2 BULLETIN 129, U. S. DEPAKTMENT OF AGEICTJLTTJRE. 
three trees was distilled. Further, the average yields from the heart- 
wood 1 of several trees were in a few instances compared with the 
yields from lumber of the same species. Differences in yields may 
also occur in trees of the same species grown in different localities, 
and for this reason the results obtained are averaged separately when 
more than one locality is represented. At least two distillations 
were made of each kind or form of material tested. 
Different forms of wood — such as body wood (wood free horn, bark), 
slabs, limbs, etc. — were distilled, but the proportion of each used in 
commercial practice varies with different plants and localities, so that 
it is not possible to assume a proportion representing average con- 
ditions. For this reason the yields from different forms of wood of 
the same species are presented separately. The corresponding yield 
per given weight of wood, made up of any proportion of the various 
forms, can readily be calculated. However, as a basis of comparison 
between the species, the average yields from all heartwood runs (in- 
cluding lumber) are taken arbitrarily as the species value. The mean 
of the heartwood and slab-wood yields is also given wherever both 
forms were distilled. 
THE RETORT. 
Figure 1 shows the construction of the retort in which the distilla- 
tions were made. The retort proper A was surrounded by the oil 
jacket B, which was filled with a high-flash-test cylinder oil. The 
outlet pipe connected the retort with an ordinary worm condenser 
(not shown). The pyrometer tubes, 1, 2, 3, 4, and 5, made it possible 
to measure the temperature at various places within the retort. The 
retort was mounted on an iron stand, was insulated on all sides, and 
heated by a row of gas burners underneath. The flames from the 
burners played chiefly up one side of the cylinder and induced a fairly 
good circulation of the hot oil around the retort. 
PREPARATION OF MATERIAL. 
The forms of material used varied to some extent with different 
species, but most of them consisted of round bolts. These were sawed 
into slabs and heartwood in about the same proportion as would occur 
in ordinary sawmill practice, and the percentage of bark on the slabs 
was roughly determined. Sticks were prepared from 6 to 8 square 
inches in cross section and a trifle less than 18 inches long. Just before 
each charge of wood was weighed, six 1-inch sections, each cut from 
a different stick and in each case from a different part of the stick, 
were taken for moisture determinations. 
In the comparative distillations on bark and sapwood the material 
was taken from the same bolts. When limbs were used they weie 
1 The term "heartwood " as used in this paper applies to the material left after the slabs have been removed 
from a bolt or log. It was in all cases entirely free from bark, but small amounts of sapwood sometimes 
remained. Lumber is considered as made from heartwood as thus denned. 
