14 J. WILLIAM BUCHANAN 



liquids were poured off and replaced by fresh. The above opera- 

 tions of filling and taking samples were then repeated as often 

 as necessary. In experiments requiring a record of the rate of 

 oxygen consumption over a considerable period of time, the re- 

 quired samples were taken after the first ten to twenty hours 

 after cutting. The flasks were then set aside at room tempera- 

 ture, the solutions changed frequently, and on designated days 

 the rate of oxygen consumption over a certain number of hours 

 was taken in the same manner as that of freshly cut pieces. 



The oxygen content of each sample was measured by the 

 Winkler method. This method has been so widely used that 

 no extended discussion or account of it is necessary here. Due 

 precautions were exercised to exclude casual air or other sources 

 of oxygen from the system. With slight modifications, the methods 

 of analysis described by Birge and Juday ('11) and modified 

 by Hyman ('19 a) were followed. Data presented by Hyman 

 show that with apparatus for obtaining samples very like my 

 own the error introduced is not greater than two hundredths of a 

 cubic centimeter when well-aerated water is used, and since 

 in all cases in this work well aerated water or solutions were 

 used, the error is considered to be plus or minus two hundredths. 

 When the rate of oxygen consumption per milligram of weight 

 of the animals is calculated, this error becomes very small in- 

 deed, and variations in results must be considered to be due to 

 variations in the physiological condition of the animals from 

 day to day, due to temperature changes in the laboratory and 

 other possible sources. The titrations were not affected by 

 the presence of the anesthetics, so far as can be determined. 

 As Hyman has shown ('19 a), the planarians add something to 

 the water which decreases its iodine-absorbing power, and the 

 oxygen content of water in which the planarians have been living 

 measured by this method is not really exact. Whether or not 

 the pieces add more of this substance or less when subjected to 

 an anesthetic, I do not attempt to say. Certainly, the error 

 thus introduced must be very small when compared with the 

 relatively great differences of the oxygen content of samples 

 from the flasks containing water and those containing solutions 

 of anesthetics. 



