XIII, A, 1 Heise and Behrman: Water Analysis in Field 3 



length aluminium tube, which is filled with the water under ex- 

 amination. The color of this column of water, viewed longitud- 

 inally, is matched by disks of colored glass that have been 

 rated in parts per million to correspond to the platinum-cobalt 

 standard. 



Iron is conveniently determined with the same outfit as used 

 for color, as described by Leighton. The only extra equipment 

 required is a series of prepared colored disks corresponding 

 to those produced by treating standard solutions of iron. These 

 disks have not been available. In lieu thereof, red and yellow 

 glasses from the Lovibond tintometer have been employed, in 

 connection with two matched Nessler tubes in galvanized iron 

 outer tubes. When 100 cubic centimeters of water were used 

 in a determination, it was found that a summation of 6.0 on 

 the Lovibond scale was very nearly equal to 1 part per million 

 of iron (as Fe). The following is the procedure employed: 



To 100 cubic centimeters of the water under examination in 

 a Nessler tube add 4 cubic centimeters of concentrated nitric 

 acid. Mix thoroughly by pouring six or seven times from one 

 tube to another and allow to stand at least five minutes to insure 

 complete oxidation. Then add 6 cubic centimeters of a 2 per 

 cent solution of potassium sulphocyanide, mix thoroughly by 

 several pourings, and allow to stand ten minutes for the color 

 to develop. Exactly at the end of ten minutes make the color 

 comparison with the Lovibond glasses under the empty Nessler 

 tube, using a piece of white paper to reflect the light. Hold 

 the tubes with one hand sufficiently high to reflect all the light 

 possible. Interchange the tubes several times to avoid inequal- 

 ities of light. The tubes should be held in such a position that 

 both may be seen with one eye. Obviously, the final reading 

 may be made either by using all the glasses under the empty 

 Nessler tube or with some under the water as well. In this 

 way intermediate values sometimes not otherwise obtainable 

 may be found. 



In all cases the nitric acid used should be tested beforehand 

 for iron, this being a not infrequent impurity. 



Turbidity is determined with the electric turbidimeter de- 

 scribed in Leighton's paper. By means of an electric flash light, 

 a cross of light is provided at the bottom of a long graduated 

 tube. The well-shaken, turbid water is poured in until the sharp 

 image has disappeared and the hazy cross of light just dis- 

 appears. This is taken as the end point in the lower part of 

 the tube. In the upper part of the tube (that is, for slightly 

 turbid liquids) there is no hazy cross of light, and the end point 



