METHODS OF MEASURING RESPIRATORY EXCHANGE 49 



Another titration method worked out by Schiitzenberger and Risler 

 is somewhat more complicated, but can be used on water containing 

 organic impurities (see Henze [1910]). 



Carbon dioxide is determined by boiling it off from a sample of 

 the water after acidification and in a current of CO 2 free air. The 

 carbon dioxide boiled off is generally taken up in baryta solution and 

 titrated (Warburg [1909]). 



The determination of the carbon dioxide liberated in respiration 

 experiments is especially difficult and uncertain, partly because the 

 quantity in question is generally a small fraction only of the total 

 quantity present in the water, and partly also because the animals ex- 

 perimented on are often likely to give off substances (urine, faeces and 

 mucus) which will yield some carbon dioxide when the water is boiled 

 after acidification. It is certainly best, therefore, to acidify slightly 

 only, and to drive the carbon dioxide off by a current of air at a 

 comparatively low temperature, but even with this precaution too 

 high results for carbon dioxide are often unavoidable. 



A very serious difficulty and source of error in experiments on 

 aquatic respiration is the action of bacteria as pointed out by Knauthe 

 [1898]. It is generally impracticable to make experiments under 

 aseptic conditions, but when the water employed contains few bacteria 

 only it will take a certain time before they multiply so far as to 

 influence the results seriously. Winterstein [1909] found that when 

 pure sea water is kept in darkness at ordinary temperature without 

 any animals in it, the oxygen content decreases steadily, during the 

 first days about OT c.c. per litre per day. 



It has been shown by Winterstein [1908] and Henze [1910, 2] that 

 the respiratory exchange of most aquatic animals is practically unaf- 

 fected even by considerable variations in the oxygen content of the 

 water. This fact is of importance methodically and facilitates con- 

 siderably the determination of the oxygen absorption. 



While satisfactory respiration experiments cannot be made on air- 

 breathing animals by the simple method of determining the changes 

 in composition of a known volume of air in which the animal experi- 

 mented on is confined, because the rise in carbon dioxide percentage 

 very soon makes the respiration abnormal and influences the respiratory 

 exchange, the method of enclosing an aquatic animal in a known 

 volume of water, the gases of which have been determined, and analys- 

 ing the water after a certain time, is as excellent as it is simple. The 

 slight increase in CO 2 tension produced when J - J of the oxygen is 



4 



