6 BULLETIN 681, U. S. DEPARTMENT OF AGRICULTURE. 
might be expected that the velocity of the flame through a cloud of very fine dust 
would more nearly approach the velocity attained in gas explosions. As a matter of 
fact, this is true. Taffanel has brought this out in tests made at the French Experi- 
ment Station. Results obtained by the Bureau of Mines lead to the same conclusion. 
The rate at which inflammation travels through a gas mixture is dependent upon 
at least two factors, the inflammability of the .gas and the percentage of the gases in 
the mixture. The rate at which inflammation travels through a dust cloud seems to 
be dependent upon two similar factors, namely, the inflammability of the dust and 
the amount of dust in suspension. A third important factor is the fineness of the dust. 
It will be seen, therefore, that gas and dust explosions are similar in many ways, 
and that a gas explosion is only a limited case of a dust explosion. 
A thorough knowledge of the nature of dust explosions and of the accompanying 
phenomena is necessary in order to devise means for the prevention of such explosions 
and for the stoppage of explosions if by mischance they should start. It is, therefore, 
necessary to know how dust ignites, the various means by which it may be ignited 
or inflamed, and the chemical processes that take place. Another important factor 
to know is the ease with which the dust ignites and propagates a flame. This property 
we may call the "inflammability" of the dust. The term ' ; explosibility " has also 
been applied in this connection. 
It is important to note that investigations of dust explosions 
indicate that, in general, two reports are heard, the first being sharp 
and quick, and the second of a loud, rumbling nature and accom- 
panied by more flame. It is thought that the first report is due to 
the ignition of a small quantity of fine dust in suspension, and that 
the resulting concussion is sufficient to disturb the dust that has 
settled at places farther away, thus forming an additional explosive 
mixture. The flame from the first explosion furnishes the source of 
ignition for this newly formed dust mixture, so that the big explosion 
with its loud roar is propagated throughout the entire dust zone. 
APPARATUS USED IN THE INVESTIGATION. 
For the purpose of conducting the experiments on grain-dust 
explosions a small, light frame building (15 by 15 by 24 feet), covered 
with galvanized iron, was erected by the Department of Mechanical 
Engineering, Pennsylvania State College (PL I). A small dust room 
was afterwards built at the side of the main structure at the dis- 
charge end of the screw conveyor. The grinding floor is elevated 
about 6 feet above the basement, which is level with the ground. 
A 16-inch attrition mill (PL II) was installed on a platform 14 
inches high, with a removable hopper underneath. Two elevators, 
one dump bin, a stock hopper over the mill, and a small bin for 
receiving ground materials were also provided. A screw conveyor 
10 feet long was placed 2 feet below the floor to receive the material 
from the mill, and to carry it to the chop elevator or out into the 
dust room. The attrition mill, screw conveyor, and elevator are 
driven by a 15-horsepower direct-current motor, located in the 
basement. 
The elevator legs and feed bin can be seen on the left of figure 1, 
Plate II. Immediately over the mill hopper are the spouts extend- 
