COAL-TAR AND WATER-GAS TAR CREOSOTES. 55 
in dynes per square centimeter, T the absolute temperature in degrees 
centigrade and K and A constants for the oil. When the viscosity is 
determined at the two temperatures, T x and T 2 , then 
F t - J/ log V, -log y; , 
F.-r.^iogr.-iogr, A 
substitute the value of A in the first equation and solve for K. The 
equation thus obtained will give the values of V at any temperature, 
providing the original determinations were accurate. The equations 
6 54(10) 18 
for the two curves shown in figure 28 are F= — — ™ for the creo- 
sote and V= ' Tl2 6 3 — for the carbolineum. 
The viscosity of oil is supposed to have an effect on its penetrance 
into wood, and it seems reasonable to suppose that a limpid fluid 
would be easier to inject than a more viscous one. Weiss (16) stated 
that penetrance is some inverse function of the viscosity. Bond (17) 
showed that, in the different mixtures of creosote and carbon-free 
tars tested, there was no apparent relation between the viscosities 
of the creosotes and the penetration obtained. He also showed that 
the same thing was true when mixtures of normal tar and creosote 
were used. 
Later, Teesdale and MacLean (18) stated that there was no appar- 
ent relation between the viscosity and the penetrance of the tar 
mixtures used. Their statements are, however, based on the vis- 
cosity as determined by the Engler viscosimeter. Since then their 
data have been recalculated to absolute viscosity and show that a 
very definite relation exists between absolute viscosity and pene- 
trance, and that this relationship is capable of mathematical treat- 
ment. However, the data are not as yet sufficiently extensive to 
show the relationship of other variables which enter into the pene- 
trance. 
Figures 29 and 30 show the relationship between longitudinal pene- 
tration and absolute viscosity of various oils, including creosote oils, 
tar mixtures, tars, and asphaltic oils into noble fir and long-leaf pine. 
The equations given in the figures are of value only when the other 
conditions used in the test are held constant. Other factors which 
may influence the penetration are the time of treatment, the pressure 
used, and the mositure content in the wood. In all probability a 
full equation should read M, P, T, XY ' = K 2 , where M is some 
function of the moisture content, P some function of the pressure 
used in treatment, T some function of the time of pressure, Y the 
absolute viscosity, X the longitudinal penetration, and K 2 a con- 
stant. In these experiments the moisture content, pressure of treat- 
