320 ANNUAL REPORT 
The ball-mill has been discussed sufficiently under the topic of 
raw grinding and was quoted as having a capacity of 12 to 16 barrels 
per hour through the 20 mesh sieve, requiring 30 to 40 horsepower and 18 
to 24 barrels with 40 to 50 horsepower. It is not an economical ma- 
chine as an intermediate grinder. 
As to the comparison between the Griffin and the tube mill, it 
must be stated that the Griffin mill is the more economical of the two 
in spite of its heavy repairs. Its tendency to separate particles of differ- 
ent specific gravity cannot be made a point of criticism here inasmuch 
as the material is homogeneous in character. The capacity of a Griffin 
mill is from 8 to 10 barrels per hour, requiring 25 to 30 horsepower. 
The capacity of the tube mill, 22 feet by 5 feet, is 14 to 20 barrels 
per hour, requiring 70 to 75 horsepower to operate and 125 horsepower 
momentarily in starting. 
Fineness of Grinding.—It has been claimed repeatedly that the 
product of the tube mill is far higher in finest material than that 
of any other machine, and hence cement ground by this machine is 
supposed to be higher in quality owing to the large amount of cement 
dust present. The writer investigated this point by collecting himself 
samples of cement ground by several types of machines and having me- 
chanical analyses made, using for separating the fine particles alcohol re- 
distilled over caustic lime. The results of these analyses are shown 
in the following table on page 321. 
These results, of course, cannot be used to condemn any machine, 
as other factors enter into the consideration of grinding, so that the very 
attempt to draw general conclusions in regard to the quality of the 
grinding would be absurd. However, in comparing the figures of column 
TO it is seen that they are quite in favor of the Griffin mill and show 
that in practice this type of machine is more than the equal of the tube 
mill. It is also observed that the finer cements show a higher tensile 
streneth with 1:2 briquettes, just as is to be expected, for the sand test is 
to some extent a test for fineness of grinding. There is perhaps a tend- 
ency in practice to overcrowd the tube mills or to feed them irregularly. 
This may account for the poorer showing of the tube mill. 
These experiments, in addition, show that the usual sieve analysis 
is not a satisfactory and accurate means of measuring the real fine- 
ness of cement, since the 100 mesh sieve commonly employed tells 
practically nothing about the amount of cement dust present. 
In the method proposing to) use only an intermediate grinding 
machine, together with an air or centrifugal separator, it would prob- 
ably be best to use the Kent mill or rolls. The latter have proven quite 
efficient at the works of the Edison Company. 
