572 
Journal of Agricultural Research 
Vol. XXXI, No. & 
not obtained from these two lots of flaxseeds, and the scattering results 
will not be included. 
Again, on September 15,1920, another lot of flaxseeds was started in 
the cages, but owing to the freakish circumstance that adults emerged 
about 20 days later in the season than usual the number finally 
recovered by the cages was too small to give any data worth record¬ 
ing. In a faithful daily examination of the cages made throughout 
the period, it was discovered that ants were attacking and carrying 
away the flaxseeds; because of the unforeseen and extremely unusual 
delay in emergence, they were able to destroy quite a large number of 
them. This is a condition likely to occur whenever the cages are 
used, and it can not be considered other than a disadvantage. 
DISCUSSION 
Of a total of 38,500 flaxseeds used in these experiments, records 
were made on 17,500, from which one may expect 13,125 flies to have 
emerged, and of these 6,278 were recovered by the various cages. An 
additional 385 were recovered by the cages in the first trial, making 
a total of 6,663 flies recovered and counted during the experiment. 
Probably the best arrangement of the information obtained in the 
preceding experiments, and one by which a comparison can be made 
for selection of a cage for any special purpose, is that of a tabulation 
of the relative efficiency of the cages used. Table III shows, on a 
percentage basis, the conditions in each cage as compared to natural 
conditions—the natural temperature, evaporation, and rainfall being 
taken as 100 per cent. Efficiency in recovering flies is shown by the 
average per cent recovered from the number reasonably supposed to 
have emerged in each cage. The results from the indoor trial were 
not included in this rating, since conditions under which this partic¬ 
ular experiment was run were unique. 
Table III .—Cage efficiency, on percentage basis 
Cage 
Temperature 
Evapo¬ 
ration 
Rain 
Flies 
Mini¬ 
mum 
8 a. pi. 
12 m. 
4 p. m. 
8 a. m. 
5 p. m. 
recov¬ 
ered 
A... 
95 
99 
99 
99 
76 
85 
35 
B... 
96 
99 
97 
90 
66 
95 
21 
C_____ 
97 
97 
92 
96 
33 
58 
4 
D... 
94 
97 
89 
95 
15 
00 
48 
E____ 
97 
99 
97 
98 
64 
84 
20 
F 
96 
99 
99 
98 
68 
82 
12 
Control_ __ 
100 
100 
100 
100 
100 
100 
100 
It is found that cage A comes the nearest to being suitable for the 
determination of the date when flies are emerging in the field. For 
the actual recovery of flies, the dark cage D was the most efficient. 
For counting the flies, cages B, E, and F have their advantages, 
whereas cage C seems to be entirely impractical, as is shown in 
Table III. 
