THE SENSE OF TASTE 



527 



value to match .34 sucrose. There is no supplemental 

 action by this computation. The apparent supple- 

 mental action with one set of transformations but not 

 the other is due to the attempt to add arithmetically 

 one linear to one nonlinear function. This example, 

 although derived theoretically from the electrophysio- 

 logical response curves, can be matched almost 

 exactly by empirical data from psychophysical ex- 

 periments on man (44). Further study of taste mixtures 

 by the electrophysiological method is desirable. The 

 additive analysis presented in figure 16 is theoretical 

 except that the response curves for the individual 

 sugars are based upon experimental points. 



BEH.^VIOR.^L EFFECTS 



FIG. 16. Response of rat chorda tympani nerve to different 

 concentrations of different sugars. The dashed line SUC. & 

 GLUC is the summated response to be e.xpected when sucrose 

 solutions are added to 0.2 M glucose. See text for discussion. 

 (From Hagstrom, E. C, unpublished observations.) 



M sucrose solutions in a mixture can be made from 

 figure 16 by adding the sucrose curve to the response 

 of 0.2 M glucose at A, so obtaining the dotted line. 

 The total sensory effect of the mixture should be the 

 sum of the two functions at point B. B equals 62 

 units, a magnitude of nerve discharge that could be 

 produced either by .34 sucro.se or .94 glucose, indi- 

 vidually. The empirical match to the original mixture 

 can be stated as: 



.2G + .2S = .34S 



CO 



In sucrose equisweet solutions where .2G = .04.S, the 

 equation (/) becomes 



.04S -|- .2S = (sucrose match) 



C^). 



But since the arithmetic sum of .04S -\- .2S is .24 

 and not .34 (a difference of .10) there is supplemental 

 action. That is, the empirical match shows a stronger 

 sucrose concentration than could be predicted by 

 the simple addition of equisweet sucrose solutions 

 (i.e. the mixture is sweeter). Setting equation (/) in 

 terms of glucose where .2S = .74G we have: 



.2G -(- .74G = .94G (and .94G = .34S). 



The arithmetic sum of .2G -|- .74G equals exactly 

 .94G which is the same as the equivalent glucose 



Taste stimulation is most directly related to food 

 taking and the rejection or avoidance of noxious 

 stimulation. Hence, the manifold of four basic tastes 

 of salt, sour, bitter and sweet can be reduced to two 

 behavioral classes, acceptance 01 rejection. Certain 

 substances are rejected in all concentrations; others 

 may be accepted at low but rejected at high concen- 

 trations; still others may be accepted at all concen- 

 trations. Acceptance or rejection may also be in- 

 fluenced by postingestion effects, by the metabolic 

 condition of the organism and by past conditioning 

 or learning. Thus taste is only one of several determi- 

 nants of appetitive behavior. 



Richter (177) in his classical studies of self-selection 

 demonstrated that the animal's behavior was a part 

 of "'the total adjustive mechanism working toward the 

 constancy of the internal en\ironment. " The adrena- 

 lectomized rat will increase the intake of salt solution 

 to such a degree that it not only survives but gains 

 weight. It is significant that the albino rat under 

 normal conditions, i.e. when not salt hungry or 

 hormone deficient, displays a striking preference for 

 sodium chloride and other sodium salts (15, 176, 203). 

 This behavior is exaggerated after adrenalectomy as 

 evidenced by a lowered preference threshold and 

 greater intake of all salt solutions above threshold 

 in preference to water. An excess of .salt in the diet 

 reduces the preference and may even wipe it out, 

 leaving only an aversion over the entire stimulus 

 range. Figure 1 7 summarizes these behavioral phe- 

 nomena together with a plot of the magnitude of the 

 afferent nerve discharge of the chorda tsmpani nerve 

 in the rat. The solid line curves show the relative 

 preference in per cent (cc salt 'cc salt -|- cc H2O) 

 under four conditions: adrenalectomized, normal (jV), 



