22 THE RELATION OF BPARROW8 TO A.OSIOULTUBE. 
the roasting-ear stage, and feed on it from the time it is put in the 
crib until wheal comes in the milk again in June. There is scarcely 
a grain thai they do not injur*', while with the native sparrows the 
reverse seems to be true. The latter eal a little grain, but seldom 
does it amount to more than 5 percent of the year's food, a modest 
fee for their service when it is considered thai the meadowlark, one 
of the besl birds of the farm, takes 13 percent of its food in grain, 
the crow 35 peicent. and the crow blackbird 17 percent. 
The most serious charge thai can be brought against sparrows is 
that they distribute noxious plants, the seeds of which pass through 
their stomachs and germinate when voided from the body; and 
this, though not strictly germane to tin- subject under considera- 
tion, will he treated of here as the most appropriate place. Spar- 
rows do nol distribute catbrier, poison sumach, and poison ivy. as 
do many birds, but it is probable that they do, to a certain extent, 
disperse the seeds of such weeds as amaranth, gromwell, and spurge. 
However, it seems tikely that this agency of seeding down farms to 
weeds is infinitesimal when compared witli the dispersion of weeds 
caused by the use of manure containing weed seed and the plant- 
ing of impure seed, which often contains seeds of foreign weeds 
of the worst stamp. The digestive apparatus of sparrows has the 
power to crack or crush tin 4 seeds of crab-grass, pigeon-grass, pig- 
weed, lamb's-quarters, and most other seeds, including the hard 
drupes of the blackberry. 1 have examined thousands of stomachs 
of sparrows containing ragweed, and have never found an unbroken 
seed. The outer ribbed shell of the akene is cracked and not 
swallowed, but parts of the true seed coat in the shape of angular 
fragments •"> to :> mm. long, which are dirty gray externally and green- 
ish white internally, are usually found during stomach examination. 
Uncrushed cotyledons are seldom met with. These facts, which 
hold also when seeds of wild sunflowers and polygonums are eaten, 
seem to preclude the possibility of subsequent germination. Con- 
cerning the likelihood of the germination of the seeds of weeds that 
are grasses ii may be staled that time and again tree sparrows 
which have \^'<\ on pigeon-grass have been examined, and it has been 
found thai while their gullets contained from l<»<» to 300 whole pigeon- 
grass seeds with the inner glumes removed, the gizzards were tilled 
with a pasty mass of endosperm containing not more than a dozen 
whole seedx. But with the harder, smaller seeds the possibility of 
germination i^ better. The digestive organs, although they have the 
power of cracking such seeds, nevertheless occasionally allow some to 
pass OUt in a perfect condition, as was shown by an experiment with 
a captive song sparrow in which amaranth seeds were voided unin- 
jured and germinated very well. Birds take seeds for food, however, 
and it seems probable that such use would preclude the evacuation of 
any but a most insignificant proportion of uninjured seeds. 
