108 
PACIFIC SCIENCE, VoL III, April, 1949 
for taking off. The albatrosses, despite the fact 
that they nest beneath clumps of shrubs, do not 
alight in the midst of the bushes; they always 
walk in to the nest from the nearest open space, 
although in some instances this distance is more 
than 100 feet. The widely spaced plantings of 
ironwood trees do not seem to reduce the num- 
ber of Laysan Albatrosses nesting beneath them, 
perhaps because there is room to fly between 
the trunks and because the plantings are inter- 
spersed with roads, building areas, and other 
open spaces. However, if a great part of the 
surface were covered with dense, low-growing 
shrubs I think the number of Laysan Albatrosses 
would be reduced. Certainly the beach-loving 
Black-footed Albatrosses would not, under such 
conditions, invade the central areas of the islands 
as they now do. The Sooty and Gray-backed 
Terns might also be affected by more plant 
growth; on Midway, colonies of these birds 
seem to center in open spaces and simply overlap 
adjoining Scaevola-c overed areas. 
To establish successfully many of the exotic 
plants it was necessary to bring in soil. After 
1906 the cable company brought in the equiva- 
lent of several shiploads (about 9,000 tons) of 
soil to get trees, shrubs, and grasses started 
around its compound. Later, soil was brought in 
to develop the nursery. The maintenance of a 
small farm with a few cows and chickens has 
added to the fertility of Sand Island. Not only 
has this soil been beneficial in producing a plant 
growth; it has also provided a topsoil better 
adapted to the needs of burrowing birds. It is 
not uncommon to see the burrows of petrels 
and shearwaters and the nests of albatrosses 
filled and covered by blowing sand in areas 
where there is neither humus nor plant roots to 
hold the sand. Burrows in pure sand collapse 
easily-, trapping the birds; entrances to burrows 
of Bonin Island Petrels here are sometimes cone- 
shaped, 5 feet in diameter and 3 feet deep, 
caused by the constant sifting down and col- 
lapsing of the dry sand. It may be observed that 
where possible the birds select partly vegetated 
areas for burrowing. 
Another act of man that will in the future 
help the birds is the addition to the area of 
the islands by filling with material dredged from 
the reef. Many acres of land have been created 
in this manner. At present only a few Laysan 
and Black-footed Albatrosses are using this new 
area which is barren of vegetation and has no 
topsoil. Further, it will take some time for the 
birds to extend their breeding areas to encom- 
pass this new region. Judging from other areas 
occupied by the various species the first forms 
to utilize this coral fill will be the albatrosses, 
Sooty Terns, and perhaps the Blue-faced Booby. 
As soon as low grasses appear the Noddy Terns 
and Gray-backed Terns will move in as they 
have already done on other parts of the island 
which have been reclaimed from war use. The 
Fairy Tern, Hawaiian Tern, and Red-footed 
Booby will not utilize the area until the Scaevola 
gets a good growth, nor will the Red-tailed 
Tropic Bird, which is a ground nester but is 
usually found beneath the Scaevola. It also is 
unlikely that Bulwer’s Petrel and the Christmas 
Island Shearwater will nest on this hard-packed, 
smooth fill until some shrubs or piles of coral 
provide cover. 
There are few reliable figures on the avian 
populations of Midway. Scattered notes on num- 
bers of individuals of various species are to be 
found, but it was not until the "Tanager” Expe- 
dition visited Midway in April, 1922, that any 
attempt was made to secure data on the popu- 
lations of all species inhabiting the atoll In the 
period 1922 to 1941 no papers of any kind 
were published on the birds of Midway. Hadden 
( 1941 ) gives incomplete estimates of some 
populations based on observations over a period 
of at least a year. Fisher and Baldwin (1946: 
4 ) estimated the numbers of all species present 
in May, 1945; using the same method whenever 
possible, I took a census of all species present 
in December, 1946. It may be observed from 
this brief history that not only are there few 
census data available but that the data have 
not been secured in comparable months of the 
year and that no standard procedure has been 
