368 BOTANICAL GAZETTE [November 



the flora present in them and the almost complete inhibition from 

 them of plants other than bog xerophytes. Among the various 

 conditions which are factors in this, the toxic properties of the 

 water have been shown to be very important. He has also (13) 

 published evidence tending to show that the toxic properties of 

 this water are due, at least in part, to the presence of matter in a 

 colloidal state. The present paper gives fuller details of the 

 experimental work pointing to this conclusion. It also reports 

 further experimental evidence tending to support this conclusion 

 and furnishes some evidence as to the colloidal state in which 

 the toxic substances are present. 



Experimental work 



The samples of bog water used were collected in most cases 

 by digging a hole, usually not more than 30 inches deep, in the 

 substratum, and dipping from this the water that accumulated 

 within a few minutes. When the bogs were too dry to admit of 

 securing water in this way it was squeezed from handfuls of the 

 decaying material. In all cases it was collected in glass con- 

 tainers and taken to the laboratory, where it was strained through 

 washed cheesecloth and then filtered twice through filter paper. 



Samples collected from several bogs and filtered in this way 

 were treated with various electrolytes [NaCl, MgS04, (NH,)S04, 

 and Na2HP04] to determine whether there was any material in 

 the water that could be precipitated by this means. 



To a 250 cc. sample from North Mud Lake bog a like volume 

 of saturated solution of (NH2)S04 was added. After shaking 

 thoroughly this was allowed to stand. At the end of 2 days no 

 precipitate had appeared, but at the end of 5 days a precipitate 

 could readily be seen. 



When 90 cc. of water from Fauntleroy bog was saturated with 

 (NH4),S04by adding the salt gradually and shaking no precipitate 

 appeared at once. When this had stood over night, however, 

 there was a considerable quantity of precipitate, consisting of 

 brown, somewhat flaky particles, some at the surface, some at the 

 bottom, and some remaining suspended in the liquid. This pre- 

 cipitation by complete saturation with (NH4)2S04 was repeated 



