14 BULLETIN 1025, U. S. DEPARTMENT OF AGRICULTURE. 
present begins to affect the color of the juices, and even after such 
prolonged use the earth may give considerable further service in ~ 
treating heavily colored juices. 
DIFFICULTIES ENCOUNTERED IN FILTERING TREATED JUICES. 
In the earier work various treatments were employed in the attempt 
to devise satisfactory methods of filtration after treatment with the 
earth which would be usable on a small scale. Some difficulty in ob- 
taining a perfectly transparent filtrate arises from the fact that dia- 
tomaceous earth contains a considerable quantity of excessively fine 
fragments of broken frustules which pass through ordinary filters and 
later partially settle out upon the walls of the containers. Lenher(/7) 
has shown that when silica or orthoclase is ground until most of the 
particles are less than 0.004 mm. in diameter, shaken up in water, and _ 
allowed to stand for several weeks the larger portion of the suspended 
material settles quickly, but the finer particles form strata in the 
liquid which persist for 10 days. At the end of 15 days the solution 
is stable, but a portion of the material is in colloidal solution and 
shows Brownian movement and Tyndall effect. It is clear from re- 
sults obtained in the laboratory that analogous conditions result 
when diatomaceous earth containing minute fragments of broken 
frustules is added to a fruit juice, as a progressive settling extending 
over some days or weeks occurs when a sample of treated juice which 
has been filtered through paper is allowed to stand undisturbed. 
As the centrifuge is successfully used in a wide variety of otherwise 
difficult separations and has repeatedly been suggested as a substitute 
for filtration in the clearing of fruit juices (JZ), its possibilities as a 
means for clarifying ciders treated with diatomaceous earth were 
thoroughly tested. After being thoroughly mixed with the earth, the 
juices were repeatedly passed through a gear-driven laboratory super- 
centrifuge having a maximum speed of 25,000 revolutions per minute. 
A single passage through the machine was practically as effective as 
three or more, since only a very small trace of material was recovered 
by any treatment after the first. While the appearance of treated juices 
was greatly improved, a considerable amount of colloidal material 
remained in suspension even after repeated stirring with earth fol- 
lowed by passage through the centrifuge. Juices so treated and sub- 
sequently pasteurized and stored remained for 18 months as hazy 
colloidal solutions. Attempts further to clarify such centrifuged 
juices by ordinary filtration were unsuccessful, as any filter having 
sufficiently close texture to hold back the suspended particles quickly 
became choked by slime. | 
For these reasons the complete removal of the added earth from 
the juices by any of the more generally available methods of filtra- 
tion appeared somewhat difficult at the outset. The best types of 
