103 



ployed, were exceedingly bright and clear from the first until running 

 had quite ceased altogether. Another disappointment, however, awaited 

 inquiry into the actual improvement as to purity secured. The expo- 

 nent, on the average, was raised not materially to exceed one per cent, 

 of total solids attributable to the coal, exclusive even of sweet- waters. 

 A few analyses, taken at random from the laboratory records, sufficiently 

 illustrate this. In every case the non-filtered and filtered samples rep- 

 resent, as nearly as practicable, the same juice, tfor the large presses 

 these were taken in equal volumes at the discharge openings of defe- 

 cators and presses, respectively, at intervals of three minutes, always so 

 as to represent by pairs identical defecators of juice and identical defe- 

 cations, before and after filtration, which, following adequate admixture 

 of each series, as obtained from individual defecators, were re-sampled. 

 This was permitted by the admirable arrangement of the coal-mixing 

 receivers,which contained, each, precisely the amount from one defecator, 

 and which were filled and emptied alternately in rotation. The effect 

 of a thorough cake washing, the sweet- water being mixed back propor- 

 tionately with the filtered juice, of which it was the after-product, is 

 shown in the last two analyses. 



After that, due to the use of 10 or 15 per cent of lignite on the weight 

 of sugar present, no commensurate eft'ect was observed to be produced 

 in the direction of increased purity by the addition of further quanti- 

 ties. This loll off very slightly or not at all, however, as filtration pro- 

 ceeded towards its finishing point, as also more or less in lixiviation, 

 depending, as seemed shown, upon a lower or higher percentage of coal 

 employed. Believing the application of the process to Louisiana juice, 

 condemned by the excessive quantities of lignite found essential to suf- 

 ficiently rapid filtration and by its failure to realize a higher gain in 

 purity, before reaching conclusive knowledge of these minutiae it 



