Vor. 11, Pt. 11] LOOMIS—A REVIEW OF THE TUBINARES 181 
pelago. They were never observed far from land, the greatest 
distance being about fifty nautical miles. They frequented 
both the bays and open sea, and often as many as a score or 
more were in sight at a time. One day at Bindloe Island about 
thirty, in a compact flock, were seen hovering over the water, 
appearing like gigantic black butterflies. Ordinarily, their 
flight is bat-like, but not as erratic as the flight of the Gala- 
pagos Petrels. 
No nesting places were discovered, although these petrels 
undoubtedly breed in the archipelago. The genital organs 
were beginning to enlarge in specimens taken April 28, were 
very large in those taken July 7, and dormant in those obtained 
November 27. 
They often came about the schooner Academy, attracted by 
the refuse thrown overboard, being particularly fond of the 
turtle’ fat. They usually pick up their food while fluttering 
close to the water with their feet dipping. The remains of very 
small fish were taken from the stomach of one bird, the sea 
furnishing it other food besides invertebrates. 
The color in life of the soft parts of adults were: bill and 
orbital ring black; irises dark brown; feet black with a light 
yellow patch in the center of each web. 
In the eighty-seven Expedition specimens the usual point-of- 
wing-formula is: primary 2>3>1>4. The borders of the 
greater wing-coverts and inner secondaries are decidedly white 
in certain specimens. 
The series at hand is sufficiently comprehensive to indicate 
the character of the moult, although every month of the year 
is not represented by specimens. Some renewal is in progress 
in April, May, June, and July specimens, but only in one (se- 
cured June 7) does it affect the primaries. In the single Octo- 
ber specimen, taken on the 14th, a complete moult is un- 
doubtedly progressing ; seven of the innermost primaries have 
already been shed. Late November, December, and early 
January specimens (fifty-nine in number) show the postnuptial 
moult in various stages of advancement. Some individuals 
have lagged far behind, in part accounting for the tardy spring 
and summer moulting. 
