496 



Comparative Anhnal Physiology 



songs, and an intermediate period in which the chirps from the two partners 

 alternated in a regular manner. Regen -^^ then tried to substitute an artificial 

 partner. When he used males which had just become adult he was able to 

 get them to alternate with a wide variety of musical sounds of such a nature 

 as to indicate a sound range of 400 to 28,000 cycles. However, males which 

 had previously sung with other males did not sing with the artificial partner. 

 In both crickets and katydids"''^ bilateral extirpation of the tympanic organs 

 abolished the normal responses, but did not abolish all sensitivity to sound. 



Quality Discrimination. Regen's experiments demonstrate conclusively 

 that insects are capable of detecting sounds and also that (at least in the 

 two species studied) they are capable of a high degree of quality discrimina- 

 tion. However, he observed that unmated female crickets responded to the 

 male chirp over the telephone even when it was so distorted as to be unrec- 

 ognizable to the human ear. If it is assumed that the mechanism of recog- 

 nition is the same in both crickets and katydids, then it seems that the qual- 

 ities of sound which facilitate recognition in insects are not the same qualities 

 which facilitate recognition by man. 



We recognize sounds primarily on the basis of frequency and to a lesser 

 extent on the basis of amplitude modulation. The structure of the tympanic 

 organ indicates that it functions poorly as a harmonic analyzer, as compared 

 with the human ear. In tettigonids and gryllids the scolopidia of the 

 tympanic organs are arranged in a row and graded in size. In cicadas the 



*30 



lOOO io,ooo 



Fig. 172. Frequency response curves for different types of acoustic receptors. (After 

 Pumphrey and Rawdon-Smith''")- Ordinate, sound pressure above or below 10 dynes 

 per sq. cm., the zero point. Abscissa, frequency in cycles per second on log scale. A, 

 normal human threshold; B, threshold of electrical response in cereal nerve of Gry/Ziis; C, 

 threshold as measured electrically in the tympanal nerve of Lociista. From Wiggles- 

 worth."'" 



scolopidia are in a single mass, and all are about the same size. In acridiids 

 the scolopidia are in three groups on different parts of the tympanum. There 

 is no experimental evidence that these structures are very effective as fre- 

 quency discriminators, but neither is there much evidence that they are not. 



