280 Birge — The Crustacea of the PlanJcton. 



tered. As a matter of fact, the net caught about twice as many 

 Crustacea as the tube, thus indicating that its coefficient is about 



two. 



In this method of determining the coefficient the quantities 

 compared are by no means uniform; indeed, it is known that 

 the number of Crustacea caught in a given haul of the tube may 

 be only one-half the number caught in a second haul within a 

 few seconds. A single comparison has therefore very little 

 value and accuracy in the determination of the coefficient by 

 this method can be reached only by a considerable number of ob- 

 servations. In my own work I made use of six sets of obser- 

 vations, taken on May 14th, October 12th and 25th, 1895, Feb- 

 ruary 25, May 18th, and July 11th, 1896. By distributing 

 the observations over so long a time it was possible to get at 

 the coefficient of the net at different times in its life and under 

 different conditions of plankton. In May the number of Crus- 

 tacea is at a maximum, and the amount of algae is small. In 

 October the number of Crustacea is considerable, but the veget- 

 able life is at a maximum ; while in February the amount both 

 of animal and vegetable life is of course small. From four to 

 six pairs of observations were taken in each set. The ratio of 

 the catch of the tube to that of the net was computed for each 

 observation in the set, and the average of these ratios was com- 

 puted, using the method of least squares. As a result of 

 these determinations, the following ratio was established : 

 Tube : net :: 49.85 : 100. The probable error of the deter- 

 mination is ± 1. The appended table shows the general results 



Several facts appear from the table. It will be noticed 

 that the amount of difference between the maximum and mini- 

 mum numbers caught varies greatly on different occasions. It 

 is plain also that the net shows no greater amount of variation 

 on the whole than does the tube. On the contrary, on those 

 occasions where the numbers are approximately constant in the 

 tube, they are similarly constant in the case of the net; and 

 ■where the numbers vary considerably in the case of the net, 

 they vary to much the same degree in the case of the tube. 

 There is therefore no reason to suspect any considerable irregu- 

 larity on the part of the net due to the stoppage of its openings, 

 or to any other cause. 



