to emphasize their rarity. 



No attempt was made to determine the precision of the counting 

 method. Whipple (1927, p. 101) stated that two examinations of the 

 same sample by the Sedgwick-Rafter method seldom differ by more than 

 10 per cent. Allen (1921) made a statistical study of his method of 

 enumerating marine algae, but did not describe the method in detail. 

 He found that the mean deviation from the mean of several counts of the 

 same sample could be kept within 10 per cent. This indicates a 

 precision of about one half that reported by Whipple. Allen concluded 

 that errors in counting were less important than those arising from 

 inadequate mixing of the sanple before transfer to the counting slide. 



Many published records of phytoplankton counts have lost much 

 of their value through a failure of the authors to record the units used 

 in counting the various algae. Some algae appear as small individual 

 cells while others appear as great filaments or colonies. With the 

 ordinary method of enumeration each kind is given equal weight regard- 

 less of size. In an attempt to avoid this difficulty, Whipple devised 

 a method of counting on the basis of standard units of area or volume 

 (see Whipple, 1927, p. 12li), This method has been used widely by 

 students of sanitation, but not by limnologists in general. The 

 principal drawback to the method is the additional labor involved. 

 When such a method is not used, published records should indicate the 

 units used for each form counted. 



In the present investigation the following units were used: 

 Each unicellular alga, such as Synedra, Navicula, and Stephanodiscus, was 

 counted as one, regardless of size. With algae like Pediastrum, 

 Coelastrum, Oocystic, and Coelosphaerium, which occur in rather definite 

 colonies, each colony was counted as a unit, regardless of size. Some 

 difficulty was encountered in deciding on a satisfactory method of 

 counting the filamentous forms and those which occur in colonies easily 

 broken up in the centrifuging process. In such cases it was necessary 

 to choose units arbitrarily. For Melosira, Oscillatoria, Lyngbya, 

 Anabaena, and Aphanizomenon, a filament 300 microns long was counted 

 as a unit; for Fragilaria a filament 100 microns long. Units for colon- 

 ies of variable size were; Scenedesmus, k cells; Tabellaria, Diatoma, 

 Crucigenia, and Asterionella, 8 cells; Dinobryon, 5 cells; and 

 Merismopedia, l6 cells. In counting, no record was kept of the abundance 

 of different species of the same genus; all species of a genus were 

 counted together. The protozoa which lack plant- like characteristics 

 were so rare in the plankton that they were disregarded in counting, 



IhS 



