TOXIC EXUDATES AND SOIL TOXINS. 147 



colouring of the fruit. Such results were obtained at Ridgmont when the 

 ground was grassed up to 5 or 6 feet from the stem of the tree. 



"To what distance grass should be removed from a tree so as to have no 

 effect on it, must, naturally, depend on the nature and size of the tree, as 

 well as on the nature of the soil; with freshly planted standard apple trees, 

 in soil which was not specially favourable to the action of the grass, a very 

 considerable effect was produced when the grass was 4 feet away from the 

 stems: on the other hand, keeping a space free of grass extending only 6 

 inches from the stems of freshly planted dwarf trees was found to have some 

 beneficial effect, even in the Ridgmont soil. The proportion of roots extend- 

 ing into the grassed ground which are sufficient to make the grass-effect 

 apparent, is remarkably small, amounting in some cases examined to only 

 ^tf'ijffth of the weight of the whole tree. 



"Forest trees appear to be affected by grass in the same way as fruit trees 

 when the grass is sown immediately after planting; six different kinds were 

 examined, both at Ridgmont and in some light sandy soil. The only differ- 

 ence in the behavior of them and of the fruit trees was, that, in the case of 

 conifers planted in Hght soil, the effect was much less than with other trees, 

 and some recovery occurred with them as time went on, instead of the effect 

 becoming intensified. 



"The action of eighteen different grasses on apple trees was examined with 

 the general result that the action in all cases was considerable, but was 

 greater with the strong-growing grasses than with the weaker ones. Clovers 

 had a similar stunting effect, but the hghtness in the colour of the leaves, 

 conspicuous with trees under grass, was absent when clover was grown. 



"The question of the action of grass being explicable by its affecting the 

 aeration of the soil, by its altering the amount of carbonic acid present, or 

 by its effect on the soil-temperature, was investigated some years ago, and 

 any explanation on such grounds was found to be inadequate. The question 

 of soil-moisture and of food-supply was also investigated, with a similar 

 result, and further evidence has much strengthened these conclusions. 



"As regards soil-moisture, there are general grounds for regarding a defi- 

 ciency of such moisture as affording no explanation of the effect of grass on 

 trees, for this effect is produced in wet seasons as well as in dry ones, and trees 

 which are affected show none of the usual signs of suffering from drought; in- 

 deed, when vegetation suffers from drought, it is the grass which shows the 

 effect much sooner than the deeper rooted trees. Determinations were made 

 of the water contents of grassed and tilled soil at Harpenden at intervals 

 throughout a year, and it was found that the grassed soil was shghtly the 

 wetter of the two from the beginning of January till the end of March, after 

 which it became the dryer, but the water contents never fell below the hmit 

 which has been found to be favourable for plant-growth, although in this 

 very soil, when grassed, the trees were showing all the symptoms of grass- 

 injury. In the plots at Ridgmont, where dwarf apple trees have suffered 

 so much from grass, various determinations have all shown that the grassed 

 soil during the summer is actually wetter than the neighboring tilled ground. 

 What the explanation of this anomalous state of things may be, is not known, 

 but it effectually disposes of the view that the grass-effect there is due to lack 

 of moisture. In some experiments the moisture in the soil has been increased 

 to various extents by supplying the trees every week with water through 

 pipes under their roots, and, though such trees were slightly benefited by 

 this treatment, they still continued to show the effect of the grass very strong- 

 ly, and were far less vigorous than similar trees in tilled soil, though this was 

 much dryer. Still more conclusive experiments were made by growing trees 

 i n pots, and keeping the water contents up to the same point, by watering 



