THE CONTROL OF THE RESPIRATION 389 



as in the previous experiment, the periodic nature of the respiration is 

 immediately changed to the regular breathing if 2 is introduced into 

 the tube. The interest of the experiment lies in the fact that a similar 

 relative elongation of the dead space is probably accountable for the 

 periodic breathing seen in the winter sleep of hibernating animals. Dur- 

 ing this condition, on account of the depression of metabolism less 2 

 is required and less C0 2 is produced, so that the exchange of gases 

 through the pulmonary endothelium is greatly diminished. The dead 

 space, however, remains of the same capacity, which amounts to the 

 same thing as if the latter had been prolonged under unchanged con- 

 ditions of pulmonary gas exchange. 



Important evidence that changes occurring in the tensions of 2 

 and C0 2 in the alveolar air, and therefore in the arterial blood of 

 the respiratory center, are largely responsible for periodic breathing 

 has been secured by studying the condition that develops after a period 

 of apnea produced by voluntary forced breathing. The results of such 

 observations are given in the curve shown in Fig. 131, in which 

 the thin line represents the 2 tension of the alveolar air, the thick 

 line the C0 2 tension. The double line running across the chart repre- 

 sents the average tension of C0 2 during quiet normal breathing. The 

 respiratory movements are represented by the tracing at the foot of 

 the curve, along the abscissa. It will be observed that the oxygen ten- 

 sion falls very rapidly during the apneic period, until just before breath- 

 ing recommences it may be as low as 30-35 mm. Hg instead of the nor- 

 mal of about 95. Meanwhile the C0 2 tension rises from the very low 

 level of 12 mm., at first very rapidly, then more gradually, although, 

 when breathing recommences, it has not yet gained the normal level. 

 As a result of the first periods of breathing, the 2 tension suddenly 

 increases, but the C0 2 falls only slightly. During the next apneic stage 

 the 2 quickly comes down again, and the C0 2 rises so as almost to at- 

 tain normal tension before breathing again supervenes. As the apneic 

 periods subsequently become less pronounced, the C0 2 tension comes to 

 stand almost at its normal level, whereas considerable variations in the 

 2 tension continue to occur. If the lungs are filled with oxygen by in- 

 haling the gas with the last few deep breaths the return of breathing 

 is delayed and it is not periodic in character. 



Besides affording substantial support to the hypothesis which was 

 stated above, there are several other interesting features of these results 

 which demand attention. In the first place, it is plain that the body 

 is possessed of some mechanism by which it can prevent great fluctua- 

 tions in the C0 2 tension of the blood, whereas towards 2 no such 

 "buffer action" is displayed. It will further be observed that the CO, 



