62 BULLETIN 1036, U. S. DEPARTMENT OF AGRICULTURE. 
From a comparison of Table 21 with Table 22 it will be seen that 
the loss from piling is much less than the loss from poles, although in 
some of the piling there was a greater percentage of creosote distilling 
below 245° than there was in the poles, the loss from the butt in the 
poles being on an average somewhat greater than the loss from the 
air section in the piles, and considerably greater when it is considered 
on an average yearly basis. Table 23 shows the analyses of creosote 
extracted from woods that have been subjected to different lengths of 
service. Xo estimate of the loss on this material can be made, but 
attention is called to the fact that, in spite of the very heavy loss that 
was probably experienced (see Tables 21 and 22), the fractions dis- 
tilling below 245° have not disappeared entirely. These fractions 
are the ones that usually contain the tar acids and the tar bases, 
both of which are known to be very toxic. Not all the investigators 
made determinations for tar acids, but in 65 per cent of the analyses 
that were made the acids were found. The absence of tar acids does 
not necessarily indicate that these compounds have volatilized, 
because it is possible for them to undergo chemical changes, as pointed 
out by S. Cabot (24), or for them to attack the cellulose or wood 
substance in such a way as to produce materials resembling bakelite, 
or for them to be dissolved in the water contained in the wood, and a 
part of them at least to penetrate into the apparently untreated por- 
tion of the wood. In any one of these conditions the tar acids would 
probably be missed by the investigator. 
Xo short test is known to the writer that will give the volatility 
of creosote in any but a comparative way. For this purpose an 
examination of the percentage of distillate would probably serve 
as well as any other test. A number of tests have, however, been 
made both by evaporation from open dishes or pans and by evapora- 
tion from the treated wood. Figure 31a shows the loss of creosote by 
evaporation from open pans or dishes under two conditions of heating, 
and the percentage distilling at 270° C. The points fall roughly in 
a straight line. Figure 31b shows the same kind of relation between 
the amount distilling at 270° C. and the loss from creosoted wood. 
The relation here is represented by a nearly straight line, but the 
slope is much steeper than before. Furthermore, the maximum loss is 
less than half that in the dish test, although the oils were much 
lighter in character. 
Yon Schrenk and Kammerer (25) have reported a number of tests 
on sLx creosotes under three sets of conditions; that is, evaporation 
from open pans at room temperature for 304 days, evaporation for 
the same length of time from maple blocks, and evaporation from 
pine blocks. Using this length of time, the investigator shows that 
the loss by evaporation from the open pan is practically the same 
as the loss by evaporation from wood under the same conditions for 
