98 J- LlNttHARD. 



Priestley, it is at once obvious, that this is not taken from the 

 tidal air but simply and solely from the reserve air, and as the 

 maximum of the carbonic acid under a normal respiration occurs 

 at the end of the expiration, this sample must necessarily give too 

 high percentages for the carbonic acid. Expiration of the reserve 

 air, as pure and simple active movement, can even less than the 

 ordinary expiration occur momentarily; it cannot do so when one 

 breathes out into the air, still less when the breathing takes place 

 through a tube, which, as it is to be easily closed by means of the 

 tongue, must be of comparatively small bore. 



The objections, which I believe we may raise against Haldane 

 and Priestley's absolute numbers, do not necessarily affect the 

 results, however, so far as the variations are concerned. That an 

 experimental series is affected by a systematic error, does not make 

 it unusable, if we only know the direction and size of the error. 



There is, however, still another source of error, the effect of 

 which is much more difficult to estimate; as pointed out also by 

 Hasselbalch, the method presupposes, that the depth of the single 

 respirations is quite the same, especially that no superfluous or 

 unusually deep expiration immediately precedes the taking of the 

 sample. The uncertainty which may arise in this way, is not small; 

 in Haldane and Fitzgerald's paper above-cited we several times 

 find differences of up to 1 "/o CO 2 in twin-determinations. This 

 uncertainty becomes obvious when we calculate the "dead space" 

 from the directly determined values. In one series given by Hal- 

 dane and Priestley the highest value for this is twice as great as 

 the lowest. And even if the "dead space" is not an invariable 

 quantity, not even in the same individual, yet such a series as mine 

 given below, where the "dead space" enters into the calculation as 

 a constant, will hardly be thinkable, if variations only approximately 

 as large as those noted by Haldane and Priestley were possible. 



As too high a percentage of carbonic acid gives on calculation 

 too high values for the "dead space", and as this is always estimated 

 considerably higher by Haldane and Priestley than by other 

 investigators, who for the rest work upon different methods, I see 

 in this a support for my view that the method of Haldane must 

 give too high values for the percentage of alveolar carbonic acid. 



To test the agreement between the two methods as also to obtain 

 a basis for the comparison of my investigations with those made 

 by Haldane and his co-workers, 1 have undertaken, as mentioned, 

 a series of investigations of my alveolar air according to Haldane's 

 method; the results of these observations are contained in the tabular 

 summary given below. 



