No. 382.] METHODS IN PLANKTOLOGY. 741 
the liability of error cannot remove the defects inherent in the 
process itself. The defects noted by Whipple ('96) are involved 
in the method of concentrating the sample, v7z.—(1) the funnel 
error, arising from the adherence of organisms and amorphous 
débris to the sides of the funnel ; (2) the sand error, caused by 
organisms passing through the sand ; (3) the decantation error, 
resulting from the adhesion of organic matter to the particles 
of sand, and from the capillary retention in the sand of the 
water used in washing the sand during decantation ; to the 
above should be added (4) the destruction of the very delicate 
organisms by the sand in the process of decantation. The 
practical value of the method for comparative results in the 
hands of different workers is invalidated by the multiplicity of 
conditions affecting the results; among these are the nature 
and amount of the sand, the care and skill of the worker, and 
particularly the nature of the sample to be filtered. 
Kofoid (Science, vi, 153, Dec. 3, 1897, “On Some Impor- 
tant Sources of Error in the Plankton Method ”) found that 
filter paper (No. 575 Schleicher & Schiill) was more effective 
than the sand filtration method, giving 75% to 85% of the 
planktonts as compared with 40% to 65% given by the sand 
filters. Kofoid has detected the advantage of filtration through 
very delicate porous media, and finds that fine infusorial earth 
is very efficient, and in spite of minor difficulties connected 
with the final separation of the planktonts from the infusorial 
earth he regards this as the most satisfactory method thus far 
devised. I might add that the total weight of material (organ- 
isms, organic and inorganic débris) suspended in water is of 
fundamental importance and can be determined with consider- 
able accuracy by this method, though I see no way to ascer- 
tain the relative proportion of organisms and débris except 
very roughly through the enumeration of the individual organ- 
isms and comparison of the apparent bulk of the masses of 
living and dead material as seen under the microscope. 
Experiments have been made by adding various quantities 
of either corrosive sublimate, picric acid, acetic, and other acids, 
alcohol, and formalin to known quantities of water, with a sub- 
Sequent determination of the volume and constituent elements 
