Microanatomy of the human tunnel of Corti structures and cochlear partition-tonotopic variations and transcellular signaling

Auditory sensitivity and frequency resolution depend on the optimal transfer of sound-induced vibrations from the basilar membrane (BM) to the inner hair cells (IHCs), the principal auditory receptors. There remains a paucity of information on how this is accomplished along the frequency range in the human cochlea. Most of the current knowledge is derived either from animal experiments or human tissue processed after death, offering limited structural preservation and optical resolution. In our study, we analyzed the cytoarchitecture of the human cochlear partition at different frequency locations using high-resolution microscopy of uniquely preserved normal human tissue. The results may have clinical implications and increase our understanding of how frequency-dependent acoustic vibrations are carried to human IHCs. A 1-micron-thick plastic-embedded section (mid-modiolar) from a normal human cochlea uniquely preserved at lateral skull base surgery was analyzed using light and transmission electron microscopy (LM, TEM). Frequency locations were estimated using synchrotron radiation phase-contrast imaging (SR-PCI). Archival human tissue prepared for scanning electron microscopy (SEM) and super-resolution structured illumination microscopy (SR-SIM) were also used and compared in this study. Microscopy demonstrated great variations in the dimension and architecture of the human cochlear partition along the frequency range. Pillar cell geometry was closely regulated and depended on the reticular lamina slope and tympanic lip angle. A type II collagen-expressing lamina extended medially from the tympanic lip under the inner sulcus, here named "accessory basilar membrane." It was linked to the tympanic lip and inner pillar foot, and it may contribute to the overall compliance of the cochlear partition. Based on the findings, we speculate on the remarkable microanatomic inflections and geometric relationships which relay different sound-induced vibrations to the IHCs, including their relevance for the evolution of human speech reception and electric stimulation with auditory implants. The inner pillar transcellular microtubule/actin system's role of directly converting vibration energy to the IHC cuticular plate and ciliary bundle is highlighted.

Postnatal structural development of mammalian Basilar Membrane provides anatomical basis for the maturation of tonotopic maps and frequency tuning

The basilar membrane (BM) of the mammalian cochlea constitutes a spiraling acellular ribbon that is intimately attached to the organ of Corti. Its graded stiffness, increasing from apex to the base of the cochlea provides the mechanical basis for sound frequency analysis. Despite its central role in auditory signal transduction, virtually nothing is known about the BM's structural development. Using polarized light microscopy, the present study characterized the architectural transformations of freshly dissected BM at time points during postnatal development and maturation. The results indicate that the BM structural elements increase progressively in size, becoming radially aligned and more tightly packed with maturation and reach the adult structural signature by postnatal day 20 (P20). The findings provide insight into structural details and developmental changes of the mammalian BM, suggesting that BM is a dynamic structure that changes throughout the life of an animal.

Morphological Immaturity of the Neonatal Organ of Corti and Associated Structures in Humans

Although anatomical development of the cochlear duct is thought to be complete by term birth, human newborns continue to show postnatal immaturities in functional measures such as otoacoustic emissions (OAEs). Some of these OAE immaturities are no doubt influenced by incomplete maturation of the external and middle ears in infants; however, the observed prolongation of distortion-product OAE phase-gradient delays in newborns cannot readily be explained by conductive factors. This functional immaturity suggests that the human cochlea at birth may lack fully adult-like traveling-wave motion. In this study, we analyzed temporal-bone sections at the light microscopic level in newborns and adults to quantify dimensions and geometry of cochlear structures thought to influence the mechanical response of the cochlea. Contrary to common belief, results show multiple morphological immaturities along the length of the newborn spiral, suggesting that important refinements in the size and shape of the sensory epithelium and associated structures continue after birth. Specifically, immaturities of the newborn basilar membrane and organ of Corti are consistent with a more compliant and less massive cochlear partition, which could produce longer DPOAE delays and a shifted frequency-place map in the neonatal ear.

The secondary spiral lamina and its relevance in cochlear implant surgery

OBJECTIVE

We used synchrotron radiation phase contrast imaging (SR-PCI) to study the 3D microanatomy of the basilar membrane (BM) and its attachment to the spiral ligament (SL) (with a conceivable secondary spiral lamina [SSL] or secondary spiral plate) at the round window membrane (RWM) in the human cochlea. The conception of this complex anatomy may be essential for accomplishing structural preservation at cochlear implant surgery.

MATERIAL AND METHODS

Sixteen freshly fixed human temporal bones were used to reproduce the BM, SL, primary and secondary osseous spiral laminae (OSL), and RWM using volume-rendering software. Confocal microscopy immunohistochemistry (IHC) was performed to analyze the molecular constituents.

RESULTS

SR-PCI reproduced the soft tissues including the RWM, Reissner's membrane (RM), and the BM attachment to the lateral wall (LW) in three dimensions. A variable SR-PCI contrast enhancement was recognized in the caudal part of the SL facing the scala tympani (ST). It seemed to represent a SSL allied to the basilar crest (BC). The SSL extended along the postero-superior margin of the round window (RW) and immunohistochemically expressed type II collagen.

CONCLUSIONS

Unlike in several mammalian species, the human SSL is restricted to the most basal portion of the cochlea around the RW. It anchors the BM and may influence its hydro-mechanical properties. It could also help to shield the BM from the RW. The microanatomy should be considered at cochlear implant surgery.

Effects of basilar membrane arch and radial tension on the travelling wave in gerbil cochlea

The basilar membrane velocity of gerbil cochlea showed discrepancy between theoretical model and experimental measurements. We hypothesize that the reasons of such discrepancies are due to the arch towards the scala tympani and radial tension present in the basilar membrane of the gerbil cochlea. The arch changes the bending stiffness in the basilar membrane, reduces the effective fluid force on the membrane and increases the basilar membrane's inertia. The existence of the radial tension also dampens the acoustic travelling wave. In this paper, the wave number functions along the gerbil basilar membrane are calculated from experimentally measured physical parameters with the theoretical model as well as extracted from experimentally measured basilar membrane velocity with the wave number inversion formula. The two wave number functions are compared and the effects of the tension and membrane arch on the wave number are studied based on various parameters of the model. We found that the bending stiffness across the gerbil basilar membrane varies (1-2 orders along the cochlea in the section 2.2 mm-3 mm from base) more than the calculated value in the flat basilar membrane model and the radial tension increases the damping of the travelling wave in gerbil cochlea significantly (5 times more than that without radial tension). These effects of arch and radial tension in the basilar membrane elucidate the discrepancy between previous theoretical model and experimental measurements in gerbil cochlea.

Two-compartment passive frequency domain cochlea model allowing independent fluid coupling to the tectorial and basilar membranes

The cochlea is a spiral-shaped, liquid-filled organ in the inner ear that converts sound with high frequency selectivity over a wide pressure range to neurological signals that are eventually interpreted by the brain. The cochlear partition, consisting of the organ of Corti supported below by the basilar membrane and attached above to the tectorial membrane, plays a major role in the frequency analysis. In early fluid-structure interaction models of the cochlea, the mechanics of the cochlear partition were approximated by a series of single-degree-of-freedom systems representing the distributed stiffness and mass of the basilar membrane. Recent experiments suggest that the mechanical properties of the tectorial membrane may also be important for the cochlea frequency response and that separate waves may propagate along the basilar and tectorial membranes. Therefore, a two-dimensional two-compartment finite difference model of the cochlea was developed to investigate the independent coupling of the basilar and tectorial membranes to the surrounding liquid. Responses are presented for models using two- or three-degree-of-freedom stiffness, damping, and mass parameters derived from a physiologically based finite element model of the cochlear partition. Effects of changes in membrane and organ of Corti stiffnesses on the individual membrane responses are investigated.

Quantitative polarized light microscopy of unstained mammalian cochlear sections

Hearing loss is the most common sensory deficit in the world, and most frequently it originates in the inner ear. Yet, the inner ear has been difficult to access for diagnosis because of its small size, delicate nature, complex three-dimensional anatomy, and encasement in the densest bone in the body. Evolving optical methods are promising to afford cellular diagnosis of pathologic changes in the inner ear. To appropriately interpret results from these emerging technologies, it is important to characterize optical properties of cochlear tissues. Here, we focus on that characterization using quantitative polarized light microscopy (qPLM) applied to unstained cochlear sections of the mouse, a common animal model of human hearing loss. We find that the most birefringent cochlear materials are collagen fibrils and myelin. Retardance of the otic capsule, the spiral ligament, and the basilar membrane are substantially higher than that of other cochlear structures. Retardance of the spiral ligament and the basilar membrane decrease from the cochlear base to the apex, compared with the more uniform retardance of other structures. The intricate structural details revealed by qPLM of unstained cochlear sections ex vivo strongly motivate future application of polarization-sensitive optical coherence tomography to human cochlea in vivo.

Probing cochlear tuning and tonotopy in the tiger using otoacoustic emissions

Otoacoustic emissions (sound emitted from the ear) allow cochlear function to be probed noninvasively. The emissions evoked by pure tones, known as stimulus-frequency emissions (SFOAEs), have been shown to provide reliable estimates of peripheral frequency tuning in a variety of mammalian and non-mammalian species. Here, we apply the same methodology to explore peripheral auditory function in the largest member of the cat family, the tiger (Panthera tigris). We measured SFOAEs in 9 unique ears of 5 anesthetized tigers. The tigers, housed at the Henry Doorly Zoo (Omaha, NE), were of both sexes and ranged in age from 3 to 10 years. SFOAE phase-gradient delays are significantly longer in tigers--by approximately a factor of two above 2 kHz and even more at lower frequencies--than in domestic cats (Felis catus), a species commonly used in auditory studies. Based on correlations between tuning and delay established in other species, our results imply that cochlear tuning in the tiger is significantly sharper than in domestic cat and appears comparable to that of humans. Furthermore, the SFOAE data indicate that tigers have a larger tonotopic mapping constant (mm/octave) than domestic cats. A larger mapping constant in tiger is consistent both with auditory brainstem response thresholds (that suggest a lower upper frequency limit of hearing for the tiger than domestic cat) and with measurements of basilar-membrane length (about 1.5 times longer in the tiger than domestic cat).

Dual traveling waves in an inner ear model with two degrees of freedom

We calculate traveling waves in the mammalian cochlea, which transduces acoustic vibrations into neural signals. We use a WKB-based mechanical model with both the tectorial membrane (TM) and basilar membrane (BM) coupled to the fluid to calculate motions along the length of the cochlea. This approach generates two wave numbers that manifest as traveling waves with different modes of motion between the BM and TM. The waves add differently on each mass, producing distinct tuning curves and different characteristic frequencies (CFs) for the TM and the BM. We discuss the effect of TM stiffness and coupling on the waves and tuning curves. We also consider how the differential motions between the masses could influence the cochlear amplifier and how mode conversion could take place in the cochlea.

A computational model of human auditory signal processing and perception

A model of computational auditory signal-processing and perception that accounts for various aspects of simultaneous and nonsimultaneous masking in human listeners is presented. The model is based on the modulation filterbank model described by Dau et al. [J. Acoust. Soc. Am. 102, 2892 (1997)] but includes major changes at the peripheral and more central stages of processing. The model contains outer- and middle-ear transformations, a nonlinear basilar-membrane processing stage, a hair-cell transduction stage, a squaring expansion, an adaptation stage, a 150-Hz lowpass modulation filter, a bandpass modulation filterbank, a constant-variance internal noise, and an optimal detector stage. The model was evaluated in experimental conditions that reflect, to a different degree, effects of compression as well as spectral and temporal resolution in auditory processing. The experiments include intensity discrimination with pure tones and broadband noise, tone-in-noise detection, spectral masking with narrow-band signals and maskers, forward masking with tone signals and tone or noise maskers, and amplitude-modulation detection with narrow- and wideband noise carriers. The model can account for most of the key properties of the data and is more powerful than the original model. The model might be useful as a front end in technical applications.

Gerbil middle-ear sound transmission from 100 Hz to 60 kHz

Middle-ear sound transmission was evaluated as the middle-ear transfer admittance H(MY) (the ratio of stapes velocity to ear-canal sound pressure near the umbo) in gerbils during closed-field sound stimulation at frequencies from 0.1 to 60 kHz, a range that spans the gerbil's audiometric range. Similar measurements were performed in two laboratories. The H(MY) magnitude (a) increased with frequency below 1 kHz, (b) remained approximately constant with frequency from 5 to 35 kHz, and (c) decreased substantially from 35 to 50 kHz. The H(MY) phase increased linearly with frequency from 5 to 35 kHz, consistent with a 20-29 micros delay, and flattened at higher frequencies. Measurements from different directions showed that stapes motion is predominantly pistonlike except in a narrow frequency band around 10 kHz. Cochlear input impedance was estimated from H(MY) and previously-measured cochlear sound pressure. Results do not support the idea that the middle ear is a lossless matched transmission line. Results support the ideas that (1) middle-ear transmission is consistent with a mechanical transmission line or multiresonant network between 5 and 35 kHz and decreases at higher frequencies, (2) stapes motion is pistonlike over most of the gerbil auditory range, and (3) middle-ear transmission properties are a determinant of the audiogram.

Psychophysical estimates of level-dependent best-frequency shifts in the apical region of the human basilar membrane

It is now undisputed that the best frequency (BF) of basal basilar-membrane (BM) sites shifts downwards as the stimulus level increases. The direction of the shift for apical sites is, by contrast, less well established. Auditory nerve studies suggest that the BF shifts in opposite directions for apical and basal BM sites with increasing stimulus level. This study attempts to determine if this is the case in humans. Psychophysical tuning curves (PTCs) were measured using forward masking for probe frequencies of 125, 250, 500, and 6000 Hz. The level of a masker tone required to just mask a fixed low-level probe tone was measured for different masker-probe time intervals. The duration of the intervals was adjusted as necessary to obtain PTCs for the widest possible range of masker levels. The BF was identified from function fits to the measured PTCs and it almost always decreased with increasing level. This result is inconsistent with most auditory-nerve observations obtained from other mammals. Several explanations are discussed, including that it may be erroneous to assume that low-frequency PTCs reflect the tuning of apical BM sites exclusively and that the inherent frequency response of the inner hair cell may account for the discrepancy.

Basilar membrane tension calculations for the gerbil cochlea

Anatomical studies suggest that the basilar membrane (BM) supports a radial tension, which is potentially important in cochlear mechanics. Assuming that the tension exists, we have calculated its magnitude from measurements of BM stiffness, longitudinal coupling, and geometry using a BM model. Results for the gerbil cochlea show that the tension decreases from the base to the apex of the cochlea and generates a tensile stress that is comparable in magnitude to the stress generated in other physiological systems. The model calculations are augmented by experiments that investigate the source of BM tension. The experimental results suggest that BM tension is maintained by the spiral ligament.

The cochleogram of the guinea pig

The cochleogram is an important tool to relate properties of the cochlea (e.g. hair cell loss, damaged hair cells) to their position in the cochlear turns, to calculate the average hair cell density, and to measure the length of the whole cochlea. In this work different methods of plotting cochleograms are compared. We suggest that a sector-wise division of the cochlea for counting a cochleogram has advantages over line diagrams that provide a higher spatial resolution but might lead to misinterpretations of the degree of missing hair cells. The scanning electron microscopic analysis of 171 guinea pig cochleas revealed a mean basilar membrane length of 16.4 +/- 1.4 mm (mean +/- standard deviation) with sector lengths of 6.9, 4.2, 3.2, and 1.9 mm, thus adding relevant information to the morphology of the guinea pig cochlea.

Cochlea's graded curvature effect on low frequency waves

In the ear, sound waves are processed by a membrane of graded mechanical properties that resides in the fluid-filled spiral cochlea. The role of stiffness grading as a Fourier analyzer is well known, but the role of the curvature has remained elusive. Here, we report that increasing curvature redistributes wave energy density towards the cochlea's outer wall, affecting the shape of waves propagating on the membrane, particularly in the region where low frequency sounds are processed.

Developing a Physical Model of the Human Cochlea Using Microfabrication Methods

Advances in micro-machining technology have provided the opportunity to explore possibilities of creating life-sized physical models of the cochlea. The physical model of the cochlea consists of two fluid-filled channels separated by an elastic partition. The partition is micro-machined from silicon and uses a 36-mm linearly tapered polyimide plate with a width of 100 microm at the basal end and 500 microm at the apex to represent the basilar membrane. Thicknesses from 1 to 5 microm have been fabricated. Discrete aluminum fibers (1.5 microm in width) are machined to create direction-dependent properties. A 0.5 x 0.5 mm opening represents the helicotrema. The fluid channels are machined from plexiglas using conventional machining methods. A magnet-coil system excites the fluid channel. Measurements on a model with thickness 4.75 microm show a velocity gain of 4 and phase of 3.5 pi radians at a location 23 mm from the base. Mathematical modeling using a 3-D formulation confirm the general characteristics of the measured response.

A self-mixing laser-diode interferometer for measuring basilar membrane vibrations without opening the cochlea

A laser-diode forms the basis of a displacement sensitive homodyne interferometer suitable for measurements from poorly reflective surfaces. The compact and cost-effective interferometer utilizes the self-mixing effect when laser light reflected from a moving target re-enters the laser cavity and causes phase dependent changes of the lasing intensity. A piezo positioner was used to displace the interferometer with known frequency and amplitude as a basis for real-time calibration of the interferometer's sensitivity. The signal-processing algorithm is described that allows measurements in presence of high amplitude noise leading to variation of the interferometer's operating point. Measurements of sound-induced basilar membrane displacements were made in the intact cochleae of rodents by focusing the laser beam of the interferometer through the transparent round window membrane. The interferometer provides a viable means for making subnanometre mechanical measurements from structures in the inner ears of small mammals, where opening of the cochlea is not practicable.

Two modes of motion of the alligator lizard cochlea: measurements and model predictions

Measurements of motion of an in vitro preparation of the alligator lizard basilar papilla in response to sound demonstrate elliptical trajectories. These trajectories are consistent with the presence of both a translational and rotational mode of motion. The translational mode is independent of frequency, and the rotational mode has a displacement peak near 5 kHz. These measurements can be explained by a simple mechanical system in which the basilar papilla is supported asymmetrically on the basilar membrane. In a quantitative model, the translational admittance is compliant while the rotational admittance is second order. Best-fit model parameters are consistent with estimates based on anatomy and predict that fluid flow across hair bundles is a primary source of viscous damping. The model predicts that the rotational mode contributes to the high-frequency slopes of auditory nerve fiber tuning curves, providing a physical explanation for a low-pass filter required in models of this cochlea. The combination of modes makes the sensitivity of hair bundles more uniform with radial position than that which would result from pure rotation. A mechanical analogy with the organ of Corti suggests that these two modes of motion may also be present in the mammalian cochlea.

The guide to plotting a cochleogram

The cochleogram is commonly used for illustrating hair cell loss after insult, yet standardized procedures for plotting either individual or averaged cochleograms are lacking despite more than 40 years of use. Due to the intra-species variation in basilar membrane (BM) length, it is important that length is plotted on the cochleogram in percent and not millimeter. It is also of interest to correlate the location of lesion to frequency by using a frequency-place equation. However, there is no consensus as which equation is most suitable for the species under study. This is an important issue since two different equations can result in significantly different frequency-place maps for the same cochlea. The purpose of this presentation is to suggest procedures for standardizing the cochleogram. The guidelines include: (i) basilar membrane length should be plotted as percent instead of millimeter due to the biological variation that exists in BM length within a particular species and strain, and the total length in millimeter stated on the cochleogram; (ii) the equations used for frequency-place maps should be stated on the cochleogram; (iii) different basilar membrane lengths should be normalized to percent before averaged cochleograms are made. These procedures are illustrated and discussed.

Cochlear development: hair cells don their wigs and get wired

PURPOSE OF REVIEW

Hair cells and spiral ganglion neurons form functional pairings in the cochlea that transduce the mechanical energy of sound into signals that are carried to the brainstem. Mutations of genes affecting the development and maintenance of these two cell populations cause deafness in humans and other animals. This review highlights recent findings regarding the development of hair cell stereocilia and spiral ganglion neurons in the cochlea.

RECENT FINDINGS

Genes underlying Usher syndrome 1A have shed light on possible molecular participants in the development and structure of the hair cell stereocilia. Analysis of deaf mouse mutants have uncovered genes involved in stereocilia elongation and the orientation of the stereociliary bundles. Studies on the regulation of spiral ganglion neuronal survival and guidance suggest that the timing of expression of specific growth factors along the cochlear spiral is involved in maintaining the divergence of vestibular and cochlear nerve fibers.

SUMMARY

Examining human and mouse genes affecting hearing has not only provided insight into causes of human deafness, but has also opened a window into how stereociliary bundles are constructed and spiral ganglion neurons are preserved and guided during development. Synthesis of information from diverse lines of research pinpoints genes for screening or repair in the genetic medicine of the future and dramatizes the intimate relationship between strict adherence to complex developmental programs and hearing. In addition, future improvements in the efficacy of cochlear implants may depend on the preservation and manipulation of adult spiral ganglion neurons. Developmental mechanisms promise to yield insight into possible interventions to redirect or reconnect spiral ganglion neurons in damaged cochlea.

[Optic coherent tomography: a new high-resolution technology of visualization of tissue structures. Communication II. Optical images of benign and malignant entities]

This is the second communication of a series of publications on Russian studies in the field of optical coherent tomography (OCT), the newest noninvasive highly resolving technology of visualization of the structure of biological tissues. By using the investing tissues as an example, this paper demonstrates the universal types of changes in their optical properties. Optimal images permit differentiate benign and malignant processes with a high degree of diagnostic accuracy. Diverse benign processes occurring in the epithelium are detected on the OCT images as changes in its height, the scattering properties and stroke of a basilar membrane. The absence of any structure on the image is the main OCT criterion for malignancy. The diagnostic efficiency of OCT is high in recognizing neoplasia of various mucous membranes: the sensitivity of the technique is 77-98%; its specificity and diagnostic accuracy are 71-96 and 81-87%, respectively.

Basic maps in the auditory midbrain

Topologic maps at consecutive levels of sensory pathways indicate behaviorally relevant features of stimuli at increasing degrees of complexity. In the auditory system, except for tonotopic maps, the nature of represented features is unknown. In a model analogous to visual map formation we show that in the auditory midbrain, layers of neurons with preference to the same frequency (isofrequency planes) may hold maps of two basic, mutually orthogonal parameters--instantaneous amplitude and phase--of basilar membrane displacement at the cochlear location responding to that frequency. The proposed neural tuning to frequency, amplitude, and phase implies that sound is transformed into specific temporal trajectories of neural activation, with consequences for experimental design and interpretation of neural response behavior.

Anatomical differences in the peripheral auditory system of mammals and man. A mini review

The major anatomical differences among animal models and man are briefly reviewed. Differences are described in the length and width of the basilar membrane, the number of inner (IHCs) and outer hair cells (OHCs), and the length of cilia on both cell types. Significant differences in the innervation pattern of the IHCs among these species include the number of afferent nerve terminals per IHC, the degree of branching of afferent fibers and the number of synapses per afferent nerve terminal. At the OHCs, the number of afferent and efferent nerve terminals, the presence or absence of presynaptic bodies, reciprocal synapses and the presence of dendrodendritic synapses in the outer spiral bundles may have important physiological functions. In the cochlear nerve, significant differences are described in the number of spiral ganglion cells (SGCs) and cochlear nerve fibers. Furthermore the percentage of myelinated SGCs and the presence of synapses on SGCs varies enormously.

Cochlear dimensions obtained in hemicochleae of four different strains of mice: CBA/CaJ, 129/CD1, 129/SvEv and C57BL/6J

Because homologies between mice and human genomes are well established and hereditary abnormalities are similar in both, mice present a valuable animal model to study hereditary hearing disorders in humans. One of the manifestations of hereditary hearing disorders might be in the structure of cochlear elements, such as the gross morphology of the cochlea. Cochlear dimensions, however, are one factor that determines inner ear mechanics and thus hearing function. Therefore, gross cochlear dimension might be important when different strains of mice are compared regarding their hearing. Although several studies have examined mouse inner ear structures on a sub-cellular level, only few have studied cochlear gross morphology. Moreover, the sparse data available were acquired from fixed and dehydrated tissue. Dehydration, however, produces severe distortion of gel-like cochlear structures such as the tectorial membrane and the basilar membrane hyaline matrix. In this study, the hemicochlea technique, which allows fresh mouse cochlear material to be viewed from a radial perspective, was used to provide an itemized study of the dimensions of gross cochlear structures in four mouse strains (CBA/CaJ, 129/SvEv, 129/CD1 and C57BL/6J). Except for the CBA/CaJ, these strains are known to possess genes for age-related hearing loss. The measurements showed no major differences among the four strains. However, when compared with previous data, the thickness measures of the basilar membrane were up to 10 times larger. Such differences are likely to result from the different techniques used to process the material. The hemicochlea technique eliminates much of the distortion caused by dehydration, which was present in previous experiments.