LOCAL OXYGEN TENSION 411 



When used in blood, the electrode should have its recess filled with 

 agar containing salt to keep the red cells out of it, as these would 

 elicit abnormally high currents by virtue of their function as oxygen 

 sources. 



The recessed electrode has been found to give calibration constants 

 which are reproducible to ±3% over a period of weeks. Continued 

 exposure to tissue fluids for some hours will result in a drift in the 

 calibration to yield oxygen tension values which are too low. How- 

 ever, an equal exposure to 0.15 M sodium chloride will bring the 

 values back to the correct level. Davies and Brink suggest that this 

 difficulty may be obviated by calibration in the tissue fluids. Later 

 they found that by filling the recess with distilled water and then 

 covering^ the tip of the electrode with a collodion membrane, the 

 drift in calibration during exposure to tissue fluids can be made 

 very small. 



Open Electrodes. The open type of microelectrode is prepared 

 by fusing a platinum wire, 25 /j, diameter, into a soft-glass tube so 

 that one end of the wire is flush with the sealed end of the tube. 

 While the open electrode cannot be used for the measurement of 

 absolute oxygen tension, it can serve to measure rapid changes in 

 tension. Thus, Davies and Brink were able to record, to about 0.1 

 sec, the sudden oxygen consumption occurring when a muscle fiber 

 contracts. 



The variable properties of the open electrode make for consider- 

 able instability in its performance. However, polarization for several 

 niin. at 1.0-1.2 volts effects some stabilization; even so the repro- 

 ducibility is only about 15% under favorable circumstances. Cali- 

 bration before and after each experiment will hold the error to a 

 minimum. Poorly defined plateaus are obtained in the current-poten- 

 tial curves, but as a rule with up to 0.8 volt nothing but oxygen is 

 electrolyzed in oxygen-free solution. A linear relation may be found 

 between current and oxygen tension in calibration experiments; 

 however, when applied to tissue, particularly near blood vessels, 

 there may be little correlation between the actual tension and that 

 derived from a calibration curve which is obtained with a solution 

 such as Ringer's. 



