RESISTANCE OF HARD WOODS TO CREOSOTE. 11 
creosote. In oak, hickory, and sycamore, for example, the rays were 
conspicuous because of their resistance to penetration in marked con- 
trast to the surrounding treated tissue. 
SPRINGWOOD AND SUMMERWOOD. 
In the hardwoods the effect of springwood or summerwood on 
treatment was not so important as the effect of the peculiarities of 
the different types of structure and the arrangement of the elements, 
such as vessels, fibers, rays, etc. In very heavy treatments the creo- 
sote not only passed through the cavities or lumena of the cells in both 
spring and summer wood, but sometimes penetrated the walls. 
CONDITION OF GRAIN. 
More or less difficulty was experienced in treating the woods in 
Avhich the fibers were interlaced or cross-grained. 
DENSITY. 
The ease or difficulty of securing a satisfactoiy penet ration does not 
appear to depend upon the density. Woods having high specific 
gravities were sometimes treated with greater ease than species of 
much lower specific gravity, and vice versa. 
RADIAL, TANGENTIAL, AND LONGITUDINAL PENETRATION, 
In most of the species tested the radial and tangential penetrations 
were very much less than the longitudinal penetration. In general, 
the species that were difficult to treat showed less difference than those 
which were easy to treat. 
GROUPING OF SPECIES. 
GROUPING WITH RESPECT TO. PENETRATIONS AND ABSORPTIONS. 
Table 2 (see Appendix) gives the average longitudinal and radial 
penetrations obtained in the penetrance tests, and the average ab- 
sorptions of the specimens treated in the cylinder. In this table 
and in the diagram, figure 3, the species are arranged in order of 
increasing absorptions in the impregnation test. In figure 3 are also 
shown the corresponding average longitudinal and radial penetra- 
tions. While the longitudinal penetrations show a general tendency 
to increase as the absorptions increase, there is nevertheless a con- 
siderable variation. The radial penetrations were very small in 
most of the woods treated and do not seem to bear an important 
relation to the absorptions. In some cases it is possible that the 
longitudinal penetrations would have shown a closer relation to the 
absorptions if the average had been based on the same number of 
specimens in both series of tests. The penetrations represent the 
