936 SWIFT 
markable nests consist essentially of mucus, secreted by the salivary 
glands above mentioned, which dries and looks like isinglass. Their 
marketable value depends on their colour and purity, for they are 
often intermixed with feathers and other foreign substances. The 
Swifts that construct these “edible” nests form a genus Collocalia, 
of which the number of species is uncertain; but they inhabit 
chiefly the islands of the Indian Ocean from the north of Madagascar 
eastward, as well as many of the tropical islands of the Pacific so 
far as the Marquesas,—one species occurring in the hill-country of 
India. They breed in caves, to which they resort in great numbers, 
and occupy them jointly and yet alternately with Bats—the 
mammals being the lodgers by day and the birds by night.1 
The genus Cypselus, as noted by Willughby, with its American 
ally Panyptila, exhibits a structure of the foot not otherwise ob- 
served among birds. Not only is the hind-toe constantly directed 
forwards, but the other three toes depart from the rule which 
CyYPSELUS. ACANTHYLLIS. MACROPTERYX. 
(After Swainson.) 
ordinarily governs the number of phalanges in the Bird’s foot,—a 
rule which applies to even so ancient a form as Archxopteryx (FOSSIL 
Brrps, p. 278),—and in the two Cypseline genera just named the 
series of digital phalanges is 2, 3, 3, 3, instead of 2, 3, 4, 5, which 
generally obtains in the Class dives. Other Swifts, however, do not 
depart from the normal arrangement, and the exception, remarkable 
as it is, must not be taken as of more value than is needed for the 
recognition of two sections or subfamilies admitted by Mr. Sclater 
in his monographical essay on the Family (Proc. Zool. Soc. 1865, 
pp. 5938-617). Mr. Hartert (Cat. B. Br. Mus. xvi. pp. 434-518) 
recognizes three subfamilies with nine genera and 78 species. 
Their geographical distribution is much the same as that of the 
Hirundinide (SWALLOW, p. 926); but it should be always and most 
clearly borne in mind that, though so like Swallows in many respects, 
the Swifts have scarcely any part of their structure which is not 
formed on a different plan ; and, instead of any near affinity existing 
between the two groups, it can scarcely be doubted by any un- 
1 Mr. H. Pryer has given one of the latest accounts of some of these caves 
in North Borneo (Proc. Zool. Soc. 1885, pp. 532-538), which may be read to 
advantage. 
