52 PLANKTON OF WISCONSIN LAKES 



play the predominant role, the increase in ash is correspondingly 

 marked because tttair silicious shells make a very substantial contribu- 

 tion to the ash. 



Constituents of the Ash 



Silica (Si0 2 ). Quantitative determinations of the silica were made 

 for substantially all of the samples and the results are indicated in 

 the general table (No. 43, p. 202) ; a summary of the results is given 

 in table 14 (p. 190). The latter table shows that the maximum per- 

 centages of silica for the five full years ranged from 31.6 per cent of 

 the dry weight of the sample in some material collected in 1912 to 37.2 

 per cent in another sample secured in 1915 ; that is, in three of the five 

 complete years samples were obtained in which the silica constituted 

 more than a third of the dry weight of the net plankton, while in the 

 other two years the maximum percentages fell only a little below one- 

 third. In 1911 the highest percentage of silica was found in August, at 

 which time a crop of diatoms predominated; in 1912 and in 1915 the 

 highest percentage was found in material collected in October, while in 

 1913 it was noted in December and in 1916 in September. The smallest 

 percentage of silica noted for the entire series of net catches from Lake 

 Mendota was found in sample No. 314 which was collected on June 30- 

 July 3, 1913. The mean percentage for the complete years was lowest 

 in 1911 and highest in 1915. 



In 1914 and in 1917 the observations were discontinued on July 2 

 and June 1, respectively, and they show much smaller maximum per- 

 centages of silica than the complete years, more especially the former. 

 This is accounted for by the fact that the largest crops of diatoms come 

 in late summer as in 1911, or in the autumn or early winter as in the 

 other four full years. The mean percentage of silica for 1914 is very 

 much lower than those of the complete years, but that of the part year 

 1917 is somewhat higher than that of 1911, but distinctly lower than 

 those of the other four full years. 



A comparison of the mean percentages of ash and silica for the com- 

 plete years shows that the latter is the most important constituent of 

 the ash. In general the mean percentage of the silica comprises about 

 half or more of the mean percentage of ash. This does not hold true for 

 the individual samples, however, because there is a wide variation in 

 the proportion of silica in the ash ; that is, the former comprised as little 

 as one twenty-fifth of the ash in one sample and more than four-fifths 

 in another. It is derived mainly from the diatoms of the net plankton 

 so that the amount is smallest when the diatoms are scarcest and largest 

 when they are most abundant. The silica, in fact, serves as a good index 

 of the relative abundance of diatoms in the samples. Thus, with the 

 exception of 1911 when the maximum amount of siMca was found in 



