36 



refused to run; if the juice was alkaline it would filter much better, 

 but gave highly colored products. 



Last year Dr. Wiley advised the use of sand. This gave good results 

 for a time, but gradually ran slow and failed to give satisfaction. The 

 size of the filters, in proportion to the juice worked, was very large, and 

 it soured easily. 



During the past winter experiments were constantly carried on with 

 the hope that something practical, cheap, and easily handled would be 

 discovered. Experiments were made with bone black, coal, sand, 

 gravel, oat straw, wheat straw, grasses, sedges, excelsior packing, and 

 many other things, all of which proved unable to do the work required, 

 were too costly, bulky, or in some other way not desirable. It was ac- 

 cidentally found that the coarse sawdust as it came from the mill 

 would do the work. 



Shallow filters are better than deep ones, and in well-conducted ex- 

 periments the juice was so well cleared of its mechanical impurities 

 that it appeared to be bleached. 



Examinations of the filters showed, among other things, soot from 

 the chimney, mud, and dirt. The juice was actually cleansed. The filter 

 used in this season's work was constructed as follows: A board twelve 

 inches wide was cut in four pieces and a box made 4 feet long by 2 feet 

 wide; a wire screen with one-sixteenth of an inch mesh was fastened 

 on the bottom, and three inches of sawdust placed within it. Care 

 should be taken that something should be placed over the sawdust to 

 break the fall of the juice and prevent guttering. 



It was found in practice that 1 bushel of sawdust was sufficient to 

 filter the juice from 15 tons of cane, and that the filter should be re- 

 newed every twelve hours. 



It may also be well to state that the hot juice as it came from the 

 evaporator was run through a sawdust filter, removing scum, scale, 

 dirt, etc. 



Double shredding. In 1885 samples were taken of the exhausted 

 chips as they came from the German diffusion battery and it was found 

 that better diffusion had taken place in small chips than from a larger 

 size; and last season this was found trife also of the battery which was 

 then being tried for the first time. All attempts to obtain a chip of 

 the size required failed, owing to the following facts : If the knives of 

 the shredder and the cutting bar were placed so closely together that 

 the small chips might be made either the shredder would not feed fast 

 enough or the knives would clog with the fine cane and stop cutting. 

 It was found this season that by making the ordinary cut first and 

 afterwards allowing the edge of the knives to project beyond the cyl- 

 inder very slightly, and by moving the cutting-bar closer and passing 

 the previously cut cane through a second time, the chips could be made 

 as fine as possible or as desirable. 



It was found in actual work that baskets of cane filled with chips of the 



