Nepigon and Saguenay Rivers. 139 



peaty matter held in suspension, but there 

 is no peat on some of the reddest streams 

 that I have fished, so acting upon the sug- 

 gestion of Secretary Small I made experi- 

 mental test-tube tannates of iron and of 

 manganese in weak aqueous solution. The 

 iron tannate was at first redder than stream 

 water and the manganese tannate was too 

 smoky. A combination of equal parts of 

 solution of iron tannate and of manganese 

 tannate gave at first too smoky a color, 

 but after standing exposed to the light 

 for a few hours an abundant flocculent 

 precipitate formed in all of the test tubes, 

 and all of the solutions assumed a tint 

 quite characteristic of that of the streams. 

 This stain was little changed at the end of 

 two weeks. Rain water coming in con- 

 tact with dead tannin-bearing trees and 

 plants would take tannin into solution, 

 and this solution percolating through a 

 soil containing iron or manganese would 

 make highly colored tannates of the met- 

 als before reaching a stream. Even in 

 peaty waters the color could be due in 

 part to the presence of tannates. It was 



