethal degree of potassium intoxication. They did not develop 

 diarrhoea, nor did they vomit; they just became weak and died. Like- 

 wise, we have seen a patient with marked limitation in tolerance for 

 potassium due to advanced pan-nephritis become fatally intoxicated 

 with potassium as a result of drinking fruit juices. 



Adolph: The study of tolerances is a very important aspect of the 

 general physiology of regulatory processes. Dr. Talbot, you estimated 

 tolerances in terms of single constituents, but in some of the situations 

 you described, such as the intravenous administrations, you were con- 

 cerned with several constituents at a time. Now when there is depletion 

 or excess of more than one constituent at a time the picture is very differ- 

 ent with respect to tolerance. For instance, there is a great difference 

 between taking pure salt and taking an isotonic solution of salt. I recog- 

 nize that this work is exploratory and that you are making your esti- 

 mates in the simplest way possible when you consider one component at 

 a time, but eventually I hope we shall have some estimates of tolerance 

 to multiple components. 



This consideration of components seems to me to extend also to your 

 studies of composition, Prof. Wallace. If you went to your statisticians 

 still more often, would you not get into the study of multiple correlations 

 which would get us further than comparisons made two at a time? 



Wallace: We have made a number of statistical multiple correlations. 

 It is often difficult to know just what they mean, once certain correlations 

 become evident. Our biggest problem has been to have any assurance as 

 to the proper parameter to which to refer growth. Should the reference 

 basis be body weight, fat-free weight, protein, ash, or water? 



Adolph : What I want to bring out is that an organism probably has 

 some way of measuring the bodily composition which is very much more 

 complicated than saying, for instance, that magnesium is the fixed con- 

 stituent around which all others revolve. I think that without a study of 

 multiple correlations we will never be able to find whether there is a key 

 fixity by which homeostasis is guided to a definite volume and concentra- 

 tion to which the organism always returns. I do not know whether 

 any of our methods of representing homeostasis will be so similar to that 

 of the organism that we can predict what it does to get back to its fixity. 



I should also like to remark on Dr. Talbot's choice of a key variable. 

 No doubt he has great reservations about the use of this term. What he is 

 trying to do, I gather, is to out-guess the organism as to what it is using 

 as a measuring stick by which it will return to its original composition, or 

 by which it will estimate what has to be done in order to defend itself 

 against disturbances. When we think that an organism is restoring its 

 potassium concentration, have we any assurance that that one restora- 

 tion is a prime objective in the adjustments which are going on? 



Talbot : We agree with you that most if not all of the variables under 

 consideration are related to each other. For instance it is known that 

 body tolerance for potassium is impaired under conditions of zero sodium 

 intake and that tolerance for sodium is abnormally limited under condi- 



