POSSIBLE CAUSES OF THE GRASS EFFECT 279 



number in the earth without grass, where the trees are flourish- 

 ing, as in the sand with grass, where the trees are suffering 

 (XIII, 123). 



Alkalinity may arise in soil through the growth of grass in 

 it, for grass in growing decomposes some of the minerals present, 

 absorbing the acid constituents of them more readily than the 

 bases. The resulting alkalinity might conceivably be an expla- 

 nation of the deleterious effect of grass on trees. To examine 

 this, trees were grown for two years in pots of sand with the 

 necessary nutrients added : in some cases the sand was made 

 alkaline by the addition of sodium bicarbonate, whilst in others, 

 either no alkali was added, or acid was supplied to them. In a 

 further set of pots the trees were grown under a surface crop of 

 grass. The mean vigour of the trees for the two years may be 

 expressed thus : 



No alkali, acid or grass . . . 100 

 Alkaline ...... 102 



Strongly alkaline . . . . 102 



Acid . . ... -97 



Grass ...... 90 



The action of the grass in these experiments was considerably 

 less than that usually found, due to circumstances which need 

 not be specified here, but it was, at any rate, appreciable, and 

 was much greater in the second year (vigour of the trees, 68) 

 than in the first ; whereas no similar action resulted from making 

 the sand either alkaline or acid, the differences of 2 or 3 per 

 cent, being well within the limits of experimental error (XIII, 96). 



The presence of alkali in soil might, however, have some 

 action, whereas none might occur where the medium was sand. 

 To examine this, a number of apple trees growing in the open 

 were watered once a week during the summer throughout three 

 years with various amounts of potassium carbonate solution, 

 the quantity supplied, where the full dose was given, being 

 equivalent to 810 grams of potash (K 2 O) for twelve trees, repre- 

 senting O'O4 per cent, of the weight of soil in the upper nine 

 inches. The mean results showed that such a dose was de- 

 leterious to the trees, as also was a half dose, though to a lesser 

 extent, but that the harm done, even by the stronger doses of 

 alkali, amounted to much less than that done by grass, where 

 the trees, at the end of the three years, were actually moribund, 

 showing during the third season a deficiency in growth of as 



