14 BULLETIN 600, U. S. DEPARTMENT OF AGRICULTURE. 
The results given in these tables are calculated on the moisture- 
free basis except in cases noted. The total percentage of ash has not 
been given. It is not practicable to weigh ash on account of its 
hygroscopic nature, the particles of unburned carbon present, and 
the variable amount of carbon dioxide present. It is almost impos- 
sible to burn to a white ash plants that are high in silica without 
volatilizing some of the constituents. The silica probably forms a 
slag with the fluxing alkalies and coats particles of unburned carbon, 
protecting it from further action. 
As an approximation to the total ash the sum of the mineral con- 
stituents is given. These figures are comparable to other results 
given for ash, except that carbon dioxide is not included and that 
partial correction has been made for the soil grains adhering to the 
plant. Many published tables show the percentage of a certain con- 
stituent in the ash, and the results here published may be compared 
with such data by taking the sum of the mineral constituents to rep- 
resent the ash. For example, in Table I, sample No. 1, alfalfa, there 
is 1.46 per cent of potash in the thoroughly dried plant, and, taking 
the sum of the mineral constituents to represent the ash, there would 
be 20.56 per cent of potash in the ash. 
Manganese in plants varies more in amount than most of the other 
elements. It appears to concentrate in the leaves, as shown in the 
case of the pine and oak samples. The presence of more than 0.3 
per cent of MnO in the raspberry plant from Acworth, N. H., is 
remarkable, and the question arose whether all raspberry bushes and 
like plants were high in manganese. Two samples, one of raspberry 
bushes and one of rambler roses from Vienna, Va., were collected and 
analyzed. The Blackcap raspberry bushes showed 0.062 per cent of 
MnO, and the rambler roses, stem and leaves, contained 0.019 per 
cent. This. would indicate that members of the rose family contain. 
more manganese than the average of plants. White* has found from 
0.108 to 0.153 per cent of Mn in sorrel plants, and from 0.059 to 
0.085 per cent of Mn in clover plants. 
There is wide variation in the composition of the same plant. 
Some of this variation may be due to the different stages of growth, 
but the greater part of it would seem to be due to the composition 
of the soil. For instance, caesium was found in plant ash from 
Acworth, N. H., and Paris, Me. Caesium beryls are known to occur 
in both these localities. Shorey? reports an extreme variation of 
the composition of plant ash due to the difference in amounts of 
lime found in two soils. He writes: 
In the case of the fiber from Sisal Plantation, where the soil was a disinte- 
grated coral, 40 per cent of the ash was lime, while in that from the experi- 
1 Ann. Rept. Pa. State College Agr. Expt. Sta. (1913-14), p. 46. 
2 Bul. 13, Hawaii Agr. Expt. Sta. (1906). 
