HYMENOPTERA. — APID®. 273 
ping off the down with their toothed jaws (fig. 91. 2.), for the pur- 
pose of forming their nests. Of this an account is given by Mr. 
Kirby, as well as the description of a nest, supposed to be of this 
species, found in the key-hole of a garden door. Although holes in 
trees are the more ordinary situations for their nests, they seem to 
have an especial liking for the latter locality, since Mr. Anderson, the 
ingenious curator of the Botanical Garden at Chelsea, has also pre- 
sented me with a nest found under precisely similar circumstances, 
and from which I reared specimens of the Anth. manicatum. There 
were twelve or fifteen cells or cases (fig. 91. 20.), consisting ex- 
ternally of a loose covering of white down (20. a), within which was 
another covering, more compact and smooth on the inside (20. b), 
and within this was contained an oval cell, of a strong coriaceous 
texture, and of a chestnut colour (20. c). This latter I consider (as 
does also Mr. Kirby from subsequent observations (see Introd. to Ent. 
vol. i. p. 439. note), to be the cocoon, formed by the larva itself, be- 
cause some of my woolly cases contained a mass of matter apparently 
consisting of dried pollen paste, the egg deposited with which had 
probably on some account proved abortive ; and in these there was no 
oval chestnut-coloured cocoon. It was in February that this nest was 
discovered, at which period some of the cells were empty, the inhabit - 
ants having forced off a circular cap (jig. 91.21.) from the top of 
the cocoon, and escaped ; others, however, contained full-sized grubs. 
As the other closed cells did not produce insects, I opened them 
after keeping them more than two years, when I found that each 
enclosed a dead female insect, which had evidently not attained suffi- 
cient power to make its way through the case. Hence I am induced 
to believe, either that the females do not appear until some time after 
the males have quitted their cocoons, or that the empty cells were 
the construction of a former year. The point at the top of my co- 
coons, which Mr. Kirby calls a funnel, does not form an aperture, nor 
is there any corresponding opening in the woolly covering. 
Some of the species of the genus Osmia are termed mason bees, 
since the materials of which they construct their nests are minute 
grains of sand, cemented together with a glutinous secretion, and 
which are placed by the insects on the angle of a wall, or the crevices 
between bricks, &c.: they are of a sufficient size to contain from three 
to eight, or, according to Geoffroy, fifteen cells, placed irregularly. 
(Réaumur, tom. vi. mém. iii.) Other species of Osmia (O. bicornis, 
VOL. II. T 
