70 Director's Annual Report. 
or eight or ten will, not infrequently, fly from the forests at the 
edge of these precipices, thousands of feet above the valley below, 
and with little concern, steer a course that will take them ina 
direct line as far as the eye can followthem. At nightfall they are 
to be seen returning from their feeding grounds lower down on the 
mountains or in the valleys flying swiftly and directly in the higher 
altitudes to roost. This daily migration over open land or from 
place to place is a very strongly marked chara¢teristic of this bird. 
A number of nests—but unfortunately no eggs—were secured. 
Iam convinced that the nesting season cannot be well defined. 
In substantiation of this opinion, I find that in the series of skins 
collected, two are immature, while perhaps five times that number 
of brownish-colored young were noted among the hundreds of 
adults seen at close range. ‘This fact, coupled with the mating 
performances and the enlarged testes and ovaries examined, would 
indicate that on this island, at least, an occasional late brood is 
reared. 
The series of eight nests do not vary in material or location 
from those elsewhere described, to a degree sufficient to warrant 
their redescription. However, a deserted nest (Mus. No. 4683) 
taken from twenty feet up in a moss-covered Ohia tree which was 
growing in the heart of the Puualu forest, 1s somewhat extraordi- 
nary in that it has returned to the original elements of which it 
was composed zz s7fu, leaving a replica of itself in living green 
moss. ‘The structure is of some years standing evidently. Old 
enough at least, so that the moss and sticks of which it was com- 
posed have had time to almost completely disintegrate. That 
portion which remains has, in the meantime, become completely 
covered with living moss, so that the shape and general appear- 
ance of the nest is retained with just enough of the old structure 
remaining to account for its form and history. 
Chlorodrepanis kalaana (Wilson). 
Without here entering into a discussion of the very minute 
characters that have been used to separate this species from the 
very closely allied forms from Maui, Kauai and Oahu, it will be 
sufficient to say that the series of seventeen birds secured on Molo- 
kai, confirm rather than disprove the conclusions formerly arrived 
[160 | 
