4 
Colorado Experiment Station 
long and 3% inches in diameter. Floats on the side keep the 
trap partly above the water. The cross section of the water 
cut is a semi-circle having an area of 3.36 square inches. The 
trap was made long enough to give a total pore space at least 
tliree times the area of the opening, in order that there would 
Fig. 1 — Seed trap of fine copper wire gauze, used in catching seeds floating on 
stream surface. 
be little or no deflection of the water to either side. Of course, 
in practice it was found that at times when there was much 
material being carried in the ditches, accumulation of debris 
in the trap soon reduced the pore space to a point which 
caused a marked deflection of the current. For example, dur- 
ing the early part of the irrigation season, the ditches are car- 
rying enormous quantities of weed seeds and other material. 
Then, the seed trap might fill to capacity in fifteen or twent}^ 
minutes. On the other hand, during the middle of the irri- 
gating season, a similar seed trap might be left in the ditch 
for twelve hours without filling to capacity. 
Metal eyes were screwed in the floats, and by means of 
cords attached to stakes on the ditch banks, the trap was held 
in mid-stream. 
The traps were collected at the end of definite periods, and 
the seeds and other material washed from them. This process 
was facilitated by a removable gauze cap at the posterior end 
of the trap. The ‘‘catch’’ was laid out on glass plates in a 
dry compartment of the greenhouse and allowed to become 
thoroughly dry. It was then broken up by hand, spread out 
on a clean glass i)late and worked over to separate out the 
weed seeds from debris. This tedious work was often made 
