BIRDS OF NORTH AND MIDDLE AMEEICA. 685 



a somewhat similar nest, but with the tube more contracted below, 

 out of seeds of Trixis divaricata, suspends it to a horizontal branch, 

 and covers the outside with feathers; as there is no shelf on the 

 inside, it is believed that the eggs arc glued against the side and 

 brooded on by the bird while in an upright position.'* 



Owmg to the extreme velocity of their flight and the great height 

 at which they usually fly, Swifts are very difficult bu'ds to collect, 

 and consequently comparatively few species of the family, are well 

 represented in collections. On this account it is not easy to obtain 

 a correct understanding of their geographic range and variations; 

 consequentl}^ our present knowledge is, in the case of many species, 

 imperfect and tentative. 



KEY TO THE GENERA OF MICROPODID.E. 



a. Toes with the uumber of phalanges abnormal (2, 3, 3, 3); hallux lateral (on inner 

 side of foot), capable of being directed forward (the foot then pauiprodactyloiis), 

 the inner toe sometimes reversible (cajxible of being directed backward), the 

 toes then paired; tarsi (sometimes toes also) feathered. (Subfamily Micro- 

 podina:.b) 

 b. Outer and middle toes directed forward, in line with axis of tarsus; feet very 

 strong, the tarsi relatively shorter. 



c. Toes naked Micropus (extralimital). c 



cc. Toes feathered. 

 d. Tail slightly forked (for less than one-fourth its length), the rectrices not 

 attenuated terminally; toes more sparsely feathered; feathers of pileum 

 and under parts distinctly outlined, broadly roimded. 



Aeronautes (p. 687). 

 dd. Tail deeply forked (for at least half its length), the lateral rectrices atten- 

 uated terminally; toes more densely feathered; feathers of pileum and 



under parts blended Panyptila (p. (590). 



bb. Outer and middle toes inclined outwards, at more or less of an angle with axis 

 of tarsus; feet proportionately much weaker, with tarsus relatively longer. 



« This style of nest is so different from that of the species of Chxtura, whose nidifica- 

 tion is knowTi, that the proper identification of the species is very questionable. 

 At the same time, it may be remarked that the suggested mode of securing the eggs 

 is highly improbable. 



& Cypselinx (not of Bonaparte, 1838) Sclater, Proc. Zool. Soc. Lond., 1865, 579; 

 Cams, Handb. Zool., i, 1868-75, 253. — Micropodinx Stejneger, Stand. Nat. Hist., iv, 

 1885, 437, in text. — Apodinx Jlartert, Das Tier., Podarg., Caprim., Macropt, 1897, 80. 



cApus (not Apos Scopoli, 1777, prior by anteriority) Scopoli, Intr. Nat. Hist., 

 1777, 483 (type, Hirundo apiis Linuyeus). — Micropus Meyer and Wolf, Taschenb. 

 Deutschl. Vog., i, 1810, 280 (type, Hirundo apus Linnaeus). — Cypsclus Illiger, Prodr. 

 Om., 1811, 229 (type, Hirundo apus Linnaeus). — Cipselus (emendation) Vieillot, 

 Analyse, 1816, 38. — Brachypus Meyer, Vog. Liv.-u. Esthl., 1815, 142 (type, Hirundo 

 apus Linnasus). — Cypcelus (emendation?) D'Orbigny, Voy. Am. Merid., iv, pt. 3, 

 1839, 357. 



This genus is numerously represented in the Eastern Hemisphere, but, so far as 

 known, there are only two American species, both of which belong to the southern 

 continent. 



