4 BULLETIN 606, U. S. DEPARTMENT OF AGRICULTURE. 
willow they are narrow as compared with those in sycamore, beech, 
and oak. 
Wood prosenchyma (Wood fibers, etc.) — Wood prosenchyma, (in- 
dicated by " X " in the plates, consists of thick-walled elongated cells 
in the form of fibers or of spindle-shaped cells with pointed, inter- 
locking ends. These specialized cells give the mechanical strength re- 
quired in the tree. In the softwoods the fibers, or tracheids, serve 
both for the conduction of liquids and for mechanical support. In 
the hardwoods, however, these two functions are separated. The 
pores or vessels act as a specialized conducting tissue, and the wood 
prosenchyma, though it shares in sap conduction to a small degree, 
acts chiefly as mechanical support. 
METHODS USED IN THE EXPERIMENTS. 
The presence of large vessels, the extent to which these are blocked 
by tyloses and gums, and the presence of other infiltrating substances 
in the cell walls are known* to be important factors in the impregna- 
tions of hardwoods with creosote. A very careful study was, there- 
fore, made of these factors. 
The resistance of the wood to treatment was studied in two kinds 
of tests : ( 1 ) Applying the creosote to a small area on a specimen and 
measuring the penetration in different directions, and (2) impreg- 
nation in a treating cylinder, blocks from each of the species being 
treated together in the same run. After treatment the specimens were 
examined under the microscope to determine the influence of the wood 
structure on penetration. 
APPARATUS. 
The penetrance apparatus in which the first series of tests was made 
is illustrated in figure 1. A hole three-fourths of an inch deep was 
bored in one face of the test specimen and the piece then clamped 
with the hole directly over the open end of pipe A, which led to the 
bottom of pressure tank B. The pipe and the lower portion of the 
tank were filled with creosote. Pressure was obtained by means of 
compressed air admitted through pipe C. A pressure regulator, K, 
was used to maintain a uniform pressure during the test. The ap- 
paratus was inclosed in a wooden oven having double glass windows 
in the front and back. Steam coils, J. heated the specimens and 
preservative to a constant temperature, which duplicated as nearly as 
possible the temperature conditions of a treating cylinder. Pres- 
sure and temperature were measured by gauge G and thermometer 
H. The specimens were placed on shelves in the apparatus and 
heated to a uniform temperature before testing. By the use of the 
