112 EXPERIMENTS WITH PLANTS 



the water by means of a piece of oilcloth about three 

 feet square placed beneath the box. How rapidly does 

 the water run off? How much remains behind? Place 

 a layer of moist cotton batting, about an inch thick 

 (sawdust or sphagnum moss such as florists use may 

 bvi employed) on the surface of the clay, and repeat the 

 experiment. Remove the clay from the sand and re- 

 peat the experiment, with and without the cotton. In 

 this experiment the cotton represents a covering of 

 vegetation on the soil: such a covering retains the 

 \v:iter in precisely the same way as the cotton. Com- 

 ] ire as well as you can by observation the amounts of 

 rani which run off from the surfaces of the following: 



Clay soil without covering. 



Sandy soil without covering. 



Soil with a covering of turf. 



Soil with a covering of tall weeds or other plants. 



Soil with a covering of shrubs or trees. 



It is believed that the floods of our great rivers, 

 s:ich as the Mississippi, could be wholly averted by 

 preserving the forests at the headwaters. Find out the 

 ,'innual damage of these floods. This represents the 

 .interest on the sum which could be profitably invested 

 in preserving these forests. What does this sum amount 

 4 ~> ? If you live in a region where damage by flood 

 <> mrs, make a similar estimate for your own region. 

 When it is not feasible to reforest these areas, the 

 natural growth of shrubs and undergrowth should be 



