30 
Journal of Agricultural Research v©i. xxm. n* » 
meal employed being 500 pounds. It was quickly seen that the cotton¬ 
seed meal showed a decided tendency to reduce or to delay the symptoms 
of sand drown while ammonium sulphate accentuated the trouble; 
There was no notable difference between the action of nitrate of soda 
and that of dried blood. Once more the injurious action of sulphates is 
in evidence, but, in addition, it is clear that only 500 pounds per acre of 
organic matter in the form of cottonseed meal exercises some preventive 
action against the disease. Apparently there is some similarity in the 
action of certain chlorids and that of cottonseed meal calling for explana¬ 
tion, Different rates of application of nitrogen in the form of dried 
blood produced no notable effect on the disease, so that neither the form 
nor the quantity of nitrogen in the fertilizer can be considered as being 
of primary importance. 
Another series of plot tests was conducted for several years at Oxford 
dealing with the fertilizing value of tobacco stalks and stems (leaf mid¬ 
rib) and barn manure for tobacco. For purposes of comparison the 
normal mixture of sulphate of potash, dried blood, and acid phosphate 
also was included in this series. The tobacco stems were applied at 
rates ranging from 1,000 to 2,600 pounds per acre, supplemented with 
sufficient quantities of dried blood and acid phosphate to furnish the 
normal application of 32 pounds nitrogen and 64 pounds phosphoric acid 
per acre. The tobacco stalks were applied at the rate of 2,70© pounds 
per acre, also supplemented with acid phosphate. The manure was used 
alone, in combination with sufficient acid phosphate and sulphate of 
potash to furnish the normal formula* and in combination with hardwood 
ashes as a source of potash. The rate of application of the manure was 
8,000 pounds per acre. As regards sand drown, the outstanding feature 
of these tests was that there were no symptoms of the disease except on 
the plots receiving only the mixture of dried blood, acid phosphate, and 
sulphate of potash. Here again is distinct evidence that vegetable forms 
of organic matter tend to prevent the development of sand drown. 
BIOCHEMICAL STUDIES OF SAND DROWN 
Because of the marked difference between the chlorid and the sulphate 
of potassium in their action on the tobacco plants, especially with regard 
to sand drown, experiments were undertaken for the purpose of compar¬ 
ing the effects of the two salts on the internal processes of the plant. A 
series of observations was made on the relative osmotic concentration of 
the expressed, filtered saps from the frozen tisisues of the leaf after the 
midrib had been removed, the material for study being taken from the 
two sets of fertilizer test plots at Oxford previously referred to. The 
samples were taken at intervals of two to four days, beginning shortly 
after the plants had been topped, that is, when the plants hadreached 
the normal stage for flowering. A single leaf near the middle of the 
plant was collected from a representative plant of each plot, and the 
same plants were used as sources of material throughout the tests. The 
leaves were collected in each case about 9 o'clock in the morning. In all 
cases the tobacco plants on the control plots receiving no potash showed 
distinct symptoms of severe potash hunger. The data obtained are 
summarized in Table I. The osmotic concentrations, in atmospheres, 
were calculated in the usual way from the observed lowering of the 
freezing point. 
