240 REPORTS ON THE STATE OF SCIENCE.—1914. 
Professor Macdonald also finds analogies between these results and 
those of walking experiments described by Douglas and Haldane in 
‘Phil. Trans.’ B, ciil., p. 245. A full consideration of the matter will 
be found in a paper communicated, June 1914, to the ‘ Proc. Roy. 
Soc.’ on the ‘ Mechanical Efficiency of Man.’ 
The section of the work dealing with the respiratory changes has 
also been continued during the past year. A number of experiments 
have been performed in which the respiratory interchange of a man, 
doing a measured amount of mechanical work upon a cycle in the 
calorimeter, has been investigated. The calorimeter is ventilated by 
means of a stream of air drawn through it at a uniform rate by a 
pump and measured by a meter placed on the distal side of the latter. 
All three are connected by tubing, through which the air flows, and 
the air as it leaves is sampled by suitable means every ten minutes. 
The samples thus obtained are examined by the gas-analysis apparatus 
devised by Dr. Haldane, and the carbon dioxide and oxygen percentage 
determined. ‘The carbon dioxide figures, when plotted against the time 
on squared paper, take the form of a curve rising steadily to a horizontal 
asymptote. 
In order to understand the figures thus obtained it was obviously 
necessary to inquire into the question of storage of gases within the 
calorimeter, and to do this a number of calibration experiments (17 in 
all) were made, in which a stream of carbon dioxide, measured by a 
gas-meter and generated by a modification of the apparatus described 
in a paper by Young and Caudwell (‘ Soc. Chem. Ind.,’ March 1907), 
was passed into the calorimeter at a uniform rate, and the ten minutes’ 
samples examined in the manner described in the experiments on the 
human subject. Attempts were then made to discover the relation 
which exists between the curve of the carbon dioxide in the leaving air 
and that of the carbon dioxide introduced from the generating appara- 
tus; but so far the results appear so complicated that no definite relation 
has been arrived at. However, quite recently Mr. G. H. Livens, M.A., 
Lecturer on Mathematics to the University, has rendered most valuable 
assistance towards solving this problem. A fairly accurate empirical 
formula has been obtained from the actual readings, but it is not of 
such good agreement with the theoretical formula as is desired, and 
further experiments are being made to detect the cause of the dis- 
crepancy. 
Owing to the appearance of a considerable error in the readings of 
the large meter used for measuring the volume of the air-flow through 
the calorimeter, it became necessary to replace it by a water-meter 
supplied by Messrs. Parkinson and Cowan, Ltd. Also, a large number 
of tests were made, both in the Physiological Laboratory and through 
the courtesy of the Sheffield Gas Company at their test-room, on the 
small meter which is used for measuring the volume of carbon dioxide 
introduced into the calorimeter in the calibration experiments men- 
tioned above. I am now certain that the error in our estimation 
on these accounts 1s well under 2 per cent. 
