one or two intermediate depths in addition were taken at all except 

 the very shallow stations. Thus, at Station 37A samples were taken 

 regularly at 0, h, 8, and 13 meters depth. From these data it is 

 possible to compare the mean number of organisms determined by samples 

 from four depths with that determined by samples from only the surface 

 and bottom. Tables h9 and 50 were designed to facilitate such a com- 

 parison. For each of the genera shown, the mean number per liter for 

 the station was first determined from two (surface and bottom) samples, 

 and then from all of the four samples. 



A comparison of the pairs of means in the tables shows that 

 some agree closely and that in every case the individuals of a pair are 

 of the same order of magnitude. That is, in no case does one mean 

 show great abundance and the other great scarcity. The conclusions 

 drawn from the data would have been the same, whether the mean was 

 based on four samples or two samples. If these data from Station 37A 

 are representative of the entire area studied and of both years, it may 

 be concluded that the data of 1929 are as valid as those of 1930 for 

 the purposesof this investigation. The data in Tables h? and 50 show 

 the adequacy of series of two samples as compared with those of four 

 samples, but not the adequacy of the series of four samples themselves. 

 However, if the algae are so distributed that samples taken at and 

 13 meters give essentially the same mean as samples taken at 0, Ii, 8, 

 and 13 meters, it is highly improbable that a further reduction in the 

 interval between sampling points would change the result materially. 



In the laboratory, a measured sample of the sample from the lake 

 was run through the centrifuge. The size of this centrifuge sample was 

 usually one liter, but sometimes smaller or larger samples were used, 

 depending on the abundance of plankton. The plankton was transferred 

 from the centrifuge bowl to a vial and made up to a known volume, 

 usually 25 cubic centimeters. 



The Sedgwick-Rafter method of counting was used. The counting 

 cell had a depth of one millimeter, and the tube length of the 

 microscope was adjusted so that the Whipple micrometer covered one 

 square millimeter of the cell. Ten squares in each of two cells (20 squares 

 in all) were counted, and the count converted to the number of organisms 

 per liter of lake water. With routine procedure, that is, with a centri- 

 fuged sample of one liter concentrated to 25 cubic centimeters, the con- 

 version factor was 1,250, With other sizes of sample and concentrate the 

 factor was as low as 625 and as high as 2,500, It must be evident that, 

 with such conversion factors, statements of the number of organisms per 

 liter in units of smaller size than a thousand would indicate an accuracy 

 which the data do not possess. For that reason figures in tables have 

 been rounded off to the nearest thousand, except those shovd.ng general 

 averages which fall below one thousand. These are shown in smaller units 



1U3 



