2 BULLETIN 372, U. S. DEPARTMENT OF AGRICULTURE. 
The first problem, therefore, seemed to be to increase the leaf 
area and thus increase the yield of oil per acre. Accordingly, in 1910 
a small plat of ground was set with plants grown from a mixed col- 
lection of seeds secured from wild plants in Volusia County. Of the 
first year's growth from this plat, 196 pounds of herb was distilled and 
yielded 0.17 per cent of oil. In the second year the yield of oil from 
plants grown on the same plat was 0.24 per cent. The content of 
total phenols in the oils was found to be 64 per cent. 
A large number of variations in the plants were noted in 1910, and 
from these about sixty individual plants showing the various forms 
were selected for seed purposes. The seed from these plants was sown 
in 1911, but the resulting plants were not true to type. It was noted, 
however, that the variations fell into two general classes — dark and 
light plants — and that to a large extent these variations came true 
to type. The dark plants had much darker leaves, more serrate, 
and with a pronounced red color in the stems and branches. The 
light plants had larger leaves, lighter green in color, and with little 
or no red color in the stems and branches. There was also a greater 
tendency to the production of leaves in the light than in the dark 
type. From 34 tests which were made to show the relative yield of 
oil of these two types of plants the following average yields of oil 
were obtained: From the dark type, 0.40 per cent of oil, which 
contained 66 per cent of phenols; from the light type, 0.42 per cent 
of oil, with 69 per cent of phenols. The specific gravity, as well as 
the percentages of oil and of phenols in the oil, was found to be higher 
in the plants of the light type. Future seed selection was therefore 
carried on with this type. 
In 1912 a further comparison was made between the light and dark 
forms and the results obtained were still more marked, but owing 
to weather conditions all samples showed a low yield of oil. The aver- 
age of 36 tests was as follows: Dark type, 0.31 per cent of oil, which 
contained 65 per cent of phenols; light type, 0.40 per cent of oil, 
which contained 68 per cent of phenols. By continuing the selection 
of the light type of plants the yield of oil has been gradually increased, 
and in 1914 the herb from a plat of two-thirds of an acre yielded 0.44 
per cent of oil, which contained 72 per cent of phenols, while in 1915 
the average was 0.42 per cent of oil, containing 74 per cent of phenols. 
During these years selection was also made to increase the size of 
the plants in order that the tonnage of herb per acre might be in- 
creased. This was also successful and a considerably increased yield 
was noted year by year. It is believed that at the present time the 
selection has been carried far enough to warrant the use of the im- 
proved form for the commercial production of thymol in the United 
States. 1 
The department has no supply of selected seed or plants of horsemint available for distribution. 
