830 
The bolting cloth (No. 20), when new, 
contains from 5,100 to 5,600 meshes per 
square centimeter, varying with the maker 
and the sample; when thoroughly shrunken 
the number of meshes increases as much as 
30%. The total area of the openings in a 
square centimeter, on the other hand, de- 
creases over 50%, falling from .133 to .066 
sq. cm., and the average area of a single 
opening is reduced from .000024 to .00001. 
sq.cm. This latter figure represents an area 
of 82x32, the average area of the open- 
ings at the maximum efficiency of the silk. 
These dimensions do not, however, consti- 
tute a precise limit to the size of objects 
which can escape through the openings, for 
many meshes exceed this area, and some 
even double it. The struggles of the im- 
prisoned organisms and the pressure of the 
filtering water also materially assist the es- 
cape of the planktonts through the yielding 
meshes of the silk. 
For the past year and a-half the leakage 
of the plankton through the silk has been a 
subject of experiment at the Illinois Biolog- 
ical Station. The results were at first so 
surprising as to require the most careful 
corroboration, and they have been witheld 
from publication in the hope that an ade- 
quate remedy might be offered therewith. 
Water from a number of sources, collected 
at different seasons of the year, and con- 
taining plankton varying in amount and 
constituency, has been subject to examina- 
tion in several ways, and their relative 
efficiency determined by the Sedgwick-Raf- 
ter counting method. In order to reduce 
the error incident to this process, the enu- 
meration of the planktonts was extended 
over from five to ten times the customary 
fraction of the catch. The leakage has 
been tested as follows: by the Sedgwick- 
Rafter sand filter; by hard-pressed filter 
paper; by the centrifuge, and by the 
Berkefeld filter. The silk catches were 
made from measured quantities of water, 
SCIENCE. 
(N.S. Vout. VI. No. 153. 
thus eliminating the uncertainty as to the 
amount of water which the drawn net filters. 
Tests were made of the filtrate from the 
silk, and also of the unfiltered water. 
Owing to diversity in the constituency of 
the plankton, the ratio of efficiency of these 
various methods cannot be precisely stated. 
In a general way the order above given is 
that of increasing efficiency. The silk, in 
the pumping method of collection, retains from 
5% to less than 0.1% of the total number of 
organisms present (excluding bacteria), as 
contrasted with the catch of the Berkefeld 
filter. 
The Sedgwick-Rafter sand filter was used 
according to the published directions of 
Calkins,* and later of Jackson} and Whip- 
ple{. This method is far more efficient 
than the silk, but proved to be subject to 
considerable loss, especially in the case of 
water richly charged with plankton. Not 
only do the minute forms, as Raphidium and 
the smaller diatoms, readily slip through 
the sand, but also the more active species, 
as Euglena and Trachelomonas, escape in con- 
siderable numbers. Hxaminations of the 
filtrate from the sand revealed the fact that 
this method captures from 40% to 65% of 
the number of organisms present, the 
greater losses occurring with abundant 
planktonts. 
The filter paper employed was No. 575 
Schleicher & Schull. It is very free from lint 
and does not easily tear when wet. As the 
filtering proceeds the plankton is condensed 
in the bottom of the funnel by means of a 
fine spray from a hand bulb. When the 
required condensation is reached the plank- 
ton can be washed from the paper by the 
*G. N. Calkins. 
of Drinking Waters. 
Health, for 1891. 
{D. D. Jackson. Onan Improvement in the Sedg- 
wick-Rafter Method. Technology Quarterly, Vol. 
IX. 1896. 
$G. C. Whipple. Experience with the Sedgwick- 
Rafter Method. Ibid. 
The Microscopical Examination 
23d Ann. Rep. Mass. Board of 
