184 Marsh — Limnetic Crustacea of Green Lake. 



XIII.) fifteen centimeters long, by five centimeters broad. The 

 frame is strengthened by three transverse braces. The frame 

 and braces are made of strips cut from sheet brass, one milli- 

 meter thick and two centimeters wide. 



Through the horizontal pieces of the apparatus are drilled 

 two holes large enough so that the heavy brass wire D E will 

 slide easily up and down. To the middle of this wire at E is 

 attached an upright piece which passes through the lower part 

 of the frame B, and strikes against the brace C. The wire is 

 held in place by a rubber band passing around the plate B. 

 The dredge is hun g from this central pin at E, and cannot be 

 detached except as the wire D E is lowered so as to throw the 

 ring off the pin. 



The releasing apparatus is fastened to the dredge rope by 

 copper wire passed through small holes drilled in the upper and 

 lower plates. The messenger is a brass cylinder five centi- 

 meters long and four centimeters in diameter. 



The work of dredging is done from a row boat which is fitted 

 with a sail. The mast is unshipped, and in the mast hole is in- 

 serted an upright about six feet long, to which is attached 

 a cross piece extendin g over the side of the boat. From this 

 cross piece the dredge is suspended by a pulley block, and upon 

 the cross piece is a hook from which the messenger is suspended. 

 The dredge is lowered vertically, and after being raised to the 

 required point, is " set off " by the messenger. When the mes- 

 senger strikes the releasing apparatus the top of the dredge 

 falls over, and it remains suspended by the middle. At the 

 same time the weight of the lead causes the cord around the 

 middle of the dredge to tighten, so that there is a double safe- 

 guard aga inst the entrance of any other organisms — the in- 

 verted top and the stricture of the suspending cord. 



There is one sour ce of inaccuracy in this dredge, and that is 

 the loss of material, when it is released, between the top and 

 the cord passing around the center. My hauls, however, were 

 made through five meter distances, and I do not think that in 

 this distance, the loss would have much effect on the results, 

 and, of course, for comparative work it need not be considered 

 at all. 



