608 R. J. BRITTEN AND F. T. MCCLURE 
and carriers has been rigorously proved or that the function of an osmotic barrier 
has been demonstrated to be unimportant. All that can be said is that the equations 
derived from the carrier model accurately describe the experimental data. Possibly 
these equations are not unique to that model. Any model which gave essentially the 
same equations would also be satisfactory (and one which gave essentially different 
equations clearly unsatisfactory). 
The permease model might be elaborated to remove some of its difficulties. Part 
of such an elaboration would have to consist of specifying the mechanism by which 
asymetry in reaction constants on the two sides of the barrier is obtained. This 
mechanism might introduce additional features which would modify greatly the 
properties of the permease model. We must ask whether any mechanism can be 
proposed which has the necessary properties, if the pool is assumed to be free in 
solution. The minimal requirements are as follows: rapid pool formation; slow loss 
in absence of glucose and/or amino acid; increase of pool with external concentration 
until saturation is reached; lack of proportionality between formation rate and pool 
size. If an exhaustive search does not yield a model with these properties, then there 
would be no reasonable alternative but to assume that the amino acid pool is not 
in free solution within the cell. In this connection it should be remembered that 
experimental data conclusively demonstrate that the pool has at least two compo- 
nents, so that if we are to assume the pool is unbound amino acid we must introduce 
at least one more osmotic barrier. Possibly it would be necessary to resort to a 
combination of a site and barrier. 
Of course, there are many problems which deserve further investigation. The 
meaning of the term “free in solution” needs to be examined both experimentally 
and theoretically. A 25°% solution of protein, RNA, etc., organized in subtle ways, 
is certainly an unusual solvent from a chemical point of view. The activity of the 
amino acids might be strikingly depressed. The results of such a study would have 
very broad implications for other processes in living cells. 
Finally, a more detailed experimental study of the rates of loss and exchange is 
needed. The study of these phenomena with pools of several amino acids and other 
compounds, such as galactosides, would supply quantitative information that might 
be helpful in deciding among the alternative mechanisms. 
With present knowledge, alternative interpretations of the concentration process 
in bacteria still remain possible, in spite of the large amount of experimental evidence 
which has given insight into many aspects of the process. The simplicity and the 
degree to which individual steps may be understood, from a chemical point of view, 
differ among the various models. It is for the future to decide which of the alternative 
approaches will be most useful. 
ACKNOWLEDGEMENTS 
In writing this article we have used freely the experiments and concepts of our 
colleagues at the Department of Terrestrial Magnetism, especially R. B. ROBERTS, 
D. B. Cowre and E. T. Botton. They have also liberally given guidance and advice. 
We are much indebted to the wife (B.H.B.) of one of us for assistance with the 
typing and editing. 
References p. 609 
