206 NATURAL SCIENCES OF PHILADELPHIA. 



one whole fiiraily — Podicipidce ; in all others rectrices occur. A small, definite 

 number of rectrices may be observed to be in inverse ratio to the size of the 

 tail as a whole. The number is ordinarily 12, in those forms with well-de- 

 veloped tail ; as in all Laridcc, most Frocellariidw, most Phalacrocoracidce, the 

 genus Plotus. A few Petrels, as some of the little Procellarice proper [Thalas- 

 sidroma auct.) are said to have only 10 ; the statement requires confirmation ; 

 most of them, as Cymochoreay mihi, and Oceanites K. and B.. have 12. Many 

 larger Petrels, as Fulmarus, Daplion, &c., have 14; Ossifraga giyantea has 16. 

 Among Cormorants the number of rectrices seems to be scarcely generic ; 

 some species having 12, others 14. As a rule Alcidcc have definitely 12 rec- 

 trices. A. impennis is said to have 18. I'haelhon has 16. Sulkke have 12 and 

 14. Tacliypetes has 12. Other genera included in the Spheniscidw, Colymbidw, 

 Pelecanidce and Anatidce have a number varying from 12 to 32, though rarely 

 over 24, and in instances where the larger numbers are approached or reached 

 the tail may have an odd number, as 21, 23.* Not taking into account the 

 small, short, rounded tails above mentioned as " typical," the Natatores furnish 

 great variations in shape of the tail as a whole. It may be long and deeply 

 forked, with more or less filamentous rectrices, as in nearly all Sterninse, some 

 Larinw, the Rhynchopince, and more particularly the Tachypetidx. It is large 

 and square in most Larince ; of corresponding size, and more rounded, or 

 even approaching a wedge-shape, in many ProccUariidie. It is moderate or 

 rather large and cuneate, in one g\x\\, Rhodosiethia rosea, and one Petrel, Ilalo- 

 cyptena mkrosoina. In certain terns [Anoiit>) it is both forked and cuneate, 

 i. e., lateral feathers graduated, central pair shorter than the next. It is short 

 and cuneate in some ducks, and in Spiheniscidie : certain of the latter, how- 

 ever, have much better developed tail, with fewer rectrices, than others. The 

 tail is long, broad, and fan-shaped in Plotida and Phalacrocoracidce, where, 

 moreover, the rectrices are exceptionally stiff. Finally, a few swimmers have 

 curiously modified central rectrices. Thus in Ilarelda glacialis these are long- 

 exserted and filamentous ; a condition exaggerated in most Lestridince and 

 carried to an extreme in P/iaethontidce. 



There are few other modifications of the feathery covering that require 

 notice. Some swimmers are ornamented with singularly modified feathers, 

 chiefly about the head. Many Spheniscidw, and most Alcidie, especially phal- 

 eridine forms, have curiously colored and shaped crests. Nearly if not all 

 Podicipidce are instructed with conspicuous ruffs of long, loose feathers ; when 

 these are deficient, as in Podilymbus, stiff bristles occurr. The bars of stripes 

 on the neck of Colymbidce are of keel-shaped feathers. Cormorants have 

 colored filoplumes about the head ; other, perhaps similar, patches are found 

 on the thighs. Some Anatidx, as Aix, and most Meryince, are conspicuously 

 crested. Petecanidx are crested, and have similarly elongated feathers on 

 other parts, as the breast. The scapular feathers of many Fulifftilince are pe- 

 culiarly curved ; one or two genera of this sub-family have velvety-piled 

 plumage. The well-known general condition of the plumage of penguins has 

 suggested the name " Squamipennes." In many families or genera, plumage 

 extends on the bill in varying degree ; it usually, but not always, reaches the 

 ankle joint, but the tarsus is never feathered, except in one genus, Tacliypetes, 

 and there only partially. No part of the body proper is naked, with the re- 

 markable exception of the gular pouch oi Steganopodes,c&VT\&(i to a maximum in 

 Pelicanus. Other dermal outgrowths ihan feathers are rare; but an example 

 of horny development is seen in the spur-winged genus Pkctropterus. 



The highly interesting modifications that make the swimming foot what it 

 is, affect the whole limb from hip-joint downward. The legs as a rule are 

 placed further back than in Insessorial and purely Cursorial forms. But there 

 is great difference in this respect among Natatores, and a regular gradation 



* I am again indebted to Nitzsch's incomparable work for part of the data for these 

 generalizations regarding numbej of wing and tail feathers. 



1869.] 



