676 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol 123. No. 4. December 2011 
FIG. 7. Scatterplots showing sexual dimorphism in size of measurements of trunk skeletons of Wake Island Rail 
(Gallirallus wakensis). • = male; O = female. 
of them as they are always concealed by a screen 
of twigs” (Weimore in Olson 1996:111). The 
experience of Spencer (1936) thal the rails could 
easily pass through 50-mm mesh indicates they 
could also squeeze through interstices in vegeta¬ 
tion that size or smaller. 
“The species has much personality, being 
serious business-like little birds....they look like 
Hungarian Partridge [Perdix perdix ] and act like 
chickens” (Lyons 1939, June). “Wake Island 
rails, larger than those of Midway, ran about here 
and there, but not underfoot. Here at Wake, they 
give a definite impression of barnyard fowl” 
(Miller (1936:694). “The bird compares in size to 
the three weeks old baby domestic chicken” 
(Spencer 1959). 
Rat control biologists, operating mainly on 
Peale Island in 1936, were concerned about the 
effect that poisoning might have on the rails. 
“One full day was given over to making an old 
chicken pen. rail escape proof. The pen was 
originally constructed with Standard Chicken 
Wire f2 inch = 50 mm mesh] and from sad 
experience it was found that the rails could 
squeeze through this size of mesh wire. Our only 
salvation was in boarding up the sides high 
enough to prevent the rails from jumping over. 
One full day and one evening was spent 
attempting to catch rails with the use of traps, 
nets, and flashlights, but to no avail. Only 15 birds 
were caught. It was quickly decided that our plan 
calling for the corralling of the rails, before 
placing the poison bait in the field was going to be 
entirely unsuccessful, as in order to corral 4-500 
