1919] 



DODGE — COLORIMETER 



63 



curacy of the determinations made near its limits of brilliancy 

 by another indicator with slightly overlapping color change. 

 Moreover, it is often desired to use a particular indicator 

 at or near the limits of its usual range. Thus methyl red, 

 extremely serviceable between P H 4.4 and P H 6.0, just fails 

 to completely cover the range of certain nutrient solutions 

 and plant juices frequently employed in the culture of fungi. 

 The limitation, however, is really in the ability of the unaided 

 eye to detect readily the slight differences when a certain 

 redness (or yellowness) is approached. The difficulty of 

 color in the medium under investigation, however, is the most 

 serious. In our work the indicator solutions have been pre- 

 pared by using the quantities recommended by Clark and 

 Lubs in 50 per cent ethyl alcohol. These are preserved in 

 amber dropping bottles. 



When a careful technique is established the degree of ac- 

 curacy sufficient for all practical purposes is assured in the 

 examination of colorless solutions by the following pro- 

 cedure: Small test-tubes containing a measured quantity 

 (usually 5 or 10 cc.) of the standard solutions are arranged 

 in open racks provided with a white paper background. A 

 series is prepared for each indicator employed, and the P H 

 values may differ by .1 or .2, depending upon the accuracy 

 required. A definite and constant quantity of the indicator, 

 usually 2 or 3 drops, is placed in each tube. The same quan- 

 tity of the sample or test solution is placed in a similar test- 

 tube and the indicator added as before. The samples are then 

 compared with the various standards in a uniform light and 

 an exact match is obtained. A characteristic of most fluids 

 or media with which the physiologist deals is color, and this 

 has operated in the past more or less to interfere with the 

 correct determinations by the indicator method. 



Early investigators were disturbed by the presence of color 

 in the solutions studied, and various methods were employed 

 to counteract this source of error. Sorensen ( '09-10) pro- 

 posed a method of dealing with colored solutions, whereby the 

 natural test solution was matched in color by means of neutral 

 dyes used in the standard solutions, before the addition of 



