A second important point is the pronounced increase in pulse-rate prior to riding on days with full diet in February which is likewise reflected in the increase noted after the first minute. Coincidental with this is an almost invariable lengthening in the time required for the return to normal.

A more complete analysis of the situation is given in table 105, in which we have summarized for the individual subjects and for each day of Professor Johnson's observations the length of time required for the pulse-rate to return to normal. Although it is somewhat inconsistent to strike a general average, an examination of the average figures for reduced diet near the bottom of table 105 shows that the shortest average period was found with Vea, who returned to the normal inside of 4.0 minutes; Pec reached normal in 4.3 minutes, Can in 5.1 minutes, and Pea in 5.9+ minutes. Of these men, Pea and Pec were trained athletes, while Vea and Can were distinctly of the non-athletic type. The records shown for Vea are of special interest as indicating how regular his return to normal usually was. With the restricted diet there were only three occasions when he required more than 5 minutes to reach the normal level. The one instance in which 10 minutes was required occurred on February 11 after the restriction in diet had ended. On the other hand, with Tom, an entirely untrained man, an average value of 16.7+ minutes was obtained for the return to normal, the individual data remaining very long throughout the observations. It is perhaps somewhat surprising that the high values of 13 and 12 were noted with Kon, Mon, and Moy, who were all physically well-trained men. Mon, particularly, shows a consistently long time for his return to normal; in only three instances was less than 7 minutes required.

A general inspection of the table seems to indicate that the longer periods for the return to the normal level occur with practically all the men about November 9 to December 7. A consideration of the figures in the last column at the right shows that this is borne out by the daily averages for the whole squad. On the other hand, most of the squad showed a long period of return on February 8 to 18, when the body-weight was rapidly increasing (see table 104) and a greater amount of nourishment was being received than during January, when the men were on a restricted diet and the weight was below normal. The average time for January 31 (a low-diet day) was 6.2 minutes. On this day the longest time required for return to normal was shown by Mon (11 minutes) and Moy (12 minutes). With the resumption of full diet, the time for return to normal increased, being on the average, 11.2 minutes on February 8, 11.8 minutes on February 11, 7.7 minutes on February 15, and 7.5 minutes on February 18. Although the number of men included in the average is not always the same, the figures give a clear indication not only that the time lengthened when the men returned to full diet, but that on the later days in February there was a tendency with most of the subjects for the pulse-rate to return to normal a little more rapidly than during the first days of full diet.

Table 105. - Time Required For Pidse-Rate To Return To Normal After Work On Bicycle Ergometer - Squad A, Subjects In Lying Position, With Food

1 These averages do not include the values for Spe and Tom.

2 More than 20 minutes required for pulsa-rate to return to normal; the averages which include these records are indicated by the plus sign.

An interpretation of the true effects of the low diet upon the return of the pulse to normal after work is, from the data with Squad A (table 105), extremely difficult. At the period in November and December, when the body-weight was approaching a minimum and the caloric intake was very low, we find some of the highest figures, that is, the longest delayed return to normal. After the return of the men from the Christmas vacation, i. e., during the month of January, the squad as a whole reached a level very considerably lower. The average values for January are, however, vitiated by the fact that both Spe and Tom were not in the squad and the pulse-rate with both of these men had previously required a long time to return to normal level. The figures for this month are distinctly at variance with those for November and December, and but few direct conclusions can be drawn. The possibility of a training should not be lost sight of. In the absence of clear-cut evidence we can only consider the difference between the return to normal of the pulse-rate during the month of January and that found subsequently after the resumption of feeding. Here the picture is very clear, namely, an increased lengthening of the time required in the latter period, but the situation is also complicated greatly by the ingestion of a large amount of food with pronounced katabolic processes, involving the deposition of fat and a general increase in weight. Unfortunately, as will be subsequently seen, reference to the data obtained with Squad B does not materially clarify the situation.

The correlation between the rapidity of return to normal of the heart rate and the physical condition of man has been especially studied by Pembrey and his associates.1 His general conclusion is that with the untrained man the return to normal is longer than with the well-trained man. "The pulse of the trained man has a slower rate at rest, a wider range in response to muscular work and a more rapid recovery after exercise".

The picture for Pea and Pec and the delayed return with the entirely untrained man Tom are in full conformity with Pembrey's views, but it is difficult to account on this basis for the rapid return to normal of Vea and Can, who were untrained men, and the delayed action of the well trained men, Mon, Moy, and Kon.

1 Pembrey and Todd, Journ. Physiol., 1908, 37, Proc. Physiol. Soc., Oct. 17, p. lxvi. Pembrey's work is best summed up in his article with Cook, Journ. Physiol., 1913, 45, p. 446.