10 BULLETIN 343, U. S. DEPARTMENT OF AGRICULTURE. 
condensation being blown off as soon as the temperature fell below 
that corresponding to the steam pressure in use. In some cases, 
after cooking the required length of time, the charge was allowed to 
stand until it had absorbed a large amount of water. After unload- 
ing, the wood was weighed and then ground as soon as possible. In 
several instances samples were taken after cooking and the bone-dry 
weight and the moisture were determined. 
GRINDING. 
Before starting the test the desired surface was placed on the stone 
by means of a mechanically controlled burr or bush roll. The stone 
was then washed until free from loose sand and an impression of the 
stone was taken by means of carbon and coated paper. (See Forest 
Service Bulletin 127.) The grinder pockets were filled, the pressure 
adjusted to the correct value, the recording instruments were placed 
in operation, and the test was then started. 
The speed and pressure were maintaimed as constant as possible 
throughout each test. Pocket binding was eliminated by the con- 
stant observation of power applied to the grinder, a falling off in power 
consumption with pressure on indicating binding. When a piston 
was raised, instead of allowing the speed to increase, it was held con- 
stant by regulation of the voltage on the motor armature. 
In series tests in which the surface of the stone was not altered, 
but the speed, pressure, or temperature were, the data secured may be 
more or less open to criticism, since the surface undoubtedly changed 
slightly from one test to another. However, this change was very 
slight.1. Likewise, in short tests it was impossible to heat up the 
stone thoroughly and here also a very slight error was introduced.! 
During the tests readings were taken of power, speed, pressure, 
temperature, and the like, the intervals of reading being 15 minutes 
in most cases, although 5-minute intervals were sometimes employed. 
LOSSES IN CONVERSION. 
While it is highly desirable that the losses occurring in the barking 
of different woods be determined, it was impossible to investigate 
this point satisfactorily. Only small amounts of wood were used, in 
some cases not more than 8 cords, and determinations on such small 
amounts would not yield reliable results. 
The amount of wood used during any test and the amount of wet 
pulp manufactured was accurately determined by weighing. Moist- 
ure samples were taken of both the pulp and the screenings and the 
bone-dry weight of each determined. In some cases the amount of 
wood fiber in the white water was secured by measuring the amount 
1 This point is discussed in more detail in Forest Service Bulletin 127, ‘‘The Grinding of Spruce for 
Mechanical Pulp.’’ 
