matter of some difficulty to locate a sandhill crane's nest, never-
tneless their presence in summer, usually made apparent by VOLS and
behaviour, is unlikely to escape notice. Therefore an answer in the
negative to this enquiry may be considered information of value. If
it is correct, and it would seem to be so, that the species does not
nest as far north as the latitude of Prince George then there is a
very considerable gap of unoccupied territory between the nesting
grounds of this species and those of the lesser sandhill crane.

In the Baker Creek Valley during May, one occasionally was seen
or heard in flight, and on May 20 a nest containing two eggs was
found on a muskeg about three miles north of Puntchesakut Lake, where
a pair had been located a week earlier. This is a large muskeg sepa-
rated into two sections by a high forested island. The portion fre-
quented by the cranes was fairly dry and contained no cover other
than scattered dwarf birch. Along the edge this growth is more
plentiful and interspersed with small dwarf spruce (Fig. 2)s-0 Just
before I reached the edge of the muskeg the female crane rose from
somewhere in this section and, after flying a short way, alighted
again close to the edge; through this cover her cinnamon-coloured
back was visible as she moved from place to place. Once when she
came into the open it could be seen that she walked with long strides--
her neck swaying from side to side--and how brilliant became the bare
crimson skin of her face when the sunlight caught it. An attempt to
obtain a photograph at close range was not successful. She rose some
300 yards away and flew down the muskeg (Fig. 39}. Later the male
joined her and both were heard calling.

The nest consisted of a flat platform 18 inches in widest diameter,
made of dead dwarf birch twigs and a small amount of grass (Fig. 40).

Virginia Rail--Rallus limicola Vieillot.

At Maclure Lake, July 15, the familiar call of this rail was heard
in three different parts of the shoreline marsh. None was seen here or
elsewhere in the region. It is regrettable that a record extending the
species' known limit of range a considerable distance northward. i's not
supported by the evidence of a specimen collected.

Sora--Porzana carolina (Linnaeus).

Recorded, on the basis of voice only, at two localities, viz. ;

Puntataenkut Creek, May 16, May 31, and at two places on Dragon Lake,
June 9.

= 7255