EXPERIMENTAL

57, 200-211

NEUROLOGY

(1977)

Effects of Visual Attention on Tone Burst Evoked Auditory Potentials LYNN U.S.

Army

C.

Huwan

OATMAN

B.

AND

Engineering

Laboratory,

Maryland Received

March

WAYNE

7, 1977;

ANDERSON Aberdeen

1

Proving Ground,

21005

rcvisimt

rrceivcd

May

2, 1977

Tone burst evoked potentials were recorded from unanesthetized cats with electrodes chronically implanted in the auditory cortex, cochlear nucleus, and round window. The tone bursts (irrelevant stimuli) were presented continuously (85 db SPL, l/s) as background before, during, and after the presentation of a visual discrimination task (relevant stimuli) which was used as a means of altering the attentive state of the animals. At the auditory cortex and cochlear nucleus, the mean peak-to-peak amplitudes of tone burst evoked potentials were significantly reduced during attention to the visual discrimination stimuli compared to the pretest and post-test control periods. However, at the round window (cochlear microphonic) no significant differences in amplitude were observed. Although the amplitudes of the tone burst evoked potentials were reduced at all frequencies during visual attention, much greater suppression occurred at the middle frequencies (700 to 5000 Hz) than at higher or lower frequencies. The results support the hypothesis that during visual attention a central inhibitory mechanism suppresses irrelevant tone burst evoked potentials, presumably via the olivocochlear bundle.

INTRODUCTION Worden (28) discussed two neurophysiological systems which could be responsible for filtering auditory information at the peripheral level. One system, a reticular feedback system, involves the regulation of auditory stimuli through middle-ear muscle contractions. The other system, an extrareticular feedback system, involves the regulation of auditory Abbreviation : OCB-olivocochlear bundle. 1 In conducting the research described herein, the investigators adhered to the Guide for Laboratory Ani?lzal Facilities for Laboratory Animal Resources, National Academy of Sciences, National Research Council, Washington, D.C. The authors gratefully acknowledge the electronic assistance of Joseph Mazurczak and the assistance of Timothy G. Coughlin in reducing the data. 200 Copyright 0 1977 by Academic Press, Inc. All rights of reproduction in any form reserved.

ISSN

0014-4886

VISUAL

ATTENTION

AND

AUDITORY

POTENTIALS

201

stimuli through the action of the olivocochlear bundle (OCB) . The existence of a descending neural pathway from the region of the superior olivary nuclei to the hair cells in the cochlea has been firmly established anatomically (20, 21). It has also been well established that electrical stimulation of the OCB produces (i) a suppression of the auditory nerve action potential as recorded from the round window (69)) (ii) produces an augmentation of the cochlear microphonic (7), and (iii) decreases the firing rate of single fibers of the auditory nerve (8, 27). This evidence has led to the belief that the OCB performs an inhibitory function which controls auditory input to the central nervous system at the peripheral level. Btino et al. (2) have presented direct evidence that the OCB inhibits auditory input at the receptor level in behaving animals. They reported that in unanesthetized guinea pigs, both the cochlear microphonic and auditory nerve amplitudes recorded from the round window were decreased during habituation and distraction experiments. However, those changes associated with habituation and distraction were not observed after transecting the crossed olivocochlear bundle. Oatman (lS, 19) observed that click-evoked auditory potentials at the receptor and cortical levels are suppressed in amplitude, whereas an animal whose middle-ear muscles have been severed is attentive to visual stimulation. Oatman suggested that during attention a central inhibitory mechanism suppresses irrevelant auditory stimuli via the OCB at the peripheral stages in the afferent auditory pathways. This suggestion is supported by Guzmau-Flores and Alcaraz (11) who showed that elimination of the OCB abolishes the suppression of cochlear nucleus evoked potentials during attentive behavior to a visual stimulus. The aim of the present study was to determine whether or not auditory evoked potentials in response to tone bursts at different frequencies are affected differentially during attentive behavior. The study was designed to determine if tone burst evoked potentials recorded from the auditory pathway are suppressed in amplitude while an animal whose middle-ear muscles have been cut is attending to visual stimulation. It was thought that the extrareticular descending system terminating in the OCB would suppress tone burst evoked potentials at the higher frequencies during attention to visual stimulation, because Spoendlin and Gacek (25) reported that the density of crossed efferent innervation of the outer hair cells is greatest in the basal coil of the cochlea. METHODS .C~hj0.f.~. Four fcnlalr cats with clrrtrotlrs surgically l~lacctl at the autlitory cortex, cochlear nucleus, aii(I round window serial as sultjects. A detailed explanation of the surgical preparation used in this experiment can

202

OATMAN

AND

ANDERSON

be found elsewhere (18, 19). At the time of the round window implantation, the tendons of the stapedius and tensor tympani muscles were cut. Histology. At the end of the experiment, the cats were killed with an overdose of sodium pentobarbital administered intravenously. Electrolytic lesions were produced at the recording sites of each cochlear nucleus electrode. The lesion current was 1 mA for 15 s. The brain was removed and placed 24 h in formalin and potassium ferrocyanide. All electrode placements were verified histologically using unstained, frozen sections (22). The histology slides confirmed that the electrodes were placed in the cochlear nucleus. Middle ears were examined to determine that the middle-ear muscle tendons had been completely severed. Visual and Acoustic Stimulation. Although a detailed explanation of the methods used in this experiment can be found elsewhere (18, 19), a brief summary follows. The visual stimuli consisted of two concentric rings presented successively for discrimination. The large outer ring was presented first, which served as a warning stimulus for the cat to attend to the stimuli. Then the smaller inner ring was presented. The cat had to respond to the onset of a small inner ring to receive the food reward. Figure 1 shows a schematic diagram of the presentation of the stimuli. If the cat responded between the onset of the large outer ring and the onset of the small inner ring, he received no reinforcement and the onset of the next trial was delayed 25 s. To increase the cats’ attentiveness and avoid temporal conditioning, the time interval (T,) between onset of the large and small concentric rings was varied randomly between 1 and 6 s. The exposure duration of the small inner concentric ring was 4 s and the time between trials was varied from 20 to 45 s. Auditory tone bursts were presented continuously at a rate of l/s during the presentation of the successive visual discrimination task, but they were

IT---

OUTER

F- ----In

~-

n

FIG.

1. Schematic

diagram

of the presentation

INNER

RING

RING

BEHAVIORAL

RESPONSE

of the stimuli.

VISUAL

ATTENTION

AND

AUDITORY

IOMSEC

4 M SEC ROUND WINDOW

COCHLEAR NUCLEUS

203

POTENTIALS

25MSEC AUDITORY CORTEX

FIG. 2. Averaged tone burst responses recorded from the round window, cochlear nucleus, and auditory cortex. [The round window wave form shows cochlear microphonics (CM) to a single tone burst. The peak-to-peak’conventions used to quantify the evoked potential are noted.]

not synchronized with the onset of the visual display. The tone bursts (2.5 ms rise-decay time, 15 ms duration) generated by a wide-range oscillator (Hewlett-Packard model ZOOCDR) and monitored by an electronic counter (Computer Measurement Corporation model 608) were presented through a sound-tube system which terminated at the entrance to the cat’s external meatus. The sound tube was not fastened to the pinna but was held firmly in place at its entrance by a bracket attached to the electrode plug (17). The tone bursts were presented at an intensity of 85 db SPL (re 0.0002 pbar) at a rate of l/s at each of the 10 frequencies from 200 Hz to 10 kHz. Sound pressures were calibrated with a 0.635-cm condenser microphone (Briiel and Kjaer type 4135) and placed perpendicular to and just in front of the end of the sound tube. Movements of the sound tube to different positions within the test cubicle did not change the output voltage from the microphone. Data Collectiorc apedProcedure. Four weeks after surgery, the cats were placed in the sound-attenuating test cubicle and electrodes were checked. Simultaneous recordings were obtained from the round window (cochlear microphonics), the cochlear nucleus, and the auditory cortex via a Microdot shielded cable connected to an electroencephalograph (Grass model 7). At the same time, the tone burst evoked potentials were recorded on a 14-channel FM tape recorder (Sangamo 4700) from which they were led into a signal averager (Nicolet Med-80) and written on an X-Y plotter (Hewlett-Packard 7035B). Figure 2 shows an example of a tone burst evoked response and indicates how the peak-to-peak measurement was made for each electrode placement. Peak-to-peak amplitudes of the average evoked responseswere measured by ruler to the nearest 5 ,V. Evoked po-

204

OATMAN

AND

ANDERSON

tentials influenced by bodily movement as observed in the electroenchephalogram were discarded from the data. The data w~-e collectetl tluring recording sessions consisting of three different periods designed to effectively manipulate the attentive state of the subjects: (i) a pretest control in which the cat was awake, relaxed, and not attentive to any identifiable stimuli ; (ii) an experimental period during the visual discrimination between the presentation of the outer and inner concentric rings (r,) when the cat was attentive ; and (iii) a post-test control period similar to the pretest control periods. The evoked responses to tone bursts were averaged for each of the three different attentive states, i.e., while the cat was relaxed, while attending to the visual discrimination, and when relaxed again. The tone burst evoked responses, averaged while.the cat was attending to the visual discrimination (condition ii), included only those evoked potentials presented during TI, i.e., between the onset of the large outer concentric ring and the presentation of the small inner concentric ring (Fig. 1). Evoked responses for each of the three attentive states were then collected at each of 10 different frequencies (one frecfuency per clay) in a descending order of presentation from 10 kHz to 200 Hz. RESULTS The data consist of averages of 50 tone burst evoked potentials from three electrode locations : round window (cochlear microphonics), cochlear

x z 2 :

loo 50-

O----O

z ,” 1

0

’ 100

I

1

I

FREQUENCY

AUDITORY FIG.

microvolts

3. The mean peak-to-peak as a function of frequency

D----o 1 IO00

PRETEST CONTROL EXPERIMENTAL POST TEST I I

CONTROL I IO 000

IN Hz

CORTEX

amplitude of and attentive

auditory state.

cortex

evoked

potentials

in

VISUAL

>

I

ATTENTION

z

I

3 I

AND

5 I

7 I

AUDITORY

I 1

2 I

3 1

2 250

5 -~ ri

2

--0 M

FL

FREOUENCY

COCHLEAR

IN

205

POTENTIALS

7 z;s;

I

2 CON TROY-

EXPERIMENTAL POST TEST CONTROL i

Hz

NUCLEUS

FIG. 4. The mean peak-to-peak amplitude of cochlear microvolts as a function of frequency and attentive state.

nucleus

evoked

potentials

in

nucleus, and auditory cortex. The data plotted in Figs. 3 through 5 are averages obtained from 300 measurements from each electrode placement recorded from each of four cats. These figures show the average peak-topeak amplitudes of the auditory responses plotted as a function of frequency for each of the three attentive states: pretest control group (cat nonattentive, relaxed, but awake), experimental group (during visual discrimination, cat very attentive), and post-test control group (cat nonattentive, rel.axed, but awake). Figure 3 shows that the mean peak-to-peak amplitudes of tone burst evoked potentials recorded from auditory cortex were of a smaller amplitude when the cats were very attentive to the visual discrimination than when they were nonattentive. ,4n analysis of variance (3) indicated significant differences between the pretest control group and the experimental group [F (1, 57) = 75.421 and between the post-test control group and the experimental group [I; ( 1, 57) = 72.781. An analysis of variance indicated no significant differences between the pretest control group and the post-test control group [F (1, 57) = 2.241. The data (Fig. 3) show that, although the amplitudes of the auditory cortex evoked potentials are suppressed at all frequencies during visual attention, much greater suppression occurred at the middle frequencies (1000 to 5000 Hz) than at higher or lower frequencies. Whereas the analysis of variance between the pretest control group and the experimental group was significant for the frequencies 200 to 5000 Hz [I; (1, 4.5) = 67.761, the

206

OATMAN

> 2 ::

, r---

~~

2 I/I-

3

5

AND

7

ANDERSON

I T --~---T-

2

500

3 --rl-

5

7

ENPRETEST

kJ 1

2 CONTROL

EXPERIMENTAL -7

a---o

z 5 400 w a

I

-POST

TEST CONTROL

i

g 300

;;;;

/i\

C

-----I---,---

1

L .I(,

100

I--1000

FREQUENCY

COCHLEAR

IO 000 IN Hz

MICROPHONIC

FIG. 5. The mean peak-to-peak amplitude a function of frequency and attentive state.

of cochlear microphonic

in microvolts

as

analysis showed no significant differences between the pretest control group and the experimental group for frequencies 7000 to 10,000 Hz [F (1, 12)= 2.041. The mean peak-to-peak amplitudes of the cochlear nucleus responsesas a function of attentive state and frequency appear in Fig. 4. Figure 4 shows that when the attention of the animals was focused on the visual discrimination, mean peak-to-peak amplitudes of the cochlear nucleus were also reduced. An analysis of variance indicated significant differences between the pretest control group and the experimental group [F (1, 57) = 37.761 and between the post-test control group and the experimental group [F (1, 57) = 43.171, but no significant differences between the pretest control group and the post-test control group (F < 1). Figure 4 also shows that greater reduction in evoked potentials occurred at the middle frequencies (700 to 5000 Hz) than at the higher or lower frequencies. The analysis of variance between the pretest control group and the experimental group was significant for the frequencies 300 Hz [F (1, 8) = 7.731 and 700 to 5000 Hz [F (1, IS)= 35.981; however, an analysis of v,ariance indicated no significant differences between the pretest control group and the experimental group for frequencies 200 Hz [F (1,8) = 2.011, 500 Hz [F (1,s) = 1.551, and 7000 to 10,000 Hz [F (1, 12) = 3.081. Figure 5 shows the mean peak-to-peak amplitudes of the cochlear microphonic responsesas a function of attentive state and frequency. The data

VISUAL

ATTENTION

AND

AUDITORY

207

POTENTIALS

points for 200 and 300 IIz were omitted from the graph because those waveforms were lost in the noise levels at this exposure intensity. The mean peak-to-peak cochlear microphonic responses were not changed in amplitude when the animals were attentive to the visual discrimination task. Analysis of variance indicated no significant differences between the pretest control group and the experimental group (I; < 1) or between the post-test control and the experimental group (I; < 1) . DISCUSSION This experiment shows that the attentive state of the animal can significantly modify the amplitude of tone burst evoked potentials recorded along the auditory pathway. The amplitudes of tone burst evoked potentials recorded from the auditory cortex and cochlear nucleus were significantly smaller when the cats were attentive to the visual discrimination than when the cats were nonattentive. These data are consistent with the previous findings of Oatman ( 18, 19), who observed that when the cats’ attention was focused on the visual discrimination stimuli, click-evoked potentials of the auditory cortex, cochlear nucleus, and auditory nerve were suppressed. These data are also consistent with the findings of Starr (26) and Moushegian et al. (16)) who observed a decreasein amplitude of click-evoked responses at the auditory cortex with middle-ear muscles cut. Oatman (19) previously suggested that the changes in amplitude of the auditory-evoked po-

I

2

3

57

I

2

4

5

7

’

2

FIG. 6. The percentage decrease in mean peak-to-peak amplitude for the experimental group at the auditory cortex and cochlear nucleus as a function of tone burst frequency. Zero percentage decrease indicates no difference in amplitude between the pretest control and experimental groups.

208

OATMAN

AND

ANDERSON

tential as a function of increased visual attention are due to a central inhibitory mechanism which influences auditory nerve responses at the haircell level in the cochlea, but does not involve the cochlear microphonic. The present data are consistent with the above suggestion. Presumably, attention to the visual discrimination task activates a central inhibitory mechanism which suppresses auditory stimuli via the olivocochlear bundle. The present data suggest that the central inhibitory mechanism is mediated by the OCB because the percentage suppression at the cochlear nucleus is about the same as that at the auditory cortex level, even though the latter is several synapses further along the auditory pathway (see Fig. 6). This suggestion is further supported by Igarashi et al. (13), who observed that elimination of the crossed olivocochlear bundle resulted in a significant increase in white noise distraction during a visual detection task in the cat, and Guzman-Flores and Alcaraz (ll), who showed that elimination of the OCB abolishes the suppression of cochlear nucleus evoked potentials during attentive behavior to a visual stimulus. In the present experiment, when the animals were attentive to the visual discrimination stimuli, the amount of suppression of the auditory evoked potentials changed as a function of the tone burst frequency. A greater reduction in the amplitude of tone burst evoked potentials recorded from auditory cortex and cochlear nucleus was observed at middle frequencies than at the higher or lower frequencies. This finding is in agreement with Sohmer (23, 24) and Konishi and Slepian (15), who demonstrated that electrical stimulation of the crossed olivocochlear bundle in anesthetized cats reduced the amplitude of auditory nerve responses for middle frequencies, but the same stimulation did not diminish the amplitudes of auditory nerve responses for the high frequencies. Sohmer (23, 24) presented data showing that the auditory nerve response to middle frequency (3.5 kHz) tone bursts was reduced more than that for higher-frequency (6.5 kHz) tone bursts at the same sound pressure level. He found that the suppression of the auditory nerve was 37% at 3.5 kHz and 25% at 6.5 kHz. The data in Fig. 6 from the present study indicate that the suppression of the cochlear nucleus and the auditory cortex was also greater near 3.5 kHz than near 6.5 kHz. The data in Fig. 6 show that the suppression of the auditory cortex and the cochlear nucleus during visual attention was 51% at 3 kHz and 20% at 7 kH z. The results of the present study on unanesthetized animals, although not involving electrical stimulation of the crossed OCB, accord well with Sohmer’s data. Konishi and Slepian (15) also reported that the inhibition of the auditory nerve produced by electrical stimulation of the crossed olivocochlear bundle is greatest for 4to 6-kHz tone bursts and that these results are in line with those reported by Sohmer (24).

VISUAL

ATTENTION

AND

AUDITORY

POTENTIALS

209

Although the present findings recorded from the auditory cortex and cochlear nucleus are similar to the auditory nerve findings of Konishi and Slepian (IS), these results on the cochlear microphonic potentials are not in agreement with observations by Konishi and Slepian (15) and Kittrell and Dalland (14). Those authors found an augmentation in cochlear microphonic amplitude with electrical stimulation of the crossed olivocochlear bundle, and, in addition, the augmentation of the cochlear microphonic was dependent on the frequency of the acoustic signals. In both studies, maximum augmentation was observed at 1 kHz with augmentation decreasing at higher and lower frequencies. The present experiment showed no significant changes in the amplitude of the cochlear microphonic between the experimental and control groups during visual attention, although the data in Fig. 5 indicate that the cochlear microphonics in the experimental group were slightly increased during visual attention compared to the control groups. The present data are in contrast to the anatomical findings (25) involving the spatial distribution of the OCB fibers along the cochlea. Because the density of the crossed efferent fibers of the outer cells is greatest in the basal coil of the cochlea, we expected that the higher-frequency auditory potentials would be more susceptible to OCB influence. This was not the case in our experiment as we observed maximum suppression in the middle frequencies (700 to 5000 Hz) rather than in the higher frequencies (7000 to 10,000 Hz). The functional significance of the OCB is not known. However, an important possibility has been discussed by Dewson (4). He suggested that by inhibiting the more sensitive (low threshold) fibers associated with the outer hair cells, the OCB may function to increase signal-to-noise ratios. Dewson (5) supported this hypothesis by showing that the ability of monkeys to discriminate among human speech signals in the presence of masking noise was significantly impaired by crossed OCB transection. Perhaps a central inhibitory mechanism activates both the reticular feedback system (middle-ear muscles) and the extrareticular feedback system (OCB) during visual attention to increase the signal-to-noise ratio, and act as a volume control in the cochlea (1) over the frequency range 500 to 5000 Hz. Galambos and Rupert (10) have shown in the cat that the middle-ear muscles attenuate transmission of tones primarily between 500 and 3000 Hz. Because the middle-ear muscles attenuate primarily low frequencies, it was expected that the OCB could attenuate the high frequencies. Therefore, a central inhibitory mechanism was hypothesized which could activate both the reticular feedback system (middle-ear muscles) and the extrareticular feedback system (OCB) to function simultaneously to attenuate the en-

210

OATMAN

AND

ANDERSON

tire frequency range. However, this hypothesis was not supported in the present experiment which demonstrated that maximum suppression occurred in the middle frequencies (700 to 5000 Hz). Although it is still possible that a central inhibitory mechanism activates the two feedback systems during visual attention, it is apparent that the two systems affect approximately the same frequency range, 500 to 5000 Hz. REFERENCES 1. BORG, E. 1971. Efferent inhibition of afferent acoustic activity in the unanesthetized rabbit. Exp. Newel 31 : 301-312. 2. B~No, W., R. VELLUTI, P. HANDLER, AND E. GARCIA-AUSST. 1966. Neural control of the cochlear input in the wakeful free guinea pig. Physiol. Bekav. 1: 23-35. 3. BUTLER, D. H., A. S. KAMLET, AND R. A. MONTY. 1969. A multi-purpose analysis of variance FORTRAN IV computer program. Psychon. Monograph Suppl. 2: 301-319. 4. DEWSON, J. H. 1967. Efferent olivocochlear bundle: Some relationships to noise masking and to stimulus attenuation. J. Nezbvopltysiol. 30 : 817-832. 5. DEWSON, J. H. 1968. Efferent olivocochlear bundle: Some relationships to stimulus discrimination in noise. J. Nezlrop/kysiol. 31: 122-130. 6. DESMEDT, J. E. 1962. Auditory-evoked potentials from cochlea to cortex as influenced by activation of the efferent olivocochlear bundle. J. Acoust. Sot. Amer. 34 : 1478-1496. 7. FEX, J. 1959. Augmentation of cochlear microphonics by stimulation of efferent fibers to the cochlea. Acta Otolaryngob SO: 540-541. 8. FEX, J. 1962. Auditory activity in centrifugal and centripetal cochlear fibers in cat. Acta Pkysiol. Stand. Suppl. 189, 55 : l-68. 9. GALAMBOS, R. 1956. Suppression of the auditory nerve activity by stimulation of efferent fibers to cochlea. J. Newopkysiol. 19 : 424-437. 10. GALAMBOS, R., AND A. RUPERT. 1959. Action of middle ear muscles in normal cats. J. Acmut.

Sot.

Amer.

31 : 349-355.

11. GUZMAN-FLORES, C., AND M. ALCARAZ. 1%3. Funcion de la cintilla olivo coclear en el fenomeno de distraction a estimulos acusticos. Nota preliminar. Biol. Estud. Med. Biol. 21: 87-92. 12. HERNANDEZ-PENN, R. 1966. Physiological mechanisms in attention. Pages 121-144 iw R. W. RUSSELL, Ed., Frontiers in Pllysiological Psychology. Academic Press, New York. 13. IGARASHI, M., B. R. ALFORD, W. P. GORDON, AND Y. NAKAI. 1974. Behavioral auditory function after transection of crossed olivo-cochlear bundle in the cat. Acta Otolaryngol. 77 : 311-317. 14. KITTRELL, B. J., AND J. I. DALLAND. 1969. Frequency dependence of cochlear microphonic augmentation produced by olivocochlear bundle stimulation. Laryngoscope

79 : 228-238.

15. KONISHI, T., AND J. Z. SLEPIAN. 1971. Effects of the electrical stimulation of the crossed olivocochlear bundle on cochlear potentials recorded with intracochlear electrodes in guinea pigs. J. Acoust. Sot. Am. 49 : 1762-1769. 16. MOUSHEGIAN, G., A. RUPERT, J. T. MARSH, AND R. GALAMBOS. 1961. Evoked cortical potentials in absence of middle-ear muscles. Science 133 : 582-583.

VISUAL

ATTENTION

AND

AUDITORY

POTENTIALS

211

17. OATMAN, L. C. 1968. The effect of attention on auditory evoked Potentials. Rep. No. TM 15-68, Human Engineering Laboratory, Aberdeen Proving Ground, Maryland. 18. OATMAN, L. C. 1971. Role of visual attention on auditory evoked potentials in unanesthetized cats. Exp. Neurol. 32 : 341-356. 19. OATMAN, L. C. 1976. Effects of visual attention on the intensity of auditory evoked potentials. Exp. Nertrol. 51 : 41-53. 20. RASMUSSEN, G. L. 1960. Efferent fibers of the cochlear nerve and cochlear nuleus. Pages 105-115 irt G. L. RASMUSSEN AND W. F. WINDLE, Eds., Neural Mechanisms of the Aztditory mtd Vestibular Systerm. Thomas, Springfield, Illinois. 21. RASMUSSEN, G. L. 1964. Anatomic relationships of the ascending and descending auditory systems. Pages 1-19 itz W. S. FIELDS AND B. R. ALFORD, Eds., Neurological Aspects of Auditory a& Vestibltlnr Disorders. Thomas, Springfield, Illinois. 22. SIEGEL, J. 1968. A rapid procedure for locating deep electrode placements. Pltysiol. Behav.

3 : 203-204.

23. SOHI~IER, H. 1965. The effect of contralateral olivo-cochlear bundle stimulation on the cochlear potentials evoked by acoustic stimuli of various frequencies and intensities. .4cfa Otoluryzyol. 60 : 59-70. 24. SOHMER, H. 1966. A comparison of the efferent effects of the homolateral and contralateral olivo-cochlear bundles. rlcfa Otolarylgol. 62 : 74-87. 25. SPOENDLIN, H. H., AND R. R. GACEK. 1963. Electronmicroscopic study of the efferent and afferent innervation of the organ of corti in the cat. Ann. Otol. Rhbtol.

Larylgol.

72 : 6613-686.

26. STARR, A. 1964. Influence of motor activity of click-evoked responses jn the auditory pathway of waking cats. Exp. NmroI. 10 : 191-204. 27. WIEDERH~LD, hf. L., AND N. Y, S. I&NC. 1970. Effects of electrical sthulation of the crossed olivo-cochlear bundle on single auditory-nerve fibers in the cat. J. Acozrst.

Sot.

Amr.

48 : 950-965.

28. WORDEN, F. G. 1966. Attention and a.uditory electrophysiology. Pages 45-107 in E. STEI.LAR AND J. bl. SPRAGUE, Eds., Progress in PhysioIogicol Psychologg. Academic Press, New York.