| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Laboratory of Applied Biology and Research Unit in Neurophysiology, Université Libre de Bruxelles, Brussels, Belgium
Submitted 28 April 2008 ; accepted in final form 31 October 2008
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
muscle fascicle; passive torque; pennation angle; stretching; ultrasonography
Classically, passive stretching refers to the technique of lengthening a muscle group by slowly moving a joint to its maximal range of motion and maintaining the position for a period of 
15–30 s. In the literature, it is often suggested that the passive torque produced during passive stretching is a limiting factor for some joint amplitudes (11, 33) and can be the result of both neural and mechanical factors (2, 10, 37). For example, it has been shown that the decline in spinal reflex excitability during different stretching methods is associated with the magnitude of muscle length changes (11). Collectively, these results suggest that tonic reflex activity might contribute to muscle passive resistance and at least partially limit the passive range of motion of a joint.
Studies on isolated preparation in animals have shown that the increase in passive resistive force during the lengthening of a muscle or muscle fiber is associated with intrinsic muscular factors (myofibrils; Refs. 19, 32) and the amount of connective tissues surrounding the muscle belly (7, 42). When the whole MTU is concerned, passive structures in both the muscle and tendon tissues (free tendon) can contribute to increases in passive torque during lengthening (2, 21, 29, 35). Although the contribution of the free tendon and passive muscle structures to the increase in passive resistive force is well established (19, 32, 34, 44), it is not clear which structure is the limiting factor in the elongation of the MTU (13, 38).
The development of imaging techniques, such as ultrasonography, allows one to determine the architecture of the MTU in humans (41). This noninvasive method makes it possible to measure changes in fascicle length (Lf) and tendon tissue behavior during muscle stretching (12, 25, 29, 30, 38). Surprisingly, different results have been reported. For example, Herbert et al. (12) observed in the medial gastrocnemius (MG) that a large proportion (75%) of the whole MTU lengthening during passive stretching is taken up by the tendon. More recently, however, Morse and coworkers (38) reported a similar contribution of free tendon and muscle length changes to the whole MTU lengthening in the same muscle. In addition to differences in the experimental approach between the two studies (38), these conflicting results could be due to differences in the individual characteristics of the muscle and free tendon tissues. It has been reported, for example, that the peak torque achieved (33, 37) and the changes in fascicle and tendon length (Lt) (13) during passive lengthening can differ among individuals, yet it is not clear which structures limit MTU lengthening.
The aim of the present study was to investigate whether changes in MG architecture differ between individuals who produce different passive torques during muscle stretching of the plantar flexor muscles and whether the relative contribution of muscle and tendon structures to the increase in total MTU lengthening varies among subjects.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Experimental Setup and Protocols
Subjects lay prone on a table with both legs extended and the right foot strapped to a footplate. Leg dominance was not a critical factor in the present study, as Manal et al. (36) did not observe any difference in pennation angles for various muscles in the right and left legs, and preliminary measurements in our laboratory on five subjects did not show a statistical difference (P > 0.05) in resting MG Lf between the right (53.3 ± 1.3 mm) and left (51.8 ± 1.4 mm) legs. The right foot was attached to the plate by three straps and held in place by an adjustable heel block. The first strap was placed over the dorsum of the foot, and the two other straps were attached around the ankle and calcaneum. The angular displacement of the ankle joint was monitored with a potentiometer that was mounted on the rotational axis of the footplate. The ankle joint axis was visually aligned with the rotational axis of the footplate. In control conditions, the foot was placed in the anatomical position (0°; the sole of the foot being perpendicular to the tibia axis). The footplate was connected by a steel cable to a mechanical device that enabled graduated passive dorsiflexion of the ankle. With this arrangement, the ankle angle was successively rotated from 10° plantar flexion to 30° dorsiflexion in 10° steps, at an angular velocity of 2.5°/s. The ankle was held at each angle for 20 s. All stretches were performed up to 30° of dorsiflexion, because all subjects were able to reach this range of motion while keeping the heel in contact with the footplate. Throughout the stretching maneuver, the subjects were requested to relax completely and to not resist the movement of the footplate. Joint angle, passive torque, and electromyographic (EMG) activities of the MG and soleus (Sol) muscles were recorded over the entire stretch. At the end of the experiment, the subject performed one maximal isometric voluntary contraction (MVC) of the plantar flexor muscles with the ankle joint at 0°, during which the maximum EMG activity was recorded. To compare the relative increase in MG Lt during stretching and MVC, a subset of 12 subjects performed three isometric MVCs of the plantar flexor muscles with the ankle joint at 0° to measure Lt change during a muscle contraction. Care was taken to minimize joint rotation during the MVC (4, 20).
Mechanical and EMG Recordings
The torque produced by the plantar flexor muscles during passive stretching and the MVC were recorded by means of a strain-gauge transducer (model TC 200–500, linear range 0–1,100 N; Kulite, Basingstoke, UK). The EMG activity was recorded from MG and Sol muscles by means of two silver surface electrodes (disks of 8 mm in diameter) in a bipolar configuration. The electrodes were positioned over the muscle belly of both muscles using a constant interelectrode distance of 20 mm. The ground electrodes were placed over the tibia. The EMG signals were amplified by a custom-made differential amplifier (x1,000) and filtered (10 Hz to 1 kHz). EMG was used to monitor muscle activity during passive ankle dorsiflexion (11) to ensure that the muscles were quiescent during the stretch. The signals (ankle angle, torque, and EMG activities) were acquired on a personal computer at a sampling rate of 2,500 Hz by a MP 100 data acquisition system (Biopac Systems, Santa Barbara, CA).
Ultrasonographic Recordings
The architectural changes of the MG muscle during stretching and the MVC were investigated by ultrasonography. Longitudinal images were obtained using real-time B-mode ultrasonographic apparatus (AU5, Esoate Biomedica, Firenze, Italy) with a 13-MHz linear-array probe. The probe was fixed firmly onto the right leg (at 30% of the distance between the popliteal crease and the center of medial malleolus), over the midbelly of the muscle (Fig. 1A). Once a muscle fascicle was clearly identified, the position of the probe was firmly held in place using a custom-made resin sheath strapped to the skin. The restraint ensured a constant orientation and pressure of the probe. A metallic marker was placed between the skin and the ultrasound probe to verify that the probe did not move during the recording (6, 13, 26, 43). The probe was coated with a water-soluble transmission gel to provide acoustic contact. Images were acquired on a personal computer at a sampling rate of 20 Hz with commercialized software (Pinnacle Studio Plus, version 9) and analyzed offline.
|
The passive torque, joint angle, and mean EMG amplitude of MG and Sol were continuously recorded over the ankle range of motion. To reduce the possible influence of stress relaxation (33, 44) and increased reflex EMG activity immediately after each 10° step rotation, passive torque and EMG were measured during the last 5 s of each 20-s step. The slope of the passive torque-angle curve from +20° to +30° was used to estimate passive stiffness (26). The torque produced during the isometric MVC and corresponding EMG activity with the ankle in neutral position (0°) was measured during a 2-s epoch at the torque plateau. EMG activities during stretching were expressed as a percentage of maximum EMG during the MVC.
Measurement of muscle architectural parameters. Two parameters were measured from each MG ultrasound scan: muscle Lf and pennation angle (6, 18, 30, 41). The muscle fascicle was defined as a clearly visible fiber bundle lying between the two aponeuroses (superficial and deep) by using a public domain image program (Scion image, National Institutes of Health). Pennation angle (µ) was determined as the angle between the fascicle and its insertion on the deep aponeurosis (Refs. 18, 41; Fig. 1B). The Lf was measured along the marked fiber bundle, from the superficial to the deep aponeurosis (6, 18). When the end of the fascicle extended off the acquired ultrasound image, Lf was estimated by trigonometry [total Lf = lf 1 (measured fasicle length) + lf 2 (estimated fasicle length) = lf 1 + (h/sin µ), where h is height] (Fig. 1) by assuming a linear continuation of the fasicles (2, 43). The error due to the linear extrapolation has been estimated to be 2–7% (5).
Measurement of aponeurosis displacement. The point of insertion of the MG and Sol muscle fascicles on their respective deep and superficial aponeuroses was determined on the same scan using the method proposed by Bojsen-Moller et al. (4). The longitudinal displacement of these points during passive ankle dorsiflexion was considered to represent the lengthening of the respective aponeurosis and free tendon distal to that point (4, 26). On each scan, three fascicles for each muscle were measured and averaged.
Measurement of MTU length at rest. The total MTU length (LMTU) at rest (ankle angle at 0°) was estimated over the skin with a tape measure after having identified the proximal (medial femoral condyle) and distal (superior edge of the calcaneum) insertions of the MG.
Estimation of the change in LMTU during the stretch.
The position of the ultrasound probe over the MG impeded a direct measure of the LMTU change during the stretching maneuver. Therefore, we used the regression equation provided by Grieve et al. (9) to calculate the change in LMTU during stretching. The percent LMTU change was calculated from the following equation: 
L = –22.185 + 0.30141 (90 + A) – 0.00061 (90 + A)2, where L is the change in LMTU due to change in dorsiflexion angle (A) (Fig. 2). In our experimental conditions, the mean length change of MTU corresponded to 0.85 mm/°, a value comparable to the 0.83 and 0.78 mm/° reported by Herbert et al. (12) and Morse et al. (38), respectively. Furthermore, in complementary experiments on three subjects, the change in LMTU obtained by the equation of Grieve et al. (9) was compared with the values measured directly (see above). Compared with an ankle angle at 0°, MTU lengthening at 10, 20, and 30° of dorsiflexion was, respectively, 8.8 ± 0.9, 17 ± 1.9, and 24.4 ± 1.8 mm when predicted by the equation of Grieve et al. (9) and 9.8 ± 0.5, 18.6 ± 0.6, and 26 ± 2 mm when measured directly.
|
Reliability and Validity of the Ultrasound Measurements
The reliability of the Lf measurement with the ankle joint in a neutral position (0°) was evaluated in 13 subjects by measuring the same fascicle on a given scan three times. The coefficients of variation (= SD/mean) were 0.4 and 4.2% for pennation angle and Lf, respectively. Ito et al. (18) reported a greater coefficient of variation for Lf (2%) and for pennation angle (7%). The extent of Lt relative to the change in total LMTU has been estimated indirectly from the measurement of the change in Lf (2, 12, 27). The assumption of this method is that fascicles along the muscle belly lengthen in a similar manner when the muscle is passively stretched (see Refs. 30, 39, 41).
Statistics
Conventional statistical methods were used for calculating the means, standard deviations (SD), standard errors (SE), and the coefficients of correlation. Changes for the different parameters were analyzed by means of an ANOVA with repeated measures on one factor (ankle angle). When significant main effects were observed, a Tukey-Kramer post hoc test was used to identify the significant differences among means. The comparison between the two subgroups of subjects was analyzed with a two-way ANOVA with repeated measures on ankle angle or by a Student t-test when a single comparison was performed. The level of significance was set at P < 0.05 for all analyses. Data are reported as means ± SD within the text and displayed as means ± SE in Figs. 3–6.
|
|
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Average passive torque increased exponentially (r2 = 0.99) with an increase in ankle angle from –10° to +30° (Fig. 3). Torque increased from 1.9 ± 1.6 N·m at –10° to 30.4 ± 13.1 N·m at +30°. Passive stiffness, calculated from the slope of the relation between passive torque and ankle angle between +20 and +30°, reached a value of 1.02 ± 0.5 N·m/°.
The EMG of the MG and Sol were monitored during the stretch and found to indicate relatively little muscle activity. The average EMG increased slightly with the stretch to reach values (30° dorsiflexion) of 1 ± 0.8 and 2.4 ± 1.3% of the MVC value for MG and Sol, respectively.
Changes in MG Architecture During Muscle Stretching and MVC
Lengthening of the MTU during the passive stretch was accompanied by changes in MG architecture (Fig. 4; Table 1). Pennation angle (Fig. 4B) decreased (–30.7 ± 9.7%) linearly (r2 = 0.99; P < 0.001) from 21.2 ± 4.3° at –10° (plantar flexion) to 14.5 ± 3° at +30° (dorsiflexion). Concurrently, Lf increased linearly (r2 = 0.99; P < 0.001) by 40.9 ± 11.6% with ankle dorsiflexion (Fig. 4C). The average length of the MG fascicles was 57.6 ± 8.9 mm at –10° and 80.5 ± 10.1 mm at +30° ankle angle (P < 0.001).
|
|
In a subset of 12 subjects, the changes in MG Lf and pennation angle were measured at rest and during an MVC with the ankle at 0°. In contrast to the changes observed during the stretch, Lf decreased by 28.5 ± 4.2% (from 56.2 ± 6.6 to 40.2 ± 4.7 mm; P < 0.001), and pennation angle increased by 31.5 ± 3.9% (from 20.6 ± 4.8 to 27.1 ± 3.2°; P < 0.01) during the MVC.
Estimation of the Relative Changes in Length During Stretching and MVC
The average LMTU for the ankle in the neutral position was 471.6 ± 35.4 mm. When the ankle angle was rotated from 0° to +30°, the MTU lengthened (496 ± 37.1 mm) by 24.4 ± 1.8 mm, which was an increase of 5.2 ± 0.1% relative to its initial length. The average lengthening of the resolved MG fascicles and aponeuroses corresponded to 17.5 ± 6.4 mm, which accounted for 71.8 ± 27.3% of the total increase in LMTU and meant that tendon elongation contributed to 6.9 ± 6.5 mm or 28.2 ± 27.3% of the increase in LMTU. The tendon contribution to the increase in LMTU was slightly reduced at the most dorsiflexed angles (Fig. 5; Table 2). The average longitudinal shortening of MG fascicles during the MVC when the ankle was at a neutral angle was 19.2 ± 4 mm (4.9 ± 1.1% of rest length).
|
|
Individual Behaviors
When individual values for stiffness and maximum passive torque were examined, there was a bimodal distribution among the subjects with a clear separation around 30 N·m for passive torque (Table 3). In 10 subjects (group 1), the passive torque reached a much lower value (P < 0.05) at the end of the stretch (23.7 ± 5.7 N·m, range: 15–26 N·m) compared with the other six subjects (group 2; 39.6 ± 13.4 N·m, range: 33–51 N·m; Fig. 6, A and B). These differences were observed, despite similar morphological parameters and changes in LMTU for the two groups. There were, however, different muscle and tendon contributions to the increase in total LMTU for the two groups of subjects. The length of MG fascicles and aponeuroses (Lf) in the group 1 subjects increased linearly (r2 > 0.99), whereas the increase in Lt was best fitted by a second-order polynomial equation (r2 > 0.99) that tended to plateau at greater ankle angles (Fig. 6C). In contrast, Lt increased linearly (r2 = 0.97) in the group 2 subjects, and there was less elongation of the fascicles and aponeuroses (Lf) at the maximum angles when passive torque was greater than that achieved by the group 1 subjects (Fig. 6D).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Change in Passive Torque and Muscle Architecture
Consistent with previous in vivo studies, the passive torque produced by the plantar flexor muscles increased exponentially with a progressive dorsiflexion of the ankle (25, 35, 37, 38). At 30° dorsiflexion, passive torque reached 30 N·m, which is similar to the value of 37.8 N·m reported by Kubo et al. at a slightly greater angle (24). However, Morse et al. (38) found a greater value (45.6 N·m) at maximal dorsiflexion. Nonetheless, Kubo et al. (23) reported a passive stiffness of 1.06 N·m/° at dorsiflexion angles between 15 and 25°, which is comparable to an average value of 1.02 ± 0.5 N·m/° calculated in the present study at dorsiflexion angles between 20 and 30°. In agreement with previous studies (12, 30, 38), muscle architecture changed during passive stretching; Lf increased and pennation angle decreased in the MG muscle when the ankle angle was rotated from –10° to +30°. These changes in MG architecture contributed to the increase in whole LMTU during passive muscle stretching.
Tendon vs. Muscle Lengthening
In the present study, the change in Lf recorded in one region of MG was representative of the change in Lf throughout the muscle. Because Lf was similar throughout a wide region of the MG (present study, Refs. 15, 30, 39, 41) and its change in length during passive stretching was similar along the muscle belly (present study), the assumption appears justified. Some studies have shown, however, that strain distribution can differ along the MG aponeurosis during passive stretching in rats (46) or in the Sol aponeurosis during isometric contraction in humans (5, 14).
To determine the contribution of fascicle elongation to the total lengthening of the MTU, it was necessary to take into account the angle of pennation. To estimate the change in Lf, we calculated MG Lf along the longitudinal axis of the MTU by multiplying Lf by the cosine of the pennation angle. The resolved MG Lf increased by 16.8 ± 5.2 mm when the ankle was moved from 0 to 30° dorsiflexion. To estimate the relative contributions of the free tendon and muscle structures to the total increase in LMTU during stretching, we used the regression equation provided by Grieve et al. (9). Our data indicated that, compared with the ankle in neutral position (0°), the total LMTU of the MG muscle increased by 5.2% (24.4 ± 1.8 mm) at 30° dorsiflexion. The change in MG tendon length was calculated by subtracting the change in the resolved Lf from that of LMTU (12, 38). The change in LMTU comprised an estimated relative contribution of 71.8% by the MG muscle structures and 28.2% by the MG free tendon.
The results of the present study contrast with those of Herbert et al. (12) that found most of the MTU elongation during passive stretching to be contributed by an increase in Lt. Furthermore, Morse et al. (38) reported a similar contribution of Lt and muscle length changes to the whole MTU elongation. In addition to a possible difference in the characteristics of the populations of subjects, as evidenced by a lower passive torque in the present study, the relative contributions of muscle and free tendon to the increase in whole LMTU were estimated with different methods. In the study, Morse et al. (38) measured the change in free Lt directly and inferred the changes in muscle (fascicles and aponeuroses) length. In the present study, the change in Lf was measured, and the magnitude of tendon elongation was estimated. Furthermore, the change in ankle angle was inferred from the rotation of the footplate in the current study, whereas it was measured directly by Morse et al.
Regardless of the exact reasons for these divergent results during stretching, our data on Lt during MVCs obtained with the same method produced values (5%) that are consistent with those reported in the literature (1, 35, 40). Our results further indicate that Lt is greater (P < 0.001) during an MVC than during passive muscle stretching (2.9 ± 1.3%), suggesting that the increase in LMTU during passive stretching occurred before the tendon reached its maximal possible elongation (4.9 ± 1.1% during MVC). Although care was taken to keep the heel in contact with the footplate during the MVCs, the increase of Lt may have been overestimated slightly due to ankle joint movement (20). Even with a maximal overestimation of 10–15 mm (20), however, our conclusion remains the same.
Individual Behavior
Our results indicate that the adjustment varied across individuals during passive dorsiflexion and that differences in the relative compliance of the muscle (fascicle and aponeurosis) and tendon structures among subjects influence LMTU differently during passive stretching. This hypothesis is supported by the observation that, although the majority of subjects (62.5%) displayed a linear increase in muscle Lf and a progressive decrease in the rate of Lt with increased dorsiflexion, a second group of subjects (37.5%) showed the opposite behavior. Although the magnitude of the aponeurosis displacement of the Sol was less than the MG during stretching and may have constrained MTU lengthening, the observation of a comparable displacement at 30° dorsiflexion in the two groups suggests that this is not the limiting mechanism. Furthermore, the difference between the two groups of subjects cannot be due to the fact that the increase in LMTU was calculated with a regression equation (9) rather than measured directly. Indeed, the change in LMTU as determined by the two methods produced a difference of only 1.6 mm (n = 3) at 30° dorsiflexion.
Although the morphological parameters were comparable in the two groups of subjects, the passive torque produced at 30° dorsiflexion reached greater values in the second group, and the passive stiffness was about twice that of the first group. A few parameters differed between the two groups at rest and during passive stretching. In the neutral position (ankle at 0°), the subjects of the second group showed only a tendency toward a greater EMG activity in MG compared with those of group 1. At the maximum angle of stretch (30° dorsiflexion), the increase in Lf was smaller in the second than the first group, yet Lt tended to be greater in group 2. As MG EMG activity was significantly greater for the second group compared with the first, the results suggest that a greater part of the passive torque generated by muscle fascicles during passive stretching in group 2 might be caused by tonic reflex activity (11). Because of the weak EMG activity during stretching, the major difference between the two groups of subjects, however, should be mainly the result of variation in the passive structures of the MG MTU (see below).
Possible Structures Contributing to Limit the Myotendinous Lengthening
Several structures may contribute to the difference between the two groups of subjects and the greater increase in passive torque for the group 2 subjects. Among these structures, muscle's connective tissues (endomysium, perimysium, and epimysium) are usually considered to play a major role in the development of passive torque (7). Although all components of the connectives tissues surrounding the fibers and the belly of the muscle contribute to passive resistance when a muscle is passively stretched, the relatively large amount of perimysium is often considered the major contributor to passive torque (42). It seems that the perimysial collagen network, classically referred to as the parallel elastic component, prevents overstretching of the muscle fiber bundles. Among the myofilaments, titin filaments that span from the Z-disks to the M-line may play a similar role and may also contribute to the resistive force during passive muscle stretching (8, 22). Whereas titin filaments appear to bear all passive tension in isolated frog skeletal muscle during lengthening (32), it is difficult to estimate its real contribution in vivo. However, if we assume that the MG muscle contains sarcomeres that are uniformly distributed along the longitudinal axis of the muscle (15, 30, 39, 41), an increase in Lf of 27.5% would induce some resistive force during muscle lengthening (22).
Some of the passive torque during stretching is also generated by the free tendon (28, 40, 45). However, consistent with the conclusions of Bojsen-Moller et al. (3), our observation that tendon elongation is less during stretching than during an MVC suggests that tendon tissues are not a limiting factor of ankle dorsiflexion. Furthermore, the force produced by extramuscular myofascial transmission between synergistic muscles (4, 16) or by joint structures and ligament in vivo (25) may play a role, in addition to tendon and muscular tissues (31), in the increase in passive torque during muscle stretching and particularly when the ankle is close to its maximal dorsiflexion range. Although we do not exclude the possibility that other structures contributed to the drastic increase in passive torque for dorsiflexion >20° in group 2, the comparison between the two groups suggests, however, that lengthening of muscle fascicles and aponeuroses might have been more limiting for these subjects. Further experiments are required to explain more clearly the observed interindividual differences.
In conclusion, the present study showed that the increase in passive torque during passive stretching of the plantar flexor muscles is accompanied by a progressive lengthening of both MG free tendon and muscle structures (fascicles and aponeuroses), with the latter showing a greater degree of elongation. The different behaviors among subjects, however, indicate that the relative strain of muscle and tendon structures varies among individuals, and, consequently, the relative compliance of these structures influence MTU lengthening differently during passive stretching.
|
GRANTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. Nordez, T. Gallot, S. Catheline, A. Guevel, C. Cornu, and F. Hug Electromechanical delay revisited using very high frame rate ultrasound J Appl Physiol, June 1, 2009; 106(6): 1970 - 1975. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
