8 



V. G. DETHIER 



I 5rv 



< 

 5 

 a: 

 o 



z 



a 



_i 

 o 

 i 



10 



UJ 

 (E 



I 



075 



I- 

 O 



UJ 



0.50 



z 



< 



Q 



UJ 



025 



4 5 6 7 8 9 10 

 NUMBER OF C ATOMS 



Fig. 5. Relation between the stimulating efficiency of organic salts and the chain 

 length of the anion. 



are expressed as thermodynamic activities rather than as moles, the differences 

 between successive homologues of a series are not so marked, but a plot of the 

 log of these values against the log of activity coefficients (Fig. 10) does not 

 produce a straight line of the sort that one is accustomed to expect from parallel 

 experiments on narcosis (for a complete discussion consult Ferguson, 1951; 

 Brink and Posternak, 1948; Dethier, 1954b). 



An attempt was made to obtain with human beings some data comparable 

 to those just outlined above even though the statement is frequently made that 

 man tastes only those substances which are in aqueous solution. Tests revealed 

 that it was possible to measure taste thresholds for alcohols dissolved in mineral 

 oil (Dethier, 1952). As was the case with insects, man's thresholds for the ho- 

 mologous alcohols in oil are nearly but not exactly identical for each homologue. 

 A plot of the oil threshold/water threshold ratio against the activity coefficient 

 (Fig. 11) indicates that the members seem to fall into two groups, the higher 

 members being in one group and the lower in another. If the threshold ratios 

 are plotted against the oil/water partition coefficients, it can be seen that the 

 two sets of values are not the same (Fig. 12). This lack of complete agreement 

 may be attributed either to the interference of saliva and to the fact that one 

 is working with large populations of mixed receptors, or it may indicate that 

 oil/water partitioning is not the limiting factor. 



