Manganese Sulphate upon the Growth of Barley . 577 
From this table and the resulting curve (Curve 3) it is evident that the 
possible stimulus with 1 :5,000,000 C.S. does fall within the range of 
experimental error. This was confirmed by recalculating the means of 
Curve 3. The black line shows the mean of the dry weights of twenty series of barley plants 
grown with copper sulphate and nutrient salts. The dotted lines indicate the limits of probable 
experimental error for each concentration of the toxic salt. (See Table I for concentrations.) 
i : 5,000,000 C.S., i : 10,000,000 C.S. and the control results after rejection 
of the individual results below 60. This corrected mean fell within the range 
of experimental error in the first two cases, and above it in the control, and 
indicated that no stimulus occurred with 1 : 5,000,000 C S. 
