AVAILABILITY OF POTASH IN GROUND ROCK. 23 



1905-6. Three large beds in the greenhouse, all having the same 

 dimensions (21^% feet by 6f feet by 5 inches), Avere filled with a 

 common unfertilized meadow soil taken from the surface after the 

 removal of sod and grass roots. Four pounds of ground bone phos- 

 phate were bi-oadcasted and Avell worked in. 



Bed No. 1 now received 2 pounds of potassium carbonate contain- 

 ing about 67 per cent of potash (K.O). In order that the potassium 

 carbonate might be evenly distributed, it was dissolved in 2 gallons 

 of Avater and sprinkled evenly over the bed. After drying out, the 

 earth was again well worked. 



Bed No. 2 received 17.4 pounds of finely ground potter's feldspar 

 containing 8.3 per cent of potash (KX)). This was also broadcasted 

 and well worked in. It will be seen that beds 1 and 2 contained 

 nearly equivalent amounts of potash, the only diiference being that 

 in one case the soluble carbonate was used and in the other grouuvl 

 rock. 



No potash in any form was added to bed No. 3, which was used for 

 comparison. The soil itself, however, contained about 0.5 per cent 

 of total potash, and as the filtered Potomac Iviver water which it was 

 necessary to use for watering throughout the entire experiment con- 

 tains an amount of potash equal to about 2 parts per million, the 

 tobacco on bed No. 3 did not at any time show the effect of potash 

 starvation. 



All three beds were supplied with a sufficient quantity of nitrogen 

 in the form of pure ammonium nitrate. The seed chosen for the 

 experiments was carefully selected Cuban wrapper tobacco. In 

 making experiments of this kind it is necessary to select seedlings 

 as nearly as possible of equal vigor. It is not improbable that many 

 wrong deductions have been drawn from pot or plat experiments in 

 which the need for careful selection of seed has been overlooked. 



Tobacco seedlings were set out in four rows of ten plants each in 

 the three beds. The tobacco crop is usually planted with about 8,000 

 plants to the acre, so that the experimental plats represented, as far 

 as plant growth was concerned, one two-hundredth of an acre. The 

 actual amount of potash added to plats Nos. 1 and 2 was equivalent 

 to 280 pounds to the acre. As the original soil contained a sufficient 

 amount of potash to mature some sort of a crop on plat No. 3, it is 

 not jDossible to state the actual amount of potash present in the three 

 plats. The potash in the soil was not water soluble and would not 

 have been termed available in the present sense of the word. 



The tobacco grew well throughout the experiment, and for the first 

 six weeks very little difference was noted. After this time, however, 

 it became evident that the plants in plat No. 3 were falling behind. 

 At the end of twelve weeks the ^jlants in plat No. 3 were plainly 



104 



