FUTURE ANALYSIS OF MEMORY 301 



holtz is no longer tenable, and that it may have to be 

 substituted by a physiological resonance-theory (6, 7). 

 According to Hermann, we may assume that the 

 nervous end-organs themselves are especially sens- 

 itive for stimuli of a definite period (7). A gen- 

 eralisation of this assumption would lead to an 

 understanding of the above-mentioned fact. The 

 motor organs of the larynx may be considered as 

 resonators, and this would explain why only cer- 

 tain tones cause a dog to howl and why only fric- 

 tion of a certain character i. e. t periods --causes a 

 frog to croak. The fact of the easy transmission of 

 sounds into innervations to the larynx in human 

 beings and parrots or song-birds would depend on 

 the same principle. But in theories of this character 

 we must leave some leeway for the influence of chem- 

 ical processes. The phenomena of correlation which 

 we notice in many animals during the period of heat 

 may be determined by substances circulating in the 

 blood during that period (internal secretion). This 

 may account for the change in the irritability during 

 that period. 



5. Our space-sensations are varieties of three dimen- 

 sions. The main coordinates show a definite relation 

 to the main axes of our body. This leads us to a 

 consideration of the possibility whether certain struct- 

 ural conditions of our body determine the main coor- 

 dinates of our system of space-sensations. Hering 

 has shown that the motor innervations of our eyes 

 may be reduced to three kinds, corresponding to the 



