46 THE RELATION OF 8PABROWS TO AGRICULTURE. 
replaced on oak Leaves, with which they blended perfectly, and offered 
to t he birds. All i hree sparrows immediately pounced upon the grass- 
hoppers and greedily devoured t hem. A dozen experiments were 1 pied 
with the song sparrow to test the efficacy of the protective colora- 
tion of weevils. A weevil (Sitones hispidulus) sunk almost beneath 
sand of iis own color was always instantly spied out. The results of 
these experiments, together with the facl thai protectively colored 
weevils and grasshoppers form an important pari of the contents of 
stomachs examined, indicates thai sparrows are nol baffled by such 
devices. 
Protection by means of the ejection of an offensive fluid seems to 
be more effective. Blister-beetles (Epicauta), which possess an irri- 
tating secretion, were repeatedly refused by the three sparrows. 
Disagreeable secretions when coupled with showy or warning colora- 
tion were found, in a number of cases, to repel the attacks of the 
sparrows. A gaudy orange and black harlequin cabbage bug (Mur- 
gantia histrionica) was offered to the birds. The junco and song 
sparrow refused it, bul the white-throated sparrow seized it and after 
some preliminary pecking ate it piecemeal. Two other injurious 
insects, the three-lined flea beetle (Lema trilineata) and the twelve- 
spotted cucumber beetle (Diabrolica 12-punctata), which are warn- 
ingly colored beetles with pungent secretions, were placed in the cage, 
but the sparrows, though exceedingly hungry, refused them. The 
failure of the birds to destroy these pests is in pari offsel by the facl 
that they do not prey on the useful ladybird beetles of the family 
Coccinellidae, which are also both warningly colored and ill flavored. 
Two species of these insects (Adcdia bipunctata and Hippodamia 
maculata) were offered to the birds a do/en limes, but were always 
allowed to crawl out of the cage unmolested. The results of these 
experiments coincided with the data accruing from the examination 
of the contents of stomachs in the laboratory. 
The frequency with which the smaller and less useful forms of 
ground - beet les (Carabidse) are met with in examination of sparrow 
stomachs suggested the carrying out of .some experiments with these 
insects. Carabidse emit volatile irritating fluids which would natu- 
rally be expected to render them immune from the attacks of birds. 
Just how effective this protective device actually is will appear from 
the account of these experiments. 
Two specimens of Amaru impuncticoWs, a carabid about one-third 
of an inch Ion-, were placed in the cage wherein were the three spar- 
rows. The white-throated sparrow instantly seized one of these 
beetles and the BQUg sparrow the other, while the junco was obi 
to content itself with pieces dropped by its mates as they hurriedly 
ate the lidbits. Before .-my further tesl could be made the junco 
and white-throated sparrows were, unfortunately, accidentally killed; 
the Beriee <>f experiments was continued, however, with the song 
