292 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 1 37 



stronger stimulus. Since the tympanic membranes are about 0.5 cm. 

 apart only a small fraction (about 0.02 msec.) of this latency differ- 

 ence could be due to the difference in the length of the path traveled 

 by the sound, most of it being due to the longer time taken by a weaker 

 stimulus to set off a propagated impulse in the sense cell (compare 

 with pi. 3, fig. i). Thus, the moth appears to have an effective means 

 of determining the approximate direction (right or left) of a sound 

 source, although there is as yet no behavioral evidence that the intact 

 insect is able to act upon this directional information gathered by its 

 auditory organs. 



The sensory characteristics and the behavioral significance of the 

 noctuid ear have been discussed elsewhere (Roeder and Treat, 1957) 

 and will be mentioned only briefly. Electrical recording from the 

 tympanic nerves reveals that noctuids are capable of hearing sounds 

 ranging from 3 kilocycles to well over 100 kilocycles per second, 

 although there is no evidence that one pitch can be discriminated 

 from another. A comparison of plate 2, figure 2, and plate 3, figure i, 

 suggests that the ear is much more efficient in translating sound into 

 nerve impulses when the sound is chopped into a series of short 

 bursts or pulses ; for instance, during a steady tone the impulse fre- 

 quency declines owing to sensory adaptation while the effects of a 

 brief pulse can be said to be amplified by the presence of the after- 

 discharge. This characteristic enables the noctuid ear to follow the 

 rapid succession of ultrasonic pulses emitted by an echo-locating bat 

 (pi. 4, fig. I ), and the resulting changes in flight pattern (Treat, 1955) 

 are of undoubted survival value to the hunted moth. Although other 

 behavioral functions are suggested by the observation (Roeder and 

 Treat, 1957) that the ear can detect short sound pulses emitted at 

 wing-beat frequency by another moth in flight, as well as by the rough 

 directional property mentioned above, the extremely small number 

 of sensory units suggests that certain other sensory refinements have 

 been sacrificed to speed and certainty of operation — prime character- 

 istics in the detection of an active predator. 



Since the central nervous system converts stimulus dimensions into 

 time sequences, the latter must be decoded at the motor end and con- 

 verted once more into the physical magnitudes of muscle tension or 

 limb displacement. This is shown very beautifully in insects because 

 of the small number of nerve fibers in their neuromuscular systems. 

 Thanks to the fine work of Hoyle (1955b) we have a fairly clear 

 picture of the operation of the jumping muscle of the locust men- 

 tioned above. This large muscle can catapult a locust weighing 1.5 g. 



