The Coefficient of the Dredge. 283 



ation in the coefficient of the net due to the accumulation of the 

 plankton within it as the net is drawn through the water. Un- 

 questionably the stoppage of the openings of the net by the 

 accumulating catch raises the coefficient, and if the net accum- 

 ulates a sufficient amount of plankton it will wholly cease fil- 

 tering the water. In plankton-rich lakes, therefore, serious 

 error may be introduced from this source. Since lake Mendota 

 during the summer and autumn contains very large amounts of 

 vegetable plankton, it was quite possible that the stoppage of 

 the net should cause errors. In order to determine whether 

 "these errors existed, I regularly made hauls of the net from the 

 bottom of the lake to the surface during the season of 1895 

 and compared the number of Crustacea obtained in the hauls 

 from the bottom with the sum of those caught in the six suc- 

 cessive levels of my series. I append a table showing the num- 

 ber of Cyclops caught in the months from January to July, 

 1895, in order to compare the series and the single haul. It will 

 be seen that the number of Cyclops varies, often considerably. 

 Out of 41 cases prior to July 1, the total haul exceeded the sum of 

 the series in 24 cases and fell below it in 17 cases. There was 

 thus no decided advantage on the side either of the series or 

 the single haul. If the amount of variation in this table be 

 ■compared with the amount shown in the catches of the tube in 

 Table I, it will be seen that the differences are of much the same 

 order as those disclosed by the tube. There is therefore no 

 ^evidence that under .these 'circumstances the net suffered any 

 stoppage in passing through the 18 meters of the lake which 

 altered its coefficient to any marked degree over that of the net 

 used through 3 meters. 



After the first of July Anabaena and similar small plants de- 

 veloped rapidly in the lake, and the amount of vegetable plank- 

 ton increased to a great amount. Under these circumstances 

 the number of Crustacea caught in the total haul varied widely 

 and irregularly from the sum of the series, and soon became 

 uniformly lower than the sum. It was found therefore that the 

 coefficient of the net has been raised by the amount of algae 

 present and the catches made by the total hauls were not em- 

 ployed in reckoning the number of the Crustacea after the first 



