LETTERS TO THE EDITOR

measurements to the nearest millisecond may well imply unjustified precision when using a record-ing paper speed of 100 mm/sec and time lines of 10 msec, but it is unlikely that the precision of our data would be enhanced by measuring to the nearest 5 msec on paper with 20 msec time lines, as recommended by Weissler and co-workers,1 or to the nearest 5 msec on paper with 40 msec time lines, as reported by Aronow et al.2 The STIs obtained immediately following exercise were corrected for heart rate using regression equations developed in resting supine subjects.1 3 Such an approach has been used by other workers in this field,4 and it is our opinion that such corrections would at least partiall