COAL-TAR AND WATER-GAS TAR CREOSOTES. 53 
gravities of oils, and at best the coefficient of expansion can be only 
a rough average of the various coefficients of expansion at different 
temperatures for different oils. In practice, it is usual to say that 
the change in volume is 1 per cent of the original volume for every 
22.5° F. of temperature change. This factor is approximately 
accurate and is based on the fact that creosote oil changes its specific 
gravity roughtly 0.0008 per degree centigrade, or 0.00044 per degree 
Fahrenheit. For commercial work over a short range of tempera- 
ture, this factor is probably sufficiently accurate; but for refined 
work, such as the experimental determination of absorption of creosote 
by wood, the figure of the next decimal place should be known. 
Some work at the Forest Products Laboratory shows that the figure 
derived from the change of gravity per degree centigrade lies between 
0.00077 and 0.00078, and between 0.000428 and 0.000433, with an 
average of 0.000430 per degree Fahrenheit. This work also shows 
that the actual change in gravity is independent of the initial gravity, 
provided the initial gravity at 10° C. is between 1.01 and 1.05; 
therefore a change in volume is dependent upon the initial gravity of 
the oil; that is, an oil with a specific gravity of 1.01, when heated 
through 100° F., will not have the same increase in volume as will 
an oil of 1.05 specific gravity heated through the same temperature 
difference. The volume, however, may be calculated by the use 
of the formula 
V > <*v 
G-{T-T) 0.00077 
where T and T' are temperatures in degrees centigrade; or by the 
use of the formula 
GV 
V 
'G-(T-T') 0.00043 
where T and T' are the temperatures in degrees Fahrenheit. In 
these formulae, V is the volume at the temperature T', V is the 
volume at temperature T, and G is the specific gravity at temper- 
ature T. 
VISCOSITY. 
Viscosity is a measure of the inner friction of liquids, that is, the 
friction produced by the liquid moving on itself. There are no 
instruments in commercial practice that measure this property 
directly. Most commercial viscosimeters are so constructed that the 
rate of flow of the liquid through an orifice of definite diameter under 
a definite head may be accurately measured. This does not give 
true viscosity, but does give an empirical measure of that property. 
Nearly all instruments are standardized by water at a fixed tempera- 
ture. The efflux time in seconds of the liquid under examination 
divided by the efflux time of water at the standard temperature, is 
