92 THE PHYSIOLOGY OF INSECT SENSES 



combined to form more complex arrangements, and this in turn 

 further multipHed. 



The tympanic organ is sensitive to the entire frequency range of 

 sounds produced by stridulation, but regardless of the frequency, the 

 pattern of discharge in the tympanic nerve is asynchronous. From this 

 it has been concluded that the organ is incapable of frequency discrim- 

 ination, and this may indeed be so. On the other hand, there have been 

 no recordings from single fibres in Orthoptera, and this should be done 

 to clinch the argument, even though Pumphrey (1940) has argued on 

 structural grounds that harmonic analysis, that is, different fibres 

 responding to different stimulus frequencies, is impossible. The fact 

 that no frequency discrimination has been observed in the moth 

 tympanic organ, where the presence of only two fibres permits a precise 

 analysis (Roeder and Treat, 1957), adds weight to the argument. It 

 also supports the idea that there is no frequency pattern in the nerve 

 mirroring the frequency pattern of the stimulus. On the other hand, 

 it is possible that the insect itself may be able to analyse frequency to 

 some degree by means of central interpretation of input from a 

 number of organs each having a different frequency range (Katsuki 

 and Suga, 1960; Suga and Katsuki, 1961). 



Pumphrey and Rawdon-Smith (1939) proposed that the tympanic 

 organ, being sensitive to intensity, detected changes in intensity, that 

 is, amplitude of the sound wave, and that the different sounds of 

 insects differed most importantly by the periodic changes (modulation) 

 in intensity of the sound. In other words, any sound wave of a fre- 

 quency within the detectable frequency range was merely a carrier 

 wave upon which was imposed a pattern of amplitude modulation. 

 By contrast, the human ear is very sensitive to frequency changes in 

 the carrier wave but relatively insensitive to changes of the modulation 

 frequency. This hypothesis was supported by records showing that an 

 organ stimulated by an amplitude-modulated sound discharged 

 volleys of impulses at the modulation frequency of the stimulating 

 sound. These results were confirmed in a number of species by Haskell 

 (1956 b), and as he pointed out, it would be reasonable to assume that 

 the different pulse rates found in the different songs of various species 

 would be the parameter signalled by the receptors, and hence the key 

 to the recognition of songs. Busnel (1955) and his co-workers (Busnel, 

 Loher, and Pasquinelly, 1954), on the other hand, have suggested as a 

 result of experiments in which natural and artificial songs were played 

 to species of Chorthippus that the essential element in songs that aids 

 discrimination is 'transients' (sudden changes in intensity). To test the 



