430 Birge—Vertical Distribution of Pelagic Crustacea. 
as it falls, pulls back the cover and opens the dredge. The 
dredge is then raised through the desired distance—in our ex¬ 
periments, 3 meters—and the larger or closing messenger is sent 
down. This releases the closing weight, which pulls the cover 
shut, and in so doing raises the opening weight. The dredge 
is then drawn to the surface. It has been found that there is 
no leakage of Crustacea into the dredge during raising. The 
thin brass of the cover is tightly pressed against the frame as 
soon as the upward movement begins, and the guides at the 
edge of the cover present so narrow and tortuous a passage 
that the water does not readily flow through it. Besides, the 
guides are so constructed that the water displaced by the dredge 
flows away from the groove and not into it. An examination 
of the cut (PI. VIII, fig. 1) shows that the arrangement is such 
that the animals are not likely to come in. In our numerous 
hauls it was found entirely tight, so far as Crustacea were con¬ 
cerned. Many times the dredge was hauled from the bottom of 
the lake through water whose upper levels contained thousands 
of Crustacea, and on opening was found to contain perhaps 6 
or even fewer Crustacea, with, perhaps a Chironomus or Corethra 
larva. Experimental hauls, sending down and raising the 
dredge when closed yielded uniformly negative results. 
When the messengers arrive at the surface they are detached, 
the dredge hauled up, and the contents washed down into the 
bucket. This is removed and the catch washed into a collecting 
tube, modeled on those designed by Prof. J. E. Beighard of the 
University of Michigan, made and used as follows: 
The bottom is cut from an eight drachm short homeopathic 
vial, a piece of fine cloth is tied tightly over the neck, and a 
cork fitted to the bottom. A rubber band is placed upon the 
conical tube of the bucket, so that the collecting tube can be 
pushed tightly upon it. The rubber cork is lifted from its 
place in the bucket, the catch runs out into the collecting tube, 
and the bucket is carefully washed down into the tube. The 
water drains off through the cloth, the tube is detached, filled 
from a large jar of alcohol, and placed in the jar, where it floats 
cork uppermost, and requires no more attention until a con¬ 
venient time offers. Most of our collecting was done at one 
