Bone-conduction (BC) stimulation of the ear is used clinically to diagnose and treat middle-ear disease (Carhart 1950; Naunton 1957; Snik et al. 1998). The simple view is that hearing sensations produced by BC stimulation of the forehead or mastoid result from direct stimulation of the inner ear – bypassing any pathology within the external and/or middle ear (e.g., Naunton 1957). However, the walls of the external ear, the tympanic membrane (TM) and the ossicles also vibrate in response to BC stimulation and produce BC-related sound pressures and fluid motions within the inner ear (Békésy 1932; Wever & Lawrence 1954; Tonndorf 1972; Stenfelt & Goode 2005, Stenfelt 2011). The relative importance and frequency dependences of the external, middle or inner ear pathways for BC inner-ear stimulation is subject to continuing discussion (e.g., Surendran and Stenfelt 2022), especially in animals (e.g., Mason 2003). Here we use chinchillas to examine the external-ear pathway of BC stimulation that is related to BC-induced ear canal sound pressures (Tonndorf, Greenfield & Kaufman 1966; Khanna, Tonndorf & Queller 1976; Stenfelt et al. 2003). These sound pressures are hypothesized to result from oscillating compression and expansion of the external ear-canal walls that perturb the ear-canal volume and generate sound pressures within the canal (Tonndorf et al. 1966; Stenfelt & Reinfeldt 2007). In the normal ear, these sound pressures are transmitted to the inner ear by way of the ossicular chain, following the same pathway that conducts airborne sound to the inner ear (Békésy 1932; Wever & Lawrence 1954).

BC-induced ear-canal sound pressures are commonly associated with the occlusion effect, an increase in the perceived sound level produced by BC stimulation when the open lateral end of the external ear canal is occluded (Naunton 1957; Khanna et al. 1976; Stenfelt et al. 2003). Several theories have been put forward to explain this observation. Mach (1863) hypothesized BC-induced sound-pressures and motions within the middle and inner ear drive the TM in reverse to act as a source of volume velocity into the ear canal. In such a system, the occlusion-produced sound pressure increase is proportional to the increase in acoustic impedance looking out the ear canal from the TM. This idea was challenged by experiments that compared BC induced ear-canal sound pressure in cat ears before and after removing the TM (Tonndorf et al. 1966), where little change was observed in BC-induced ear-canal sound pressures after TM removal.

A second early theory of the increased sensitivity to BC sound after occluding the ear canal was that the occlusion eliminated air-borne masking noise, effectively increasing the sound perceived through BC stimulation (Hallpike 1929; Dean 1930; Pohlmann 1930). Other animal experiments described by Tonndorf et al. (1966) demonstrated a coincidence of occlusion-produced increases in both ear-canal sound pressure and cochlear microphonic (an inner-ear sensory potential) – a finding incompatible with the masking hypothesis. While most accept the hypothesis of wall-vibration induced ear-canal sound pressures, some details remain unexplained and the occlusion effect continues to be studied via measurements of occlusion-induced changes in ear-canal sound pressure (e.g., Fagelson & Martin 1994; Stenfelt et al. 2003), and/or changes in hearing thresholds (Edgerton & Klodd 1976; Berger & Kerivan, 1983; Reinfeldt et al. 2007).

There is some controversy concerning the source of the vibration of the ear-canal walls. Békésy (1932) reasoned, from estimates of the relative mechanical impedance of the bony and cartilaginous walls, that only the cartilaginous portion of the ear canal is capable of motion magnitudes sufficient to compress and rarify the air within the canal. (A common corollary of this observation is that ‘deep’ occlusions that place the occlusion within the bony ear canal produce little occlusion effect (Mueller 1994)). Nonetheless, measurements of ear-canal sound pressures before and after deep occlusions in human subjects and cadavers (Stenfelt et al. 2003; Stenfelt & Reinfeldt 2007) suggests that vibrations of the bony walls of the canal do contribute to BC-induced ear-canal sound pressures.

Békésy also suggested that a significant source of BC-induced cartilaginous ear-canal motion in humans and many mammals is contact between the cartilaginous ear-canal walls and the lower jaw (the mandible) as the compliant jaw-skull joint permits independent motion of the mandible and skull during BC-stimulation. The contribution of jaw motion to BC hearing has been supported by observations of subjective changes in bone-conduction hearing when the jaw is clenched or relaxed (Békésy 1932; Littler, Knight & Strange 1952). Further support for the importance of the mandible comes from measurements in subjects who underwent hemimandibulectomy (removal of a portion of the mandible), where BC-induced ear canal sound pressures measured on the side of resection were diminished with respect to the sound pressures measured in the contralateral ear canal (Howell & Williams 1989). In contrast, an earlier report from two patients who underwent unilateral mandibular resection described an occlusion effect of similar magnitude in the affected and contralateral ear (Allen & Fernandez 1960). Similarly, experiments in cats and human cadavers do not support a contribution of relative canal-jaw motions to BC, as BC-induced ear-canal sound pressure and evoked cochlear potentials were little changed by jaw resection in animals (Tonndorf 1966) and ear-canal sound pressure was little affected in human cadavers (Stenfelt et al. 2003.).

The most extensive collection of work on the occlusion effect and mechanisms related to BC-induced ear-canal sound pressures in humans is that of Stenfelt and his colleagues in live subjects and cadaveric human preparations (Stenfelt et al. 2003; Stenfelt & Goode 2005; Reinfeldt et al. 2007; Stenfelt 2011, Dobrev, Stenfelt et al. 2016; Stenfelt & Prodanovic 2022; Surendran & Stenfelt 2022). This group investigated the sources of BC-induced ear-canal sound pressure by comparing measurements of these sound pressures and umbo velocity in cadaveric human ears and whole heads (Stenfelt et al. 2003; Reinfeldt et al. 2007) before and after multiple structural manipulations. These measurements allowed them to quantify the occlusion effect and address the contribution of the jaw as well as inner- and middle-ear mechanisms to BC-induced ear canal sound pressures. Their results suggest that in humans: i.) The increase in ear-canal sound pressure produced by ear-canal occlusion is as large as 20 dB at frequencies below 2 kHz. ii.) The occlusion effect is diminished, but not eliminated, by removal of the cartilaginous portion of the ear canal. iii.) Relative motions of the jaw with respect to the ear canal do not contribute significantly to the BC-induced ear-canal sound pressures. iv.) Reverse-driven motions of the TM from inner-ear generated BC sound pressures and ossicular motions contribute little to BC-induced ear-canal sound pressures. v.) When the ear canal is occluded, BC hearing is dominated by the BC-induced PTM at frequencies between 0.4 and 1.2 kHz.

Animal models enable near simultaneous measurements of acoustic, mechanical and sensory responses that cannot easily be performed in humans. Various animal models have been used to investigate the mechanisms of bone conduction and its use in diagnosing ear pathology, these include: cat (Marres 1965; Tonndorf 1966), guinea pig (Legouix 1959; Tonndorf 1966, Kanagawa & Tokimoto 1982; Zhao, Fridberger & Stenfelt 2019), mouse (Chhan, McKinnon & Rosowski 2017), and chinchilla (Songer et al. 2004; Chhan et al. 2013; Chhan et al. 2016). The choice of the chinchilla model in the current study was motivated by our ability to record acoustic, mechanical, and neural responses in normal and manipulated ears in this preparation, as well as similarities between chinchillas and humans in the frequency range of hearing and in the dimensions of the external, middle and inner-ear structures (Vrettakos, Dear & Saunders 1988). On the other hand, there are significant structural and functional differences (e.g., Rosowski, Ravicz & Songer 2006; Mason 2013). In chinchilla we see: i.) thinner bones of the ear and skull, ii.) a fused malleus-incus, iii.) a more compliant support of the ossicular system, iv.) a lack of true mastoid air-cells, and v.) a cartilaginous ear canal that projects out from the muscle layers around the skull, and is not embedded within the soft-tissues of the head as it is in human.

This paper describes measurements of BC-induced skull velocities and sound pressures at the TM made before and after manipulations of the middle ear, external ear and jaw. We also report vibration thresholds for BC-induced auditory-evoked potential (the compound action potential: CAP) in chinchillas before and after manipulations of the middle and external ear. These measurements test assumptions about the contributions of the jaw, middle ear and inner ear to BC-induced external-ear sound pressures, and also quantify the contribution of the BC-induced ear-canal sound pressure to BC hearing in the normal and manipulated chinchilla ear.