owded habitation. With the Mason bee of the Walls,
I see the number of larvae vary, in different cells, between four
and twenty-six; with the mason bee of the Sheds, between five and
thirty-six; with the three-horned Osmia, who supplied me with the
largest number of records, between seven and twenty-five; with the
blue Osmia (Osmia cyanea, KIRB.), between five and six; with the Stelis
(Stelis nasuta), between four and twelve.

The first return and the last two seem to point to some relation between
the abundance of provisions and the number of consumers. When the mother
comes upon the bountiful larva of the masked Anthophora, she gives it
half-a-hundred to feed; with the Stelis and the blue Osmia, niggardly
rations both, she contents herself with half-a-dozen. To introduce into
the dining room only the number of boarders that the bill of fare will
allow would certainly be a most deserving performance, especially as the
insect is placed under very difficult conditions to judge the contents
of the cell. These contents, which lie hidden under the ceiling, are
invisible; and the insect can derive its information only from the
outside of the nest, which varies in the different species. We
should therefore have to admit the existence of a particular power
of discrimination, a sort of discernment of the species, which is
recognized as large or small from the outward aspect of its house. I
refuse to go to this length in my conjectures, not that instinct seems
to me incapable of such feats, but because of the particulars obtained
from the three-horned Osmia and the two mason bees.

In the cells of these three species, I see the number of larvae put out
to nurse vary in so elastic a fashion that I must abandon all idea of
proportionate adjustment. The mother, without troubling unduly whether
there be an excess or a dearth of provisions for her family, has filled
the cells as her fancy prompted, or rather according to the number of
ripe ovules contained in her ovaries at the time of the laying. If food
be over-plentiful, the brood will be all the better for it and will grow
bigger and stronger; if food be scarce, the famished youngsters will not
die, but will remain smaller. Indeed, with both the larva and the full
grown insect, I have often observed a difference in size which varies
according to the density of the population, the members of a small
colony being double the size of their overcrowded neighbors.

The grubs are whit