EXPERIMENTS WITH PAPER FILTERS. 
439 
increased to two; these overcoats were pierced with long, narrow apertures, running 
from the point to the circumference. 
In the following experiments, the improved filter was tried against the plaited form. 
The improved filter, on account of its great strength, was allowed a funnel with a throat 
once and a half as broad as that given to the plaited filter. It would have borne one 
three times as broad. 
1st trial, 121:100 3rd trial, 112:100 
2nd „ 125 : 100 4th „ 115 : 100 
The fifth trial was made with a precipitate of sulphate of calcium ; the filtrates were 
82 : 100. Here the new filter seemed at fault, but on weighing the precipitates col¬ 
lected, they were as 140 : 100. The new filter being last filled, got the thicker portions 
each time, while the plaited filter got the weak top liquor. 
The outer jacket is cut by folding the filter as a plain filter and taking out a sliver on 
each edge; repeating the process increases the number of apertures. If the filters are to 
he used double, as in broad-throated funnels, the openings should extend nearly to the 
vertex; if they are used single in common funnels the openings need not extend so far, 
and the extreme point may be removed. 
A thick porous felt might be used for acid liquors as an outer filter. Cotton cloth 
would serve in alkaline solutions; gun cotton could be used with either. If asbestos 
cloth could be procured, it would probably be the best. It might be cleansed by burn¬ 
ing, and would be unaffected by anything likely to be filtered in quantitative analysis, 
nor would it harm the filtrate when estimations are to be made by permanganate of 
potassium. 
Most coarse filter-paper is “ stuffed ” with mineral matter; such paper must, of course, 
be leached in acidulated water before being used for quantitative work, where the filtrate 
is to be saved. 
When the precipitate is dry, the outer filters are thrown away, only the fine inner 
filter being burned. 
Plaited and plain filters were tried in like funnels; calling the plaited filters No. 1, 
we had:— 
i. ii. 
i. ii. 
1st trial, 116 : 100 3rd trial, 236 : 100 
2nd „ 237: 100 4th „ 136 : 100 
5th „ with sulphate of calcium precipitate, 231 : 100 
The precipitates weighed 166 : 100. 
I then tried plain filters off the same sheet against each other in like funnels; usually 
the results varied but a few per cent., though sometimes much more; the greatest diffe¬ 
rence noticed was 2 to 1; several times the results corresponded exactly on repeated runs 
of 500 cubic centimetres. 
Experiments were made to determine the difference in efficiency between the single 
and the triple sides of filters. No. 1 had its triple side covered with paraffin, leaving 
the single side free. No. 2 had the triple side free, while the single was covered; with 
paraffin the result was 175 : 100; glycerin was then tried wfith the result 200 : 100, 
showing the additional paper considerably retarded the flow. 
I thought, since the adhesion of the water to the glass is the cause of slow filtration, 
I might increase the flow by coating the funnels on the inside with paraffin, to which 
water does not adhere. No. 1, being coated, No. 2, left clean, I got 
i. ii. i. ii. 
1st trial, 200: 100 3rd trial, 100 : 100 
2nd „ 184 : 100 4th „ 137 : 100 
The filters in the third and fourth trials were the same, but the funnels were changed 
about. 
The outside or skeleton filters, above described, may be cut the same size as the inner 
filter; if much smaller the upper part of the inner filter clings to the glass; if larger, a 
part of the precipitate is liable to adhere to the outer filter, and even with great care a 
part of the precipitate would creep up and be lost.— Abstracted from an article by 
Charles E. Avery , in the ‘ American Journal of Pharmacy .’ 
