1920] Allen: Plankton of the San Joaquin River 23 



PROCEDURE . 

 Standard Concentkatton 



In preparation for counting, all catches were first brought to a 

 volume of 100 cubic centimeters by addition of water or, in a few 

 cases, by decanting some fluid. If the concentration was too great at 

 this volume, dilution was made to 400 or 800 or 1,600 cubic centimeters. 

 No other quantities were used because the increase in difficulty of com- 

 putation would offset the difficulty in counting. This was due to the 

 fact that the above quantities were used often enough to make the for- 

 mation of computing tables useful for them. 



Method of Filling Sedgwick-Raittek Cell 



In the process of filling the cell after thorough mixing a little more 

 than 1 cubic centimeter was taken quickly into a pipette. The cell was 

 then filled as rapidly as possible until the cover slipped into place. This 

 never occurred completely until there was a slight excess of fluid. The 

 excess was immediately taken back into the pipette, leaving the cover 

 glass flat. The possibility of some error is evident, but experiment in- 

 dicated that this method gave more even distribution in the cell than 

 any other and that errors were not appreciably more frequent than 

 those attending other methods. The method was followed throughout 

 the whole series and every detail was handled by the writer, so that 

 there must have been practical uniformity. This would certainly re- 

 duce the significance of any error which may have been incident to the 

 method. 



Masing and Recording the Count 



After filling the cell satisfactorily, a rapid survey was always taken 

 in order to estimate the relative amounts of plankton and non- 

 plankton. The estimate was then recorded in percentage of silt. With 

 these preliminaries completed, the work of counting was begun, fifty 

 fields being always counted. This made a total of 60. .5 cubic milli- 

 meters. Counting was begun on the proximal side of the slide at 

 point 32 on the lateral scale of the mechanical stage. The field next to 

 tlie wall of the cell was not counted, but the next five were taken con- 

 secutively. This process was repeated at point 16, then at point 80 

 on the proximodistal scale at the right end of the cell, followed by 

 point 87. Points 16 and 32 distal, 87 and 80 left, were then taken in 

 order. The detail count was completed by ten fields from point 15 to 



