152 MUSCULAR WORK 
uneventful. The pulse-rate in the third period rose rapidly, the records being: 
10 h 50 m a.m., 124; 10 h 52 m a.m., 128; 10 h 54 m a.m., 130;' 10 h 56 m a.m., 132; 
10 h 58 m a.m., 136; at which time the experiment was discontinued. 
As the result of a careful examination of the results of the experiments in 
which the indications of collapse appeared, it was decided that these symp- 
toms were due to an accumulation of carbon dioxide in the ventilating system, 
caused either by using defective soda-lime or too slow a rate of ventilation. 
Furthermore, the possible discomfort and strain on the part of the subject 
in holding his head in the position required by the rigid brass pipe to which the 
mouthpiece was attached should if possible be avoided. Several modifica- 
tions were therefore made in the apparatus and in the experimental routine. 
Since these have already been noted in the description of the apparatus 
(see p. 29), a brief description will suffice here. 
To insure complete absorption of the carbon dioxide by the soda-lime, 
the composition of this absorbent was somewhat modified in accordance with 
the results of a series of experiments which had recently been made for test- 
ing it; furthermore, two soda-lime containers were used in series so that any 
carbon dioxide which might pass the first container would be completely ab- 
sorbed in the second; finally, the efficiency of the soda-lime was tested in 
every experiment by deflecting a small part of the air-current after it passed 
the carbon-dioxide absorbers and sending it through a solution of barium 
hydroxide before returning it to the system, the presence of excess carbon 
dioxide being readily detected by this method. Since it was found that the 
rate of ventilation was not rapid enough to remove the expired air as fast 
as the subject breathed into the pipe, a larger pulley was placed upon the 
armature shaft of the motor by means of which the rate of ventilation was 
increased from 35 to 60 liters and later to 85 liters per minute, this being 
found amply sufficient. To give more freedom of movement for the head in 
riding, a flexible tube was attached to the valve-piece through which the 
subject breathed. As this addition increased the dead air-space, a supple- 
mentary piece of rubber tubing was attached and the air-current was deflected 
through this so that it passed directly by the mouthpiece, thus eliminating 
the possible effect on the respiration of so large a dead air-space. This modifi- 
cation of the apparatus is clearly shown in fig. 3, page 29. After these changes 
in the apparatus and in the experimental routine had been made, practically 
all of the experiments were performed without any indications of collapse 
except on one or two occasions when the subject suffered some distress owing 
to the use of a defective soda lime bottle. 
In the so-called fatigue experiment with M. A. M. on March 15, a col- 
lapse seemed imminent, although not actually occurring; this, however, 
was not in any way due to the presence of excess carbon dioxide in the ven- 
tilating system or to deficient ventilation, for not only was the absence of 
carbon dioxide proved by the barium hydroxide test but the amount of oxygen 
in the air was determined and found to be 22.5 per cent. The impending 
collapse was unquestionably due to the excessive muscular activity and the 
onset of fatigue. 
Possible criticism may be made that the collapse in the earlier experi- 
ments was due not to an accumulation of carbon dioxide in the system but to 
a lack of oxygen. This was not true, as an analysis of the air in the apparatus 
