CRITICAL DISCUSSION OF RESPIRATION APPARATUS. 235 



allowable in exact research work. Loewy 1 has made an investigation 

 to find whether or not the respiration and the respiratory exchange 

 are markedly affected by breathing through the Zuntz-Geppert respi- 

 ration apparatus during muscular work. In this study he measured 

 the respiratory exchange after a certain amount of muscular work had 

 been done and also carried out experiments in which the respiratory 

 exchange was measured both during the working period and in the 

 period after the work had been completed. He found that the respi- 

 ratory exchange in the period after work was not affected by breathing 

 through the respiration apparatus during the period of work. Grafe 2 

 has called attention to the criticisms of the Zuntz-Geppert apparatus 

 and Roily 3 and Jaquet 4 have also referred to the abnormal results which 

 are sometimes obtained. 



One of the criticisms brought against the Zuntz-Geppert method is 

 that the sampling is not proportional ; in other words, that the sample 

 of air does not represent the true average. In experiments which 

 were made in this laboratory a known amount of carbon dioxide was 

 introduced into a current of air passing intermittently into the meter 

 and it was found that the percentage of carbon dioxide recovered in the 

 sample taken by the proportional method agreed well with the per- 

 centage calculated. The experiments were somewhat difficult to carry 

 out, as it was not easy to arrange for the intermittent delivery of carbon 

 dioxide into the air-current in such a manner that it would be readily 

 calculated or determined. The air can be mixed to some extent 

 before it reaches the sampling tube by inserting a large bottle or flask 

 between the expiration valve and the entrance to the Elster meter. 

 In experiments in which the expired air changed rapidly in composition 

 such a procedure would be a disadvantage, as there would be a dead 

 space through which the air would have to pass before a sample was 

 taken. The mixing and sampling would thus lag behind the changes 

 in the expired air. This is true even with the present arrangement, as 

 the long tube between the man and the gas-meter is a dead space 

 which must be swept out before the sample drawn through the tube 

 will actually represent the composition of the air. This lag plays no 

 great role unless the experimental periods are of extremely short dura- 

 tion and in periods of 15 or 20 minutes it is not of much importance. 

 Geppert and Zuntz 5 point out that the capacity of the tube from the 

 valves to the sampling device should always be greater than the maxi- 

 mum expiration likely to occur during an experiment. 



The Zuntz-Geppert method of proportional sampling was checked by 

 Geppert 6 with a rabbit, all of the carbon dioxide produced being ab- 



, Archiv f. d. ges. Physiol., 1891, 49, p. 492. 

 "Grafe, Abderhalden's Handbuch der biochemischen Arbeitsmethoden, 1913, 7, p. 459. 

 'Roily, Deutsch. Archiv f. klin. Med., 1908, 95, p. 75. 

 4 Jaquet, Ergebnisse der Physiologie, 1902, 2, p. 457. 

 'Geppert and Zuntz, Archiv f. d. ges. Physiol., 1888, 42, p. 189. 

 6 Geppert, Archiv f. experimentelle Path. u. Pharm., 1886-87, 22, p. 373. 



