1900] MICBOSCOPICAL JOURNAL. 55 



water with the standards, the tubes must be shaken occa- 

 gionaly in order to keep all the material in suspension. 

 The stoppers used in the tubes should be boiled in water 

 before they are used, in order to extract any natural col- 

 oring matter which they may contain. Waters that have 

 both a high color and a high turbidity may be compared 

 with standards made by combining the silica standards 

 with the platinum-cobalt color standards. The color of 

 the water is first determined bj filtering the sample, and 

 the turbidity then observed by comparing the original 

 water with silica standards made to have the same or 

 nearly the same color. 



The various river waters recently examined have shown 

 turbidity-readings up to 3,000 of the silica standard, and 

 some of them have had also a high color. It was found 

 that the turbidity of these waters could be accurately 

 determined by dilution of the samples to bring them 

 within the range of the series above described. In most 

 cases this dilution with distilled water entirely eliminated 

 the effect of color. It is only in cases of especially high 

 color and where great accuracy is required that it will be 

 necessary to use the combined color and turbidity stand- 

 ards. In those cases where the suspended partiJiles them- 

 selves are colored, sufficient dilution will overcome the 

 diflSculty. 



The relative value of silica and kaolin for turbidity 

 standards is shown by the following comparisons: Three 

 samples of kaolin were obtained from different sources 

 and standards made from each by adding equal weights 

 of the material prepared by elutriation, to the same vol- 

 ume of water. 



Turbidity. 



Silicia loo 



First sample of kaolin 125 



Second sample of kaolin 135 



Third sample of kaolin 220 



The resulting turbidities were far from uniform, as is 



