174 STXJDIES ON RESPIRATION. I 



to place the tube in a hole bored in a block of wood and to view it 

 through another hole bored at right angles to the first. The last 

 method is very useful when there is a color (due to small suspended 

 organisms or to the giving off of coloring matter by the organism) 

 which is superimposed on that of the indicator. This color is con- 

 tained in the tube which is put into the wooden block in line with 

 the tube containing the buffer solution, so that the light passes 

 through both tubes before reaching the eye. In this way the color 

 of the indicator can be varied independently of the disturbing color, 

 which is due to the organism. 



Uniform conditions for comparison of colors were secured by the 

 use of the "Dayhght" lamp.^ This gives a uniform source of light 

 under which colors can be matched with sufficient accuracy, so that 

 the investigator is free from disturbances due to fluctuations of 

 daylight. 



The best test of the accuracy of the work is to make repeated meas- 

 urements on the same material in its normal environment at constant 

 temperature. Tests made with favorable material show that the 

 variation is very small, the probable error (in the time required to 

 produce a standard change in acidity) being less than 1 per cent. 

 Hence if greater variations are found, they must be due to the varia- 

 bility of the material, to the personal equation of the observer, or to 

 unfavorable conditions {e.g. of illumination) in comparing colors, 

 rather than to the method. A probable error of less than 1 per 

 cent must be regarded as highly satisfactory for biological meas- 

 urements. 



The results obtained by the use of the new apparatus^ are quite as 

 accurate as those obtained by the original method.^ 



Attention may be called to two things in respect to the curves: 

 (1) After each measurement there is an interval during which the 

 solution is changed; since the organism is exposed to the action of the 

 reagent during this interval the time should be included in the total 

 time of exposure (as shown in Fig. 1). (2) The rate is obtained by 

 taking the reciprocal of the time required to produce the standard 



^ Cf. Luckiesh, Science, 1915, xlii, 764. The form used is known as "north 

 Ught." 



