Subjects
10 male subjects with normal hearing and ages ranging from 21 to 33 years (average + S.D. 25.2 ± 3.2 years) formed the group of pranayama. 10 more male subjects (age 22 to 42 years, mean ± S.D. 31.7 ± 6.8), who had no training in pranayama or in concentration exercise, were also assessed for detecting effects of attention maintained on the nasal air flow and on ongoing tidal respiration that was not controlled consciously. The study was explained to the subjects and their signed informed consent taken.

Before commencement of training in pranayama, each subject of the pranayama group was tested in three sessions conducted usually at several days apart, to assess baseline variations and changes in middle latency auditory-evoked potentials (AEP-MLRs) occurring in successive periods of the 20-min session. In each session of 20 min, a total of four recordings were made during the following time periods while the subject sat awake and relaxed with eyes closed:
(i), one recording in the first 5-min period, i.e., in pre-test eyes closed baseline condition (PEC); (ii), two successive recordings in the next 10-min period, i.e., in eyes closed ‘control’ sitting without pranayama (ECC) and (iii) one recording in the last 5-min period, i.e., in eyes closed post-test condition (ECP). After acquiring data as described above, the subjects were introduced to the training in pranayama practice: 6 of them in the ujjayi type of pranayama, and 4 in the bhastrika type. After practising the pranayama for an average of 21.8 months (range 18-26 months), the subjects were reassessed in test sessions conducted to record the AEP-MLRs in the same temporal sequence of test periods mentioned above, except that the subjects practiced pranayama during the time-period PR, that corresponded in time to the ECC cited above.

Stimulation and recording settings
The AEP-MLRs were recorded from Cz-A1. In each time-period cited above, 1500 responses in the post-stimulus 70 ms time-period were averaged on a Nicolet (USA) Model Med-80-4 computer system. The preamplifier bandwidth was set at 1-1500 Hz. Click stimuli of 40ms duration, with alternating polarity, were delivered at 5 Hz binaurally through acoustically-shielded ear-phones. The intensity was set at 40 dB above the threshold of hearing, which was about 24 dB under the above mentioned stimulus parameters. It was kept moderate enough so as not to disturb the concentration of subject, and yet be adequate to evoke the potentials with consistency. Subjects were comfortably seated during recording sessions in a dimly lighted (twilight), air-conditioned and sound-attenuated cabin. and were watched throughout via a closed circuit low-light-sensitive television.

Measurements and statistical assessment
Peaking latencies of the waves were measured from the moment of stimulus delivery. Peaking amplitudes of the waves were measured taking the level existing at the time of stimulus delivery as the baseline. Averages of the peaking amplitudes and latencies of Na, Pa and Nb waves of the three repeat sessions of a subject were calculated for each time period. Since recordings were made twice in the time period ECC or PR of each of the three sessions. the average of the time period ECC or PR was based on six values. Such averaged values obtained under the conditions of ECC or PR, for the 10 subjects were subjected to the analysis of variance (Snedecor and Cochran, 1967; Zar, 1984) and to the matched pairs t-test, to infer whether significant changes occurred in the peaking of waves during pranayama. Also, the baseline data (of PEC time period) obtained before and after undergoing pranayama training were compared.

Since significant changes are observed for only Na wave, data of Pa wave are presented to only a limited extent, and of Nb wave not detailed in this paper for saving space.

Pranayamic methods
Utmost concentration has to be maintained during the following pranayamas. The subject is asked to concentrate on that point in the nasal passage where the inhaled air is first felt. This training in concentration is claimed to create, in the long run, a steadying effect on the mind (Behanan, 1937). During the pranayama, the subject sits with spine erect. In the ujjayi type of pranayama after a slow but complete exhalation through the left nostril, the next breath is inhaled through both nostrils at a uniform pace, with a partially closed glottis, and the abdominal muscles kept in a state of partial contraction. Inhalation is followed by timed breath-holding (the kumbhak phase), during which both nostrils are closed with the right hand, the glottis is also tightly closed, and the chin is lowered to touch the jugular notch (jalandhar bandha or chin lock). At the end of the kumbhak phase, the head is raised back to the normal level, the glottis is opened partially and the subject exhales through the left nostril, while the abdominal muscles go on contracting. This is followed by a phase of external breath-holding during which the abdominal muscles are specially
 

Fig.1 Illustration of typical breathing cycles in 2 types of pranayamas. (A), ujjayi type, during which the duritatio of expiration (E) is typically twice that of inspiration (1), and there is a timed breath-holding at end-inspiration (K-1), and also at end-expiration (K-E). Such cycles of breathing are repeated at about 2/min, (B), bhastrika pranayama, in each cycle of which a cycle of which a cycle of ujjayi type of breathing is preceded by a set of rapid breathing cycles of shallow breaths (KPB). The figure also illustrates for comparison the tidal level of breathing pre and post-pranayamic periods, TR, transition phase between two bhastrika cycles.

held contracted (uddiyan bandha or abdominal lock). This comprises one cycle of ujjayi which is followed by the next (Fig. 1A). Proficiency is acquired through stages of practice; the first stage involves only inhalation and exhalation in the time (duration) ratio of 1: 2, later on inhalation, timed internal breath holding, and exhalation are practiced in the ratio of 1: 1: 2 and still later on with the ratio of 1: 2: 2. In the last stage, ‘external’ breath-holding is introduced, so that the ratio becomes 1 : 2:2: 1. Six of the subjects of this study (RMC, RNG, SNT, SK, SKT and VD) had attained this level of accomplishment in the ujjayi pranayama practice. The average durations ± S.D. (in seconds) of the 3 phases, viz., inspiration, ‘internal’ breath-holding, exhalation, and ‘external’ breath-holding were 3.9 ± 1.0, 10.0 ± 3.0, 10.4 ± 1.5 and 5.0 + 1.0, respectively. The average respiratory rate of these six subjects during pranayama was 2.4 ±0.8 cycles/ min. The four remaining subjects (PRJ, SJR, STP and NVR) practiced a modified version of the pranayama (Fig. 1B). In this, each breathing cycle began with exhalation (at the end of tidal breathing), followed by very shallow, rapid breathing at the rate of approximately 2 breaths/s for a total of 5 s, on average. During this part of rapid breathing (which is known as kapalabhati), the amplitude of respiration was usually only 1/5th of the tidal breathing. This is followed by a cycle of ujjayi pranayama, with inhalation, ‘internal’ breath-holding, and exhalation in the duration ratio of
1: 4:2. There was no phase of ‘external’ breath-holding. The average durations ± S.D. (in seconds) of these three phases were 5.5 + 0.7, 22.0 ± 2.5, and 8.0 ± 1.5, respectively. During the ujjayi part of the pranayama, the average respiratory rate was 1.6 + 0.5 cycles/min. This combination of a brief kapalabhati part and a cycle of ujjayi constitutes one unit which is repeated as long as the session lasts. The glottis is kept open during the kapalabhati part, and partially or fully closed during the ujjayi part as cited above. Such a combination pranayama is called the bhastrika type.

The subjects were expected to keep up their proficiency by practising the pranayama daily for about 20 min, except for any occassional lapses.

Their proficiency was checked periodically by a pranayarna expert (teacher) to confirm that the subjects were maintaining their practice.