Skip to main content

Full text of "The structure and classification of birds"

See other formats

, . I ; ..-. '- 

" . - ft ' - i mm '' 
m :- . . : -, :-;.o: 

'' . - : m - icfM 
ap ' ''; P>lffl 


-.. - . 
- ' ' ra ' ; v 

. ' *''''. ..' . .*' 

.; , ; ..: .-. 



u-.-x.- " ? " : ' 

- 3 , <.: ; .-. 

afflSfa * ; . '.'"-.- 

*:->'.".,-;.; .,;,, ."..,.,'/.- M I ' g . ;;; 

.-"," : - I SPSS '" ' 

^.-.^'- " ... WSS : -' _, - 

M i I ! 


>--- .. . ' : 

-' V E ' - '. 


I I " ' ; - : ', 

;: r' v ' '".- : -' . 

En ' . - 


, , 3 

. - 



Marine Biological Laboratory 

l, 1 

Accession No._ 

Given By T - H. ^OJ OPV, Jr. 


,I3L Li ;.a 












P 1! E F A C E 

IT was the intention of my predecessor in the office of 
Prosector to the Zoological Society, the late Professor 
Garrod, F.K.S., to write a treatise upon bird anatomy. 
This intention was so far realised that a nearly complete 
account of the anatomy of the fowl, with the appropriate 
illustrations, was actually drawn up ; it was proposed that 
this should be followed by a second part, in which the general 
anatomical characters of the different groups of birds were to 
be stated. Of this second part I have beside me some thirty 
sheets of MS. Professor Garrod 's successor in the post of 
Prosector to the Zoological Society, the late Mr. W. A. 
Forbes, had every intention of finishing this work com- 
menced ; but unfortunately death took place before any 
actual additions had been made to the MS. left by his 
predecessor. I have, on the kind encouragement of Mr. 
Sclater, determined to make an attempt to carry out this 
plan of my two forerunners, and the present volume is the 

It must be admitted that a handbook upon bird anatomy 
was more wanted at the time that it was first conceived by 
Mr. Garrod than it is at the present day. Zoologists had 
then nothing of a general character save the incomplete 
fragment of Bronn's ' Thierreich ' and the sections devoted 
to bird anatomy in such comprehensive works as those of 
Meckel and Cuvier. We have now two treatises of first- 
rate merit, that of Fiirbringer and Dr. Gadow's completion 
of the section ' Aves ' in JSronn's ' Thierreich.' Professor 


Garrod's intended work differed from either of these in that 
he meant to preface it with a detailed account of Gallus. I 
have not thought it useful to follow him in this ; for we 
have an excellent treatise dealing with one particular bird type 
in Dr. Shufeldt's book upon the ' Kaven.' Instead of this 
I commence with a general sketch of bird structure, purposely 
avoiding histological detail and the elaborate description of 
anatomical facts which are not, in the present state of our 
knowledge, of great use in classification. The main part of 
this book is the account of the structure of the different 
groups of birds. It was upon this aspect of the subject that 
Mr. Garrod intended to dwell most fully. Dr. Gadow has 
also treated bird anatomy from this point of view ; the con- 
cluding section of his contribution to Bronn's ' Thierreich ' is 
devoted to an enumeration of the distinguishing characters 
of the groups of birds. I have, however, treated of this 
matter more fully, and have incorporated more facts (some 
of them recorded for the first time) in the systematic part of 
my book, than did Dr. Gadow. I have felt it to be useless 
to attempt to vie with Professor Fiirbringer's magnificent 
treatise upon birds. To deal with all the organs of the body 
as fully as he has done would require more space than it 
would be probable that any publisher would be disposed to 
allow me. I believe, however, that I have been able to note 
the principal facts in the anatomy of the different orders of 
birds, and that nothing of first-rate importance has been 

Moreover under each section I have referred to the 
majority of the memoirs already published, so that the reader 
can supplement, where it is necessary, the facts which I 
myself detail. These references, I may remark, have been 
(with a few exceptions) carefully verified ; and although my 
bibliography of the subject is not complete it is, I hope, 
without important deficiencies. 

The facts of bird structure contained in this book are to 


some extent drawn from published memoirs ; but the 
majority of them, especially those relating to osteology and 
muscular anatomy, have been verified by myself ; and I have 
also laid under heavy contribution the note books of my two 
predecessors already mentioned. Some of the more import- 
ant of these facts are illustrated by woodcuts and ' process ' 
blocks ; for these I am indebted to the liberality of the 
Publication Committee of the Zoological Society of London 
and to the editors of the ' Ibis.' 

Finally, I regret to have to acknowledge that in several 
instances I have used two names for the same bird, a fact 
which I did not discover in every case until it was too late 
for alteration in the text. I hope, however, that any diffi- 
culties arising from this error on my part will be obviated by 
the cross references in the Index. 


July IS'.)*. 




THE FOOT . . 1 


FKATHERS ........... 

PTERYLOSIS ...... .... 14 

ALIMENTARY CANAL ......... 19 

Tongue . 19 

Teeth . . 20 

(Esophagus ..... .21 

Stomach ........ 

Intestine ........ 

Caeca 30 

Liver ......... .31 

Gall Bladder ... 33 

Pancreas ........ 

Cloaca ....... . . 84 

Bjirsa Fabricii .... 3. r > 


THE CCELOM . . . 37 


Heart ... .48 

Arterial System ......... 52 

Venous System ......... 55 


Trachea . . ... .58 

Syrinx 60 

Lungs and Air Sacs . . . . . . . 7 1 


Muscles of the Fore Limb ....... 78 

Muscles of the Hind Limb . . . . . 90 

Muscles of the Neck and Trunk 103 

Caudal Muscles ......... 107 

Abdominal Muscles 108 

Hyoiclean Muscles . . . . . . . . . 108 

Muscles of the Head .... ... 110 

OSTE()L(><;\ . 11] 





Vertebral Column 

. Ill 


. . 119 

The Shoulder Girdle . 

. 1-20 

The Fore Limb ...... 

. . I2;j 

Sternum ....... 
Pelvis ...... 

. 127 

Hind Limb ....... 

. 134 

The Skull ... 

. . 135 


. 151 


. . 153 


. 159 

ORNITHUR.S: . ... 

. . 167 


. 167 

PASSERES ........ 


PICI ... 

. 1S3 

Picid* ........ 

. . is:; 

Bucconid* . ..... 

. LSS 

Rhamphastitlffi ...... 
Capitonidte . ..... 
ALCEDINES . . .... 
COLII . . . 

. . 189 
. 192 
. . 197 
. 201 

TROGONES . . . ... 

. . 202 

CORACLE . . . . 

. 20 1 

Coraciidu 1 ....... 

. . 204 

Meropidtc ....... 
Monaotidae . . .... 

. -MiS 
. . 210 


. 212 



Bucerotidu' .... . . 
Upupidse . ..... 

. . 215 
. 222 


. . 224 


. 2;J1 

STRIGES . . .... 
1'siTT VCI 

. . 244 

.MUiOPH.-VGI ...... 

. . 272 

. 2H2 

OPISTHOCOMI . . . . . 

. . 2N.1 

GALLI .... . . 

. 291) 

CoLUMBtf: . . . . 

. . .",05 


. ."15 


. . ;(1 ( .) 


. :',21 

OTIDES . . . ... 

. . :;:;i 

I.I.MICOL.15 .... 

. 336 

(I'ldicneiuidi!' .... 

. . 345 





Chimiididir .... 

. . 347 


. 349 

Glareolida; .... 


LaridiP .... 

. 350 



GRUES ..... 


(IrnidiP .... 


llhinochetid;: 1 



Psophiidie .... 


Eurypygidfp .... 










Sl'IIKNISCI ..... 



. 402 


. . 419 


. 420 

Ciconiidu' .... 

. . 422 

Ardeidn- .... 


Balaeuicepidie .... 

. . 433 

Plataleidse .... 

. 434 


. . 445 


. 451 


. . 456 


. 46!) 


. . 472 

Falconidsv .... 

. 472 

Pandionida? .... 

. . 478 



Cathartidio .... 





. . 498 



1 Jinornithidse .... 







via. PA;P, 

1. Wing of Golden Plover (after Goodchild) . ... 10 

2. a. Cubital Reiniges of Pheasant, b. Cubital Remiges of Golden 

Eagle (after Wray) . . 11 

3. a. Maims of Ostrich, showing Attachment of Reiniges. />. 

Manus of Ostrich, Upper Surface, showing Remiges and 
Coverts, c. Digits of Embryo, showing Reiniges and Coverts 

(after Wray) 13 

4. Feather showing Aftcrshaft (after Sclater) . . . 18 

5. A, Lower Mandible of Indian Darter, t, Rudimentary Tongue. 

-B, Tongue in Profile (after Beddard) 20 

6. Head of Lorius, showing Extended Tongue with Brush Tip 

(after Gar rod) 20 

1 '. Intestinal Loops (after Gadow) ...... 24 

8. Alligator Mississipicnsis ; Alimentary Tract (after Clutl liters 

Mitchell) 26 

9. Chaunti chavaria ; Alimentary Tract (after Chalmers Mitchell) 27 

10. Art/its yiganteus, Chick; Intestinal Loops (after Chalmers 

' Mitchell) 27 

11. Casita rius ; Intestinal Tract (after Chalmers Mitchell) . . 28 

12. Struthio cameltis; Intestinal Tract (after Chalmers Mitchell) 28 

13. Haliaetits albicilla ; Intestinal Tract (after Chalmers Mitchell) 29 

14. Ara ararauna ; Intestinal Tract (after Chalmers Mitchell) . 29 

15. Par us major (after Chalmers Mitchell) ..... 31 

16. Alimentary Viscera of the Indian Darter (after Beddard) . . 32 

17. Duodenum of Syrrhaplcs (after Brandt) 33 

18. Duodenum, Bile Ducts, and Pancreatic Ducts of another Si/r- 

rhaples (after Brandt) . . . . . . 33 

19. Duodenal Loops of Bhea amcricana (after Gadow) . . . 34 

20. Duodenal Loop of Rh. Darwin i (after Gadow} . . . . 34 

21. Cloaca of Chauna drrbiana laid open from in front (after 

Forbes) . . . . 35 

22. Two Types of Bursa (after Forbes) . . 36 

23. Respiratory Organs of Duck (after Huj-lr//} . . . s;> 


rui. I'A'-I 

'21. Respiratory Organs of Apteryx (after Hu.rlcij) . ;-','. 

23. Diagrammatic Transverse Section of Emu, to show the Projec- 
tion of Oblique Septum (after Beddard) 40 

26. Diagram of a Transverse Section through Thorax of Duck 

(after Bed/lard) .41 

27. Similar Diagram of Crow (Corvtis capcllaunx} (after Beddard). 41 
2(-'. Viscera of Rook displayed by Removal of Abdominal Walls 

(after Bed/1 /i nh . . 42 

29. Abdominal Cavity of Bit con- us (after Beddard) ... 44 

yO. Heart of Fowl, Interior of Right Ventricle (after Lankcxtcr) . 49 
yi. Heart of Aptcryx, Interior of Right Ventricle with Attachment 

of Papillary Muscle cut through (after Lankexler) . . 49 

o2. Normal Avian Carotids (after Gar rod) 52 

yy. Carotids of Bittern (/:ft<'r Gnrrod) 52 

yi. Carotids of Flamingo (after Gnrrod) 52 

y5. Carotids of Cacatna (after Garrod) 52 

:-)6. Carotids of Passerine (after Garrod) 53 

o7. Abnormal Arrangement of Carotids, where the Left is Super- 
ficial in Position (after Garrod) . . . . .";; 
:JS. Windpipe of Selcucides niara (after Forbes} . . . . 5S 
o9. Breast Region of Man ucodia (after Beddard} . . . . 59 

40. Syrinx ot Indicator, Enlarged (after Garrod} lil 

41. Syrinx of Cymbirhynchus (after Forbes) . . (>2 

42. Syrinx of Balceniceps (after Beddard) . . .'. . . I'rJ 
4y. Syrinx of Struthio (after Forbes) . . . Ii4 

44. The same Syrinx from behind . . . . . . . (14 

45. Svrinx of Vanella cayennensis, from in front (after Garrod) . (515 
4(5. The same from behind (after Ga rro/l) . . . . . 67 

47. Trachea of Tantalus loeulator. a. From Front, b. From Side 

(after Garrod) UK 

48. Syrinx of Steatorn in. Front View (after Garrod) . . . (19 

49. Tensores Patagii of Pluvnicoptenus (after Weldoit) . . K4 

50. Muscles of Leg of Pulainedea. Outer View (after Beddard 

and Mitchell) . . . 91 

51. Leg Muscles of Palantedea, Inner View, illustrating 15iceps and 

its Sling (after Beddard and Mitchell) 94 

52. Leg Muscles of Baleariea. The Ambiens Tendon is cut (after 

Mitchell) 95 

5y. Leg Muscles of OpisthoComus (after Mitchell) ... 96 

F>4. Flexor tendons of Gallus bankiva (after Garrod) . . . 100 

55. Flexor tendons of Ap/eri/.r Mantelli (after Garrod) . . . 100 

56. Flexor tendons of Tiiuutiiculuy aUtudaritts (after tlarrod} . . 101 


KM:. r.\<;\<; 

57. Fiexor tendons oi' iiuccrox rltiinn-rrox (after Gurroil) . . 101 

,08. Flexor tendons of Meanhrma (after Garrod) . . . 102 

59. A Passerine Foot (after Gar rod) . . . . . .102 

60. Caudal Muscles of Palaniedea (after Beddard and Mitchell) . 107 

61. Hyoidean Muscles of Opisthocomus (after Mitchell) . . 109 

62. Pelvis of Apteryx. From Beneath (after Mivart) . . . 113 

63. Lumbar and Sacral Vertebra of an Immature Ostrich (after 

Mivart) 113 

64. Ribs, Sternum, and Pelvis of Cliuit</a Burmeiateri (after 

Beddard) 11-1 

6"). Last Two Vertebne of Strutliio (aftt r Mirarl) . . . . 116 

66. Atlas of Emu (aflcr Mivart) ... ... 118 

67. Axis of Emu (after Mivart) . US 

68. Atlas of Cassowary (after Micarl) 118 

69. Development of Shoulder Girdle of Chick (after Lindsay) . 122 

70. Radius and Ulna of Metopidi/is (after Forbes) . ... 125 

71. Digits of Ostrich (after Wray) .126 

1'2. Sternum of Lopliophor/ts i mpeyanus (after Hurley) . 128 

73. Sternum of Podica sencgalensis (after Beddard) . . . 128 

74 Sternum of Emu (after Mivart) 129 

75. Pelvis of Dinornis (after Mivart) iy2 

76. Hyoid of Laihamus discolor (after Mivart) . . . 137 

77. Skull of Ehca. Ventral View (after Huxley] .... 140 

78. Skull of Dacelo (after Huxley) ... . . l jo 

79. Skull of Alca (after Huxley) .141 

80. Skull of Coi-vtis (after Huxley) . 141 

81. Skull of Psopliia. Lateral View (after Beddard) . . . 14iJ 

82. Skull of Larns. Dorsal View (after Garrod) . . . . 14;; 
8:j. Skull of Furnarius. Dorsal View (after Garrod) . . . 14;; 

84. Skull of ChloejjJutga Magcllanica. Back View (after (farmd) 146 

85. Syrinx of Etiryhemus. Front View (after l<\)rl>es) . . . 17y 

86. Syrinx of Cyinbirliyncluts. Side View (after Forbes) . . 17S 

87. Syrinx of Philepitta. Side View (after Forbes) . . . 181 

88. Syrinx of Philepitta. Front View (after Forbes) . . . 181 

89. Svrinxof Xenictts. A. Front View. B. Back View (after Furlirn) 181 

90. Skull of Woodpecker (Gccinus viridis). Ventral \ie\v (after 

Garrod) ... ...... 187 

91. Feather Tracts of Mcgaheiiia aniatiea (after Beddard) . . 19;j 

92. Tensores Patagii of Ccrylc ulcijon (after Beddani) . . , l<j<) 

93. Tensores PaLagii of Callalcijon nifn (after Beddard \ . ];i<( 
'.14. Tensores Patagii of Saurojudiy albicUla (after Beddard) . . 


Kill. I'VCR 

95. Skull of Col ins eaxfinioiiotus. Ventral Aspect (after Garrod) 203 

fl6. Tensoi'es Patfigii of Lcjitosont/ts (after Forbes) . . . 206 

97. Syrinx of Lcptosomus (after Forbes) . . . . . 207 

98. Foot of Todus (after Forbes) 213 

99. Foot of Momotiis (after Forbes) 213 

100. Patagial Muscles ot Bueorvus (after Beddard) . . . 216 

101. Leg Muscles of Aceros (after Bed/lard) 218 

102. Syrinx of Aceros nipalensis, Front View (after Beddard) . 220 

103. Syrinx of Bucorvits abijssinicas. Front View (after Beddard) 221 

104. Skull of Mieropus melanoleuctis. Under View (after Sltufeldl) 229 

105. Anconal Aspect of Left Humerus of Mieropus melanoleucus 

(after Shufeldt) . . .230 

106. Palmar Aspect of same Bone (after Shufeldt) . . . . 230 

107. Anconal Aspect of Left Humerus of Trochilus Alej-andri 

(after Shufeldt} 230 

108. Palmar Aspect of same Bone (after Shufeld':} . . . . 230 

109. Left Feet of Antrostomus vociferus and Nyctidromus albicollis 

(after ticlater) 231 

110. Right Foot of Podara/ts Cuvieri (after Sclater) . . . . 231 

111. Powder-down Patches of Podargus (after Sclater) . . . 232 

112. Pterylosis of Steatornis (after Gat-rod] . . . 233 

113. Syrinx of Nyctirlromiis albicollis (after Bcddar/1] . . . 235 

114. Syrinx of JEgotlielcs (after Beddard] . . . . . . 236 

115. Syrinx of BatracJiostonius (after Beddard) .... 236 

116. Fatagial Muscles of Ca/primulgus (after Garrod] . . . 237 

117. Corresponding Muscles of Steatornis, but of Left Wing (after 

Garrod) 237 

118. Skull of Caprimultjus (after Huxley] 238 

119. Fore Part of Skull of Nyctibns jamaicensls (after H n.rleij) . 239 

120. Skull of Steatornis (after Huxley) ... . . 239 
li'l. Skull of Podargus (after Huxley) . .... 240 

122. Sternum of Caprimulgus (after Sclater) 241 

123. Sternum of Podargus (after Sclater) 241 

124. Sternum of Nyctibius (after Sclater) . . . . . 241 

125. Colic Caeca of Plwtodilus (after Beddard) .... 247 

126. Skulls of Strix and Bubo (after Beddard) 24H 

127. Right Foot of Strix (after Beddard) 249 

128. Left Foot of Bubo (after Beddard) 250 

129. Syrinx, of Scops leucotis (after Beddard) 251 

130. Syrinx of Bubo (after Beddard) 251 

131. liyoid of Slritiynps (after Micart) ...... 2l>5 



132. Hyoid of Lorius flavopalliatus (after Mivart) . . . 266 

133. Hyoid of Lorins doiuicella (after Mivart) .... 267 

134. Pterylosis of Eudynamis oriental-is. Ventral View (after 

Beddard) ........... 273 

135. Pterylosis of Piaya cayana. Dorsal View (after Beddard) . 274 

136. Pterylosis of Piaya cayana. Ventral View (after Beddard) . 275 

137. Syrinx of Piaya cayana (after Beddard) ..... 277 

138. Syrinx of Centropus ateralbus (after Beddard) . . . . 278 

139. Intestines of Corythaix chlorochlamys (after Mitchell) . . 283 

140. Sternum of Opisthocomus. Side View (after Huxley] . . 287 

141. Sternum of Opisthocomus. Front View (after Huxley) . . 287 

142. Syrinx of Opisthocomus. Front View (after Gar rod) . . 289 

143. Syrinx of Pavo spicifcr. Front View (after Garrod) . . 294 

144. Syrinx of Same. Back View (after Garrod) . . . . 294 

145. Syrinx of Callipepla californica. Front View (after Garrod) 295 

146. Syrinx of Same. Back View (after Garrod) .... 295 

147. Syrinx of Male Tctrao tetrix. Front View (after Garrod) . 296 

148. Syrinx of Aburria carunculata. Front View (after Garrod) . 297 

149. Syrinx of Same. Back View (after Garrod) . . . . 297 

150. Syrinx of Megacephalon inaleo. A.. Front View. B. Back 

View (after Garrod) ........ 297 

151. Skull of Crax globicera. Side View (after Huxley) . . . 300 

152. Skull of Tetrao urogallus. Ventral View (after Huxley) . 301 

153. Sternum of Crax globicera, (after Huxley) ..... 301 

154. Horizontal Sections of Gizzards of, a, Ptilopusjambu, 6, Treron 

calva (after Garrod) . . . . . . . 306 

155. , Gizzard of Carpopliaga latrans. b, One of Horny Tubercles 

in Section (after Garrod) ........ 306 

156. Intestines of Colnmba livia (after Mitchell) .... 307 

157. Syrinx of Carpophaga latrans (after Garrod) . . . . 310 

158. Intestines of Cariama cristata (after Mitchell) . . . 323 

159. Intestines of Crex pratcnsis (after Mitchell) . . . . 323 

160. Deep Flexor Tendons of Hcliornis (after Beddard) . . . 326 

161. Patagial Muscles of Heliornis (after Beddard) . . . . 326 

162. Syrinx of Podica senegalensis (after Beddard) . . . 327 

163. Sternum of Heliornis. Ventral View (after Beddard) . . 327 

164. Lateral View of Vertebral Column, Pelvis, and Sternum of 

Podica senegalensis (after Beddard) ..... 328 

165. Skull of Podica. Lateral View (after Beddard) . . . . 329 

166. Skull of Heliornis. Ventral and Lateral Views (after Beddard) 329 

167. Skull of Podica. Ventral View (after Bedda rd) . . . 329 

168. Under View of Skull of Charadrius pluvialia (after Huxley) . 338 


Flc;. I'AGK 

169. Vomers of various Limicolce (after Garrod) . . . . 339 

170. Skull of Attagis Gcuji (after Gar rod) 349 

171. Tensores Patagii of Rliyncliops (after Beddard) . . . . 352 

172. Tensores Patagii of Larus argent at us (after Beddard from 

Forbes) 353 

173. Tensores Patagii of Litnda cirrliata (after Beddard from 

Forbes) 360 

174. The same of Syntliliborliamplius antiquus (after Beddard 

from Forbes) 360 

175. Tensores Patagii of Ceratorliina monoecrata, (after Beddard 

from Forbes) .......... 361 

176. Syrinx of Lumvia troilc (after Bedda rd) . .... 363 

177. Syrinx of Ceratorliina monocerata (after Beddard) . . . 363 

178. Certain Leg Muscles of Eliinoclietus (after Beddard) . . 370 

179. Deep Flexor Tendons of Eliinoclictus (after Beddard) . . 371 

180. Muscles of Fore Limb of Eliinoclietus (after Beddard) . . 372 

181. Syrinx of Eliinoclietus (after Beddard) ... . . 372 

182. Skull of Cliunga. Ventral View (after Beddard) . . .373 

183. Syrinx of Psopliia leucoptera (after Beddard) . . . . 375 

184. Sternum, Pelvis, &c., of Psophia leucoptera (after Beddard) . 376 

185. Skull of Phororliacos. lateral Aspect (after Andrews) . . 384 

186. Skull of Phororliacos. Dorsal Aspect (after Andrews) . . 385 

187. Pelvis of Phororhacos. Dorsal Aspect (after Andrews) . . 385 

188. Syrinx of jEclimoplwrus (after Beddard) .... 388 

189. Syrinx of Taeliybaptes (after Beddard) 388 

190. Origin of Biceps in Pelccanus and Phalacrocorax (after Filr- 

bringer) . 405 

191. Skull of Fregata. Ventral Aspect (after Beddard) . . . 410 

192. Skull of Phaeton (after Beddard) 411 

193. Stomach of Levaillant's Darter (after Garrod) . . . . 414 

194. Syrinx of Scopus (after Beddard) 421 

195. Deep Plantar Tendons of Scopus (after Beddard) . . . 421 

196. Deep Plantar Tendons of 8co%>us (after Beddard) . . . 421 

197. Syrinx of Leptoptilus (after Weldon) . . . . . . 422 

198. Syrinx of Dissura episcopus (after Beddard) .... 423 

199. Syrinx of Abdimia splienorliynclia (after Beddard) . . . 423 

200. Syrinx of Xenorhynchus senegalensis (after Beddard) . . 424 

201. Convoluted Windpipe of Tantalus ibis (after Garrod) . . 426 

202. Ventral Surface of Skull of Ardea cinerea (after Huxley) . 432 

203. Syrinx of Baltrniceps. Front View (a_ ftcr Beddard) . . . 433 

204. The same. Back View (after Beddard) . . 433 



205. Syrinx of Balaniceps, arranged to display Pessulus and Mcin- 

brana Tympaniforinis (after Beddard) . . ... 434 

206. Windpipe of Platalea, ajaja (after Garrod) .... 430 

207. Intestines of Platalea leucorodia (after Mitchell) . . . 437 

208. Intestines of Ciconia nigra (after Mitchell) .... 437 

209. Syrinx of Phcenicopterus (after Weldori) 440 

210. Intestines of Fulmarus glacial-is (after Mitchell) . . . 446 

211. Skull of Diomcdea exulans (after Huxley) . .... 449 

212. Skull of Procellaria gigantea (after Huxley) .... 449 

213. Caeca of Chauna chavaria (after Beddard) ..... 4f>3 

214. Windpipe of Palamedea '(after Beddard and Mitchell) . . 454 

215. Skull of Chauna derbiana. Ventral Aspect (after Garrod) . 455 
21(5. Mouth of Biziura lobata (after Forbes) . . . . . 458 

217. Biceps of Femoris of Duck (Bi), to show its Relations to Gas- 

trocneinius (after Weldon) ....... 400 

218. Windpipe of Sarcidiornis melanota <? (after Garrod) . . 461 

219. Same of S. inclanota ? (after Garrod) ..... 461 

220. Same of Rhodonessa caruorjhyllacea ? (after Garrod) . . 46.1 

221. Syrinx of R. carijophyllacea $ (after Garrod) . . . 462 

222. Windpipe of Metopiana peposaca $ (after Garrod) . . . 463 

223. Syrinx of Biziura S (after Forbes) 464 

224. Skull of Qiu-rqucdula crecca. Lateral View (after Huxley) . 467 

225. Ventral View of Same 467 

226. Tensores Patagii of Polyboroidcs (after Beddard) . . . 472 

227. Tensores Patagii of Serpentarius (after Beddard) . . . 480 

228. Skull of Serpentarius (after Huxley) 481 

229. Windpipe of Condor (after Beddard) 483 

230. Skull of Cathartes aura (after Huxleij) 483 

231. Cseca of Calodromas elegans (after Beddard) . . . . 488 

232. Cseca of Nothura macidosa (after Beddard) .... 488 

233. Skull of Tinau<srobust'us (after Huxley) 490 

234. Tongue and Windpipe of Rhea Darwini (after Gadow) . . 505 

235. Syrinx of Aptcryx Mantelli. Front View (after Forbes) . . 507 

236. The same, from behind .... . . 507 

237. Syrinx of Rhca americana. Front View (after For bes) . . 508 

238. The same, from behind ... .... 508 

239. Sj-rinx of Casuarius galeatus. Front View (after Forbes) . . 509 

240. The same, from behind 509 

241. Tracheal Pouch of Emu cut open (after Mitrie) . . . . 510 

242. Sternum of Rhea (a fter Mivart) 512 

243. Pelvis of Rhca (after Mivart) . . . . 513 


244. Skull of Emu (after Huxley) ....... 514 

245. Pelvis of Emu (after Mivart) ....... 515 

'24(5. Sternum of Cassowary (after Mivart) ... . 516 

'247. Pelvis of Cassowary (after Mivart) ...... 517 

248. Skull of Ostrich (after Huxley) ...... 518 

249. Sternum of Ostrich (after Mivart) ...... 519 

250. Pelvis of Ostrich (after Mivart) ...... 519 

251. Shoulder Girdle of ^Epyornis (after Andrews) . ... 523 

252. Diagram of Eelationsliips of Strutkiones (after T. J. Parker) 527 






As the aim of the present work is to detail the characteristics 
of the various groups of birds, I do not propose in this 
section to do more than give a general account of bird 
anatomy. A fuller description will be found in the part 
relating to birds by Dr. GADOW, in Bronn's ' Klassen und 
Ordnungen des Thierreichs,' where much that will be found 
here in the accounts of the several families of birds is 
treated of in the introductory chapters. The greater part of 
this I deliberately omit, to save repetition. 

The Foot 

The feet of birds show a large amount of variation, 
which is not for the most part of great value in the deter- 
mination of affinities. That the older naturalists paid great 
attention to these facts is evident from the names Palmipedes, 
Cursores, &c. No bird has, save for abnormalities, such as 
the Dorking fowl, more than four toes. The opposite ex- 
treme is reached by the ostrich, which has only two. That 
the three-toed and, a fortiori, the two-toed condition has 
been arrived at by a reduction from four toes seems to be 
shown by the condition of the feet in certain petrels, where, 



as shown by FORBES, a rudiment of the missing toe is 
present in the shape of a nodule, or of two nodules, of bone, 
hidden in extreme cases beneath the skin, and only appear- 
ing externally as a small wart with no claw. In Phebcetria 
there is an advance upon this, since there is externally a 
minute claw, and beneath the skin two minute nodules of 
bone. In the woodpeckers, Picoides and Tiga, commonly 
spoken of as three-toed birds, there is a similar vestige of 
the fourth toe. On the other hand this does not apply to 
all three-toed birds. In Rkea, Tetrax, and Pelecanoides 
FORBES searched in vain for a trace of the missing toe. The 
toes of the bird's foot are arranged in different fashions, 
giving rise to more than one form of foot. When there are 
only three toes they are all directed forwards, except in 
Picoides, where the last is directed backwards ; but then, as 
already stated, this bird has a rudimentary hallux, and it 
conforms therefore to the type seen in other woodpeckers. 
AVhen there are four toes they are rarely all turned forwards ; 
this is the case, however,' with the swifts. Most commonly 
the hallux is turned more or less completely backwards ; 
this is so with the passerines and with many other birds. 
In what is termed the zygodactyle foot, e.g. the woodpeckers, 
both the first and the fourth toes are turned backwards, and 
thus an effective grasping organ is produced. 

An anomalous form of zygodactylism, termed hetero- 
dactylism by some, is offered by the trogons, where the 
second toe is turned back. Syndactylism is an expression 
used to describe toes which are united together for a longer 
or shorter distance, such as, for example, the todies and 
kingfishers. Further details in the form of foot will be 
found under the descriptions of the several families. 

In many birds the toes are perfectly free from each other 
up to their attachment to the metatarsals. In others there 
is a condition known as webbing, where a scale-covered 
skin is stretched between the toes. This may be feebly 
developed, as in many wading birds, or complete, as in swim- 
ming birds, such as the duck. The extreme state of webbing 
is seen in the pelican tribe, where all the four toes are 


united by webs in the ducks only three being thus united, 
and the hallux free. The coots have a form of webbing 
which is characterised by applying the term ' lobate ' to the 
foot. Each toe in the case of a lobate foot is bordered by a 
flat expansion of skin, but there is no connection between 
the borders of adjacent toes. The foot is covered to a vary- 
ing degree with a horny integument, which is arranged as 
larger or smaller flat scales, or as granules of various shapes 
and sizes. This scutellation sometimes extends on to the 
tibia ; it generally occupies the tarsus as well as the foot 
proper ; the other extreme is shown by Syrrhaptes, where 
feathering extends down to the last digits of the foot. The 
form of these is sometimes useful in classification, as in the 
Passeres (q.v.) The digits are armed with claws, which are 
straighter in wading birds, and very curved in the birds of 
prey. Their relative lengths vary ; in the larks, for example, 
and in the cuckoo (Ccntropiis) that of the hind toe is enor- 
mously long. The middle toe is often serrated, as in herons, 
owls, and goatsuckers, &c. 

The order of the toes may be almost always settled by 
counting the number of the phalanges. This is progressive, 
the first toe having two phalanges, the second three, the 
third four, and the fourth five. To this general rule there 
are a few exceptions. I have described in the owl (Plwto- 
dilux} only four phalanges in the last (fourth toe), a state of 
affairs, however, which is plainly due to a fusion between 
the two first phalanges. The goatsuckers (of genus Capri- 
uinhiiis and allied forms) have a digital formula of 2, 3, 4, 4. 
In many Tubinares the formula is 1, 3, 4, 5. In the swifts 
the toes are still further reduced, for we have in that group 
the digital formula 2, 3, 3, 3. Pterocles has a digital formula 
agreeing with that of Caprinuth/iis. ZEHNTNER ' has lately 
shown 'that the digital formula of the swifts is due to 
reduction ; he has found, in fact, in Cypselus niclba, in a 
certain stage of development, four phalanges in each of 
digits 3 and 4, in which stage, therefore, the bird is only 
one digit short of the normal. 

1 'Beitriige zur Entwicklung von Cypseius melba,' Arch.f. Natury. 1890. 

B 2 


A comparison of the varied forms of feet among birds 
has an important bearing upon the origin of the foot. From 
this follows some insight into the nature of the life led by 
the possessors of the most primitive form of foot. The 
matter has been put forward in a clear fashion by FiNN, 1 
and in a correspondence which his paper elicited. As all the 
evidence at our disposal seems to show that the four-toed 
foot is the more primitive, we have to decide whether the 
palmate foot of the pelicans or the grasping foot of the 
passerine is the earlier, or whether some modification of 
these, such as the zygodactyle, or four-toed foot with a 
rudimentary hallux, is the more primitive. The four-toed 
foot of the Steganopodes is often figured as if all the toes were 
directed forwards, but this is really not the case ; they are, 
as in terrestrial birds with a more or less rudimentary 
hallux, directed at least sideways. This seems to argue that 
the original form of the foot was as it is now in the Passeres, 
a fact which is still further enforced by the foot of Arch&o- 
pteryx (see below) . The more purely terrestrial the birds are 
the more rudimentary is the hallux, until in the purely 
terrestrial bustards the hallux has disappeared altogether, 
and in the ostrich, most terrestrial of birds, the second toe 
has vanished also. Among the gallinaceous birds, moreover, 
the more arboreal forms, the Cracidas and Megapodida?, the 
hallux is better developed than in those that do not roost in 
trees. That the zygodactyle foot is a further modification of 
the anisodactyle seems to be shown by the transitional state 
of the owls, which ' always perch in the zygodactyle position,' 
the fourth toe being capable of reversion. - 


In all existing birds the upper and lower mandibles are 
invested with a horny sheath, the beak. The form of this 

1 ' The Significance of the Bird's Foot,' Natural Science, June 1894 ; see 
also July and September Nos. of the same periodical for further notes and 
correspondence on the matter. 

- A. REICHEXOW, ' Die Fussbildungen der Vogel,' J- f- O. 1871, p. 401. 


beak has been used by ornithologists for systematic purposes, 
and whole groups of birds have received their names from 
this shape, e.g. dentirostres, lamellirostres, &c. The bill, 
however, varies so greatly in admittedly allied birds that its 
use for classificatory purposes is not great. As a striking 
instance of this may be mentioned the Limicolae ; we see 
there the spatulate bill of Euonyrhynchus, like a diminutive 
spoonbill, the upturned bill of Recurvirostra, the sideways- 
turned bill of Anarhrynchus, the longer lower mandible of 
Rhynchops, and the ibis-like bill of Numenius. GADOW has 
used for classificatory purposes the complex or simple con- 
dition of the beak. In some birds, e.g. liatitae, the horny 
she'ath is composed of several pieces ; in others, the majority, 
this is not the case. In birds of prey and in parrots there 
is present a structure which has been termed the cere : this 
is simply the basal part of the beak, which has remained soft. 
Its occurrence in those two groups of birds does not appear 
to be significant of any close affinity. 

The lamellirostres afford another example of how dan- 
gerous it is to attempt any decision as to affinities from the 
form of this organ. It has been insisted that one reason for 
regarding the flamingo as a long-legged duck is the existence 
of lamellae along the beak ; but this feature is also met with 
in the stork, Anastomiis, to which group moreover the bird 
is now more generally believed to be related. The puffin is 
nearly exceptional 1 in the periodical moulting of a portion of 
the bill; but the pelican (P. trachyrhynchus) casts annually 
an excrescence upon the top of the upper beak. Sexual 
dimorphism in the bill is rare, but is exhibited in a marked 
way in HeteralocJia, where the female has a long and down- 
wardly curved bill, while that of the male is shorter and 
straighter. 2 

1 Several auks (g.r.) do so, and it has been asserted of the penguins. 
See in this matter and for variations EIILKKS, Zool. Miscell. i. (Gottingen, 


Feathers l 

A. bird may be known by its feathers ; to define a bird it 
is only necessary to refer to its covering of feathers. No 
other animal has any structures comparable to a well- 
developed feather. It is true that the filo-plumes are really 
little more than hairs. But the processes of development 
serve to place a fundamental barrier between the two kinds 
of structures. 

A hair commences as a thickening of the stratum 
Malpighii, which grows downwards into the dermis ; a feather 
is from the first a slight papilla involving the outer layers of 
the epidermis as well as the stratum Malpighii, a papilla 
which is surrounded by a circular depression. This papilla 
gradually sinks down into the skin and assumes a cylindrical 
form. The cells of the Malpighian layer commence to pro- 
liferate vigorously, and form a series of thickened folds 
disposed radially to the longitudinal axis of the feather 
papilla, and towards the central pulpa. These radially 
arranged masses of cells undergo a process of cornification, 
free themselves from the overlying cells of the horny layer of 
the epidermis, and produce a bundle of horny fibres the 
embryonic down. The feathers may retain this embryonic 
character throughout life, or further changes may take place. 
This consists in the formation below the first feather follicle 
of a second in continuity with it ; in this a feather is 
developed, which may be a down feather, like the first formed, 
or may grow into one of the stronger varieties of feathers to 
be described presently. In either case the growing feather 
pushes the down before it, and the latter is ultimately thrown 

The structure of feathers has been described at length by 

1 H. E. DAVIES, ' Beitrag zur Entwicklungsgeschichte der Feeler,' Morph. J.B. 
xiv. 1888, p. 368, and 'Die Entwicklungsgesch. d. Feder,' etc. ibid. xv. Issn, 
p. 5GO ; C. R. HEXXICKE, ' Die Entwieklung d. Feder,' MonatsscJn: dcntscli. Ver. 
Vogclscli. xiv. ISN'.I, p. 223 ; K. KLKK, ' Ban und Enlwicklung der Feder,' 
Zeitsclir. f. d. gcs. Naturw. lix. p. 110 ; see also GADOW, article ' Feather' in 
NEWTON'S Diet, of Birds. 


NITZSCH, and among recent writers more especially by 
WE AY. 1 A typical feather consists of the stern or rhachis, 
of which the lower ' quill ' region is termed the calamus. 
From the rhachis above the calamus spring a series of lateral 
branches, the rarni or barbs, which in turn give rise to 
barbules, and they to minute, often hooked, processes, the 
barbicels. At the junction of the calamus with the barb- 
beariiig rhachis arises in many feathers an aftershaft (fig. 
4), which has the character of a second smaller feather 
arising from the shaft of the first ; but in the cassowary, 
emu, and the extinct Dinornis this aftershaft is as large as 
the main feather from which it arises. The barbicels with 
their terminal hamuli give the stiffness to the feather which 
is caused by the interlocking of these processes. The bar- 
bules are of two sorts, those nearest to the root of the barb 
being different from those nearest to its tip. The former 
are shaped something like a knife blade ; they are thickened 
above and bent in the middle, gradually tapering away to 
a fine point ; just in the middle, where the bend is, are 
two or three small teeth on the upper margin. It is by 
means of these teeth that successive barbules are locked 
together. The remaining set of barbules are frayed out 
towards the end into a series of branchlets which are 
hooked at first, but the more distal set are merely fine- 
pointed branchlets ; these arise obliquely, so that a given 
barbule comes into relation with four or five other barbules. 

All feathers, however, have not so complicated a structure. 
The strong wing feathers of the cassowary consist of the 
stem alone. Filoplumes have but few radii, consisting 
almost alone of the calamus and rhachis; 

Down feathers are as a rule without the hamuli ; often 
the radii spring at once from the calamus, there being 
no rhachis. 2 A peculiar form of these feathers, called 

1 ' On the Structure of the Barbs, Barbules, and Barbicels of a Typical 
Pennaceous Feather,' Ibis, 1887, p. 420. 

- The term neossoptiles has been applied to the down covering the newly 
hatched young of many birds, in contradistinction to telcoptiles, the feathers 
(down or contour) of the adult bird. 


' powder down feathers,' } is found in many birds belonging 
to quite different groups ; they are usually aggregated into 
special patches. These are simply down feathers of which 
the tops continually break down into a dusty matter. 
These powder down patches have been asserted to be lumi- 
nous in the heron, and to aid it in attracting its prey ; but 
the assertion seems to be void of truth. 

The feathers of birds are, with a few exceptions, coloured 
either by the deposition of pigments alone, or by optical 
tints derived from the actual structure of the feathers shown 
up against a basis of dark pigment. The colours of birds' 
feathers have been chiefly investigated by CHUECH, KKTJKEN- 
BERG, and GADOW, 2 to whose papers the reader is referred. 

The arrangement of the feathers upon the wing requires 
a special description. They have been carefully studied by 
the late Mr. WRAY, S from whose paper both the informa- 
tion and some of the explanatory illustrations have been 
drawn. In the wing of the wild duck there is, as in all 
birds, a fringe of stout quills known as the remiges. These 
are attached to the fore-arm and to the hand. The border 
of the ulna, to which they are fixed, constantly bears impres- 
sions of the quills. Here the feathers stand out at right 
angles to the bone ; in the hand they become more and 
more inclined forwards until the last of the series lies parallel 
with the bone (phalanx 2 of digit II.) which bears it. Of 
these remiges it is usual to term those which are inserted 
upon the ulna the secondaries, and those upon the hand 
proper primaries. But the term cubitals is gaining ground 
as an expression for the secondaries of many writers. The 
first of the remiges is much smaller than the others, and 
has been called the remide ; it nevertheless belongs to the 
series of remiges. The rest of the feathers of the wing are 
known as the coverts or tectrices. There are four series of 

1 L. STIEDA, ' Uber den Bau cler Puderdunen der Rohrdrornmel,' Arch. f. 
Anat. u. Phys. 1870, p. 104. 

- GADOW, in Bronn's Thicrreich ('Aves '), treats of the matter in considerable 

3 ' On some Points in the Morphology of the Wing of Birds,' P. Z. S. 1887 
p. 343. 


these, which successively overlap each other and the remiges. 
The first series on the upper aspect of the wing are the 
tect rices majores, which have a perfectly definite relation to 
the remiges, there being one for each remex. To this state- 
ment there is in the duck a single exception ; this exception 
is the fifth cubital (reckoning, as it is customary to do, from 
the carpus) ; this remex appears, by reason of the fact that 
there is a gap and that the tectrix is present, to be absent. 
On the under surface of the wing there is a corresponding 
row of lower tectrices majores. It will be noticed that the 
reference of the remicle to the series remiges is justified by 
its having its proper complement of tectrices majores. 

Next to the tectrices majores comes a row of feathers, 
the tectrices medice. These are also present on the under 
surface ; the set of both, however, is not complete, that of 
the second metacarpal being wanting on the upper surface, 
and the distal four or five of the manus on the lower surface. 
The next row on the upper face of the wing is quintuple, 
and the feathers composing the five tiers are known as 
the tectrices minor es. They are scantily represented on 
the manus, where in fact there is not room for them, they 
being developed on the skin covering the muscles and on 
the patagium of the wing. This row of feathers passes on 
to the humerus and becomes there partly specialised into 
two rows ; the lower of these (sometimes called parapteron) 
are long feathers suggestive of remiges, while the row im- 
mediately above bears the same relation to the pseudo- 
remiges as the tectrices majores do to the true remiges. On 
the ventral surface of the wing are similar tectrices minores 
with a similar specialisation of an hypopteron (representing 
the parapteron above, and sometimes called axillaries), with 
its row of special coverts. The patagium is mainly filled 
up with several rows of feathers, which are collectively 
termed the marginals ; anteriorly, upon the pollex, they 
form together with the anterior feathers of the minores the 
so-called ala spuria. The ala spuria is specialised into four 
small quills with coverts, the specialisation being quite like 
that of the numerals at the other extremity of the wing. 



So much then for the arrangement of the feathers in the 
typical bird selected ; we must now consider the divergencies 
from this constituted normal. The fifth cubital, absent in 


1-1', posterior border ; 2-2', anterior border ; .1, remiges ; fi, greater whm coverts : ' '. L>. 
median coverts: E, /', rrmuiniiiir euverts. 

the wild duck, is often present in birds. The terms ' quin- 
cubital ' and ' aquincubital ' have been devised to express 
these facts. The following groups are quincubital : ' 

Crypturi, Galli, Rhinochetidae, Cuculi, many Picarians. 
On the other hand aquincubital birds are 

ColymbidaB, Tubinares, Steganopodes, Herodiones, Acci- 
pitres, Anseres, &c. 

The majority of birds, in fact, have not the fifth cubital 
remex. The most remarkable fact about this missing 
rernex is that it is either absent or present ; in no case are 
there any intermediate conditions, such as a small remex. 

See SCLATEB, 'Remarks on the Fifth Cubital Remex,' &c.. Ibis (6), ii. 1890, 

p. 77. 



The only explanation, so far as I am aware, of this 
remarkable state of affairs is contained in a suggestive paper 
by DEGEX.' DEGEN commences with the assumption that in 
the hands of the primitive bird all three fingers then freely 
movable were furnished with remiges. In modern birds 
remiges are only attached to the thumb (ala spuria) and to 
digit II. DEGEN also postulates a fourth finger (of which 
rudiments have been discovered in modern birds ; see below) 
with its remiges. 

When the metacarpal bones became fused the feathers 
of the third and fourth digits were, he supposed, forced back 


li 1-7, remiges : />.<', <lovsal tectris major : J7, ulim. 

upon the ulna, as there was no longer any room for their 
coexistence with those of the second digit upon the com- 
pressed hand. Among these the carpal remex was also 
pushed back. As this remex was attached to an unstable 
bone or cartilage, its position was not secured, and the varia- 
bility remained when the feather altered its position ; hence 
the presence or absence of the fifth remex, which is this 

The carpal remex is another variable feather. It is present 

1 ' On some of the Main Features in the Evolution of the Bird's Wing,' Bull. 
Brit. Orn. Club, July 1894 (published in I&is). See also for quincubitalism 
GEKBE, ' Sur les Plumes de Vol et leur Mue,' Bull. Soc. Zool. Fr. ii. Iw77, 
p. 289. 


and fully-sized in Nothura. It is occasionally present but 
small, and sometimes even with its covert altogether absent. 
The remiges themselves vary in number apart from the 
presence or absence of the fifth cubital, but not within 
very wide limits. Struthio and the penguin alone are 
exceptional, and will be treated of separately and later. 
GADOW has published a useful table showing the number of 
the primaries in a very large assortment of birds belonging 
to all orders. The number of primaries varies only between 
ten and twelve. The number of metacarpals has also a 
small range of variation, the smallest number presenting 
six and the largest eight. Casuarius having an abbreviated 
hand is still further reduced, the primaries being only two 
and the secondaries five. The largest number of metacarpals, 
eight, is possessed, however, by Apteryx, with an abbreviated 
hand, and by Struthio. Seven metacarpals are found in the 
grebes, flamingoes, and several, but not all, the genera of 
storks. All other birds have six. 

The two prominent exceptions to the foregoing state- 
ments are, as has been already mentioned, Struthio and the 
Spheniscidse. In the ostrich (see fig. 3) there are sixteen 
primaries, each with its tectrix major upon the upper surface 
of the wing. The other rows are perfectly recognisable, as is 
shown in the figure. The wing of the penguin is, however, 
not reconcilable with the ordinary plan of structure. It has 
thirty-six bordering feathers, which may be termed primaries ; 
Ft'iRBRiNGEE has suggested that these may be really ten 
primaries with their coverts, but in any case the wing is 
covered with about thirty rows of scale-like feathers. 

As to the general wing feathering, GOODCHILD ' has sur- 
veyed a large series of birds, and noted their peculiarities. 
Some valuable classificatory results appear to be the outcome 
of these investigations. Thus the plan characteristic of the 
humming birds resembles that of the swifts, and both are to 
be distinguished from the passerines. The picarian type 
gradually approximates to the psittacine ; Melopsittacus 

1 ' Observations on the Disposition of the Cubital Coverts in Birds,' P. Z. S. 
1886, p. isl. 



might be well referred to the picarians when judged from 
the present standpoint. On the other side the birds of prey, 
both diurnal and nocturnal, are parrot-like in the arrange- 

uient of their wing feathers. But, curiously enough, Pernis, 
Pandion, Gijpogeranus, and the Cathartidae differ from theii- 
allies. This is not the only case where the disposition 
of the feathers runs counter to the affinities to be 


derived from an examination of other structures ; for 
while Plialacrocorax is quite accipitrine the other Stega- 
nopodes are quite different. Herons agree with the 
Accipitres, while the ciconiine pattern leads towards that 
of the Tubinares, and is identical in some cases with that of 
the American vultures. The cuckoos should be, when 
judged by the feathering of their wings, placed in the 
immediate neighbourhood of the Cohimbae, from which group 
Goura ought to be separated. Ghaunq. is practically a 
pigeon in these characters, while the Limicolse are not far 
off. The Crypturi are gallinaceous. 1 


As a general rule the feathers of birds are not distributed 
uniformly over the surface of the body, but are set in the 
skin in definite tracts, between which are spaces that are 
entirely bare or covered only with down plumage. The 
feathered tracts are termed pterylse, the interspaces apteria. 
A few birds, such as the struthious, the penguins, and the 
screamers, have an uninterrupted plumage ; but this state 
of affairs, though corresponding with what one supposes 
to be the original condition, is not necessarily so in the 
birds under consideration. Thus, although the ostrich has 
an uninterrupted plumage in the adult state, the young 
embryo, as first figured by Miss Lindsay, has definite 
pterylae, thus proving that the continuous feathering is here 
purely secondary. There is a very great variety in the 
arrangement of the pterylse among birds, and for the details 

1 DE MEIJERE has devoted some pains to the arrangement of the feathers 
with reference to each other, a subject which, as he says, has been hitherto 
treated of only in a stepmotherly fashion. It appears from his investigations 
that the feathers are arranged in groups, as are the hairs of mammals. For 
example, upon the naked region of the head of Numida the feathers are 
grouped in fours, a stronger feather with two hair-like feathers, one on one side 
and one on the other. This is what is generally found, a central stronger 
feather with hair feathers surrounding it. There is here a remarkable analogy 
with the grouping of mammalian hairs, where a stronger hair is often sur- 
rounded by three or four more slender hairs. MOSELEY also in the case of the 
dodo (q.v.) has found the feathers to be grouped in threes. 


the reader is referred to the systematic part of this work. 
In all, however, there are the following tracts present : 

(1) The spinal tract runs from the head to the oil gland. 
This tract is sometimes continuous at the sides of the neck 
with the ventral tract, to be described next. It is sometimes 
a single solid tract throughout, but more usually there is a 
space developed in it, an apterion in the back, which is of 
greater or less extent. Commonly there is a break, more or 
less distinct, between the anterior and the posterior portion 
of the tract, which may be complete or may consist in an 
abrupt transition between anterior stiffer feathers and 
posterior slighter feathers. 

(2) The ventral tract is always a double tract, but the 
median apterium may be very narrow. The anterior part 
of the tract may be single, but sometimes it is double from 
its very origin. Very commonly on the pectoral region 
each half of the ventral tract gives off a lateral branch. 

(3) The humeral tract is a band of stiff feathers running 
across the humerus ; it is always present and shows no parti- 
cular modifications. 

(4) The femoral tract is a corresponding band crossing 
the thigh. There is sometimes, as in the barbets, a small 
tract lying between the femoral and the spinal ; and, besides 
the main tracts, the patagium and the lower leg are more or 
less covered with contour feathers. The study of Pterylo- 
graphy was first taken up in a systematic manner by NITZSCH ; 
since his day the number of facts has largely increased, and 
careful figures of the pterylosis of many birds, not figured 
by NITZSCH, have been published by a host of observers, 
and others. For references to these see the descriptions of 
the different orders of birds. 2 

The general facts of pterylosis must be used cautiously. 

1 See also W. MARSHALL, ' Pterologische Mittheilungen,' Zool. Gart. xiv. xv. 

- For the musculature of the feathers see HELM, ' Ueber die Hautmuskeln 
cler Vogel,' Ac., J.f. 0. 1884, p. 321. These skin muscles are either limited to 
the skin, running from feather to feather, or are parts of skeletal muscles, such 
as the cutaneous branch of the latissiimts dorsi, &c. 


On theoretical grounds it might fairly be assumed that a 
continuous covering, without any distinctions between 
pterylas and apteria, was a primitive condition. But there 
is evidence to show that where a continuous covering of 
feathers exists it is not invariably a mark of ancient stock. 
Thus the ostrich, as already remarked, which when adult 
has practically no separation into pterylae and apteria, has, 
when young, very distinct pterylae and apteria. In this 
case, therefore, it is clear that the uninterrupted feathering 
is a secondary character and not a primitive one. With the 
penguins, on the other hand, it is conceivable that the 
absence of apteria is a primitive character. As to the value of 
the various arrangements of apteria in the pterylosis, GADOW 
lays stress upon the continuous feathering of the neck, or 
the presence there of lateral spaces, but admits that the 
Indian painted snipe (Bhynchced) is an exception which 
somewhat destroys the value derivable from the considera- 
tion of the facts. FUBBKINGEB, on the other hand, uses in 
his tables of characters the dorsal tract and its modifications. 
But the variations which occur here in a single and surely 
well-marked family (e.g. Picidse) tend to shake our faith in 
the value of the exact way in which the hypothetically con- 
tinuous feathering has lost its continuity. No doubt GADOW 
is right in saying that it is of taxonomic importance ' more 
in the investigation of small than of large groups.' 

NITZSCH, for instance, lays some stress upon the ' furcate 
division and degradation of the portion of the spinal tract 
situated between the shoulder blades ' in the Accipitrinse ; 
this division includes the owls. 

But on turning over his plates one is struck by the fact 
that the peculiarity in question is by no means confined to 
that group, occurring as it does in such widely removed 
forms as Caprimulgidae, Charadriidse, and Psopliia. Nor is 
an undivided dorsal tract a distinctive mark of affinity, 
since it is to be found in such a diversified assemblage as 
that including Pavo, Alcedo, CertJiia, Todus, and various 
passerines. The ventral tracts divide each of them upon 
the breast into an outer and an inner division in Pernis 


apivora, Coracias garrulus, JRhamphastos, Musophaga, Gal- 
ius, various cuckoos, Charadrius, kc. They do not divide 
in Pandion, Cypselus, Cncuhix, Opisthocomus, Buceros, Co- 
linuba, Fulica, and Ciconia. 

GARROD made originally the apparently reasonable sug- 
gestion that the down feathers upon the apteria of many 
birds may be the remains of contour feathers, from which 
the inference is necessary that those birds with downs upon 
the apteria are nearer to the continuously feathered and 
ancestral bird than are those whose apteria are nude. But 
the whole matter is rather more complicated than this. 
There are birds with only contour feathers and nude apteria ; 
there are birds with contour feathers only upon the pterylse, 
and down upon the apteria ; there are birds with dow r ns 
everywhere ; and finally there are birds with downs only 
upon the pterylae, mixed with the contour feathers. The 
facts, therefore, when stated thus fully are not so easy of 

The evidence derivable from Archaopteryx less negative, 
perhaps, than ' negative ' evidence often is may afford us a 
clue. So many feathers of that bird are well preserved thai 
it seems possible that where feathers have not been preserved 
they were either really absent or soft down feathers. The 
latter suggestion seems to be the more probable, on account 
of the plain fact that Archaopteryx was a flying bird. Now 
the fact that the contour feathers are frequently preceded 
by downs points in the same direction, viz. that the primitive 
feathering of birds was in the form of downs. The persist- 
ence of downs, therefore, on this hypothesis is so far a 
primitive character, and the greater the persistence the more 
primitive the bird. Thus those birds which have downs 
everywhere will be the more archaic. This is so far promis- 
ing that that group contains such apparently old tj^pes as 
Palamedea, Opistliocomus, Rhinochetus, Ovrc. On this view 
the most modern of birds will be those which I elsewhere 
try to show are an ancient race, i.e. the bulk of the Pico- 
Passeres. But, as might be expected with an .ancient race, 
there is every variety shown, and members of this great 




group are found in all the divisions of birds founded upon 
the distribution of the downs. This view throws a side 
light upon the Struthiones. The feathers of those birds have 
been called intermediate between contour feathers and 
downs. It may be that they are primitive, and that the 
struthious birds have arisen from some ancient type in 
which the modern bird's feather had hardly been evolved- 
Among nearly related families the details of pterylosis do at 
least sometimes afford indications of resemblance. Thus, 

for instance, there are certain small 
likenesses between the barbets, 
toucans, and woodpeckers (see 
below), which help in establishing 
the near kinship between the three 

The size or the presence or 
absence of the after shaft appears to 
be of little use for systematic pur- 
poses. Among the ducks, for ex- 
ample, some have it and some have 
it not. It is as large as the main 
feather in the emus and totally 
absent in Rhea. Facts like these, 
which might be multiplied, throw 
doubts upon the value of this 
structure in classification. So too 
with the oil gland ' and its feather- 
ing or absence of a tuft. Cancroma, 
which in other points of its structure conforms to the heron 
type, is alone in that group in having a nude oil gland. The 
gland is absent in some parrots, present in others. GAEROD 
at one time thought that he could correlate among the 
Pico-Passeres a nude oil gland with small caeca, and a tufted 
oil gland with the absence of ca3ca ; to the vast majority 
of picarian birds there is no doubt that the correlation does 

1 A. PILLIET, ' Sur la Glande Sebacee des Oiseaux,' &c., Bull. Soc. Zool. Fr. 
xiv. 1889, p. 115 E. KOSSMANX, ' Ueber Talgdriisen der Viigel,' ZeitscJn: f. 
iriss. Zool. 1871, p. 5lis. 




apply. But the todies were found to be birds with a tufted 
oil gland and with large caeca. 

It has been pointed out that when the oil gland has a 
tuft of feathers upon its apex the rest of the gland is un- 
feathered, and that, on the contrary, when the tip is nude 
the general surface of the gland is feathered. The oil gland 
is, so far as we know, a structure special to birds ; it is, indeed, 
the only purely external glandular apparatus that exists in 
them. It is therefore possible, if not probable, that the 
organ first arose in the class that it is not an inheritance 
from any ancestor. On this view it is quite possible that 
the absence of the oil gland may not be always due to its 
disappearance ; birds without oil glands may or may not 
have lost them. It seems very likely, for example, that the 
usual absence of this structure among the struthious birds 
is rather a primitive than a secondary character. If this 
view of the matter is justifiable, the presence of a tuft may 
also be, in some cases at least, secondary ; for it is certainly 
a specialisation that may have appeared after the oil gland 
was fully developed. 

Alimentary Canal 

The tongue ' of birds is one of the most variable organs 
as to size and texture. In Plotus, for example, it is 
practically altogether absent (see fig. 5). When present it 
is larger or smaller, more or less fleshy. The long, thin, 
horny tongue of the toucans characterises those birds ; the 
parrots (see fig. 6) have a thick and fleshy tongue ; naturally 
the organ is of more use to the latter than to the former. 
A very remarkable modification of the tongue, seen in birds 
quite remote in the scale, is the pulling out of the free end 
into a tuft of fine fibres ; this is associated with the capture 
of honey or minute insects from the corollas of flowers ; it is 

1 C. S. MINOT, 'Studies on the Tongue of Reptiles and Birds,' Ann. Man. 
Boston Soc. Nat. Hist. 1880 ; NITZSCH-GIEBEL, ' Die Zunge der Vogel,' &c., 
Zeitschr. f. d. cjes. Naturw. xi. 1858, p. .19; Lumvio, PRINZ v. BAYERN, Zur 
Anatomic der Zunge (Mitnchen, 1884). 

c 2 


seen in the Trichoglossinge (see fig. 6), named so on account 
of the very structure, and in the Nectariniidse, &c. Very 
frequently the tongue is more or less spiny upon its surface, 

particularly towards the attached end 
of the organ. A very singular modi- 
fication is the extraordinarily long 
tongue of the woodpeckers, w r hich 
is, of course, associated with the ex- 
traction of grubs from the crevices in 
trees. A detailed description of the 
numerous forms of this organ would 
occupy more space than can be allow r ed ; 
but the principal varieties will be 
found described under the different 
families. The modifications of the 
tongue are not of great assistance to 


7? Trwr-r-r TV P^r FlG - 6. HEAD OF LoriUS, SHOWING 

-D, J.ONGUE IN PROFILE. m / r< .-mmn\ 


the taxonomist, except as regards smaller groups. Thus 
the Plataleidae have been distinguished from other Herodiones 
as ' Lipoglossae.' 

Teeth are not met with in living birds. KOSE, however, 
has discovered what he believes to be a rudimentary tooth 
band (' Zahnleiste ') in Sterna, 1 a discovery which may have 

This has been recently confirmed by Miss CARLSSON (' Ueber die Schmelz- 
leiste bei Sterna hirundo,' Anat. Anz xii. 72), who found it to characterise both 
jaws. For other embryonic traces of teeth see P. FHAISSE, ' Uber Ziihne u. Zahn- 


some added significance in view of the possible relationship 
to modern birds of the cretaceous Iclithyornis (cf. below). 
The need for teeth seems to have disappeared with the 
development of a horny bill, the replacement of the one 
structure by the other being, perhaps, comparable to the 
replacement of functional teeth by horny plates in the 
Ornitliorliyncluts. Functional teeth, however, existed in 
the Jurassic Archaopteryx and Laopteryx (?), and in the 
toothed birds of the cretaceous epoch, Hesperornis and 
Iclitliyornis. In the latter the teeth are in sockets, in the 
former in grooves. In both the teeth are numerous, but 
not, perhaps, extending on to the premaxillaries ; the teeth 
show no specialisation in different regions ; they are of 
dentine coated with enamel, and in Hesperornis the basal 
portion of the roots consists of osteodentine. 

The Eocene bird (from the London clay) Odontopteryx 
toliapicus has a strongly serrated upper jaw, a state of 
affairs which is paralleled in the South American passerine 
Phytotoma rara. In this latter bird, as Parker has pointed 
out, 1 there is a ' row of clearly denned denticles, both along 
the dentary and palatine ridges of the premaxillary.' He 
suggests that in these birds and in the merganser, where 
similar ' denticles ' occur, the bone of the jaw has grown 
into arrested dental papillae. 

The oesophagus dilates in a few birds into a crop, which 
is more highly specialised in Opisthocomus (q.v.) than in any 
other form. When the crop is well marked it consists of a 
spherical to oval dilatation of the oesophagus, which in 
pigeons is divisible into a right and left half and an inter- 
mediate unpaired portion. The gallinaceous birds, the 
parrots, and among the Limicolae the American genera 
TTiinocorys and Attagis, are provided with a crop. In other 
birds a slight dilatation of the oesophagus, either permanent 

papillen bei Vogeln,' J.B. nat. Gcs. Leipzig, 1882, p. 16 ; A. F. J. C. MAYEE, 
Ziihne im Oberschnabel bei Vogeln,' &c., Froriep's Notiz. xx. 1841, p. 69 ; 
BLANCHARD, ' Observations sur le Systems Dentaire chez les Oiseaux,' Coni/iir.-. 
Rend. 1. 1860, p. 540 ; M. BRAUN, ' Die Entwicklung des Wellenpapageis,' Arb- 
Zool. Zoot. lust. Wiirzb. v. 1879. 

1 In his memoir upon githognathous birds. 


or temporary, foreshadows the fully developed crop of the 
birds mentioned. 

The stomach ' consists of two compartments following 
each other, the glandular proventriculus and the more mus- 
cular gizzard. The proportions of these two segments of 
the stomach vary, and both are much reduced in the 
hoatzin, whose crop appears to take on the function of a 
gizzard. The proventriculus is usually, but not always, 
separated by a marked constriction from the gizzard, and has 
a patch of large glands which generally forms a band lining 
the upper part of the sac and continuous right round it ; to 
a proventriculus in which the glandular patch is disposed in 
this fashion the term ' zonary ' is applied. More rarely the 
patch of glands is a single oval or round patch not continu- 
ous round the proventriculus, or there may be two such 
patches. In Plotus anhinga the two patches of proven- 
tricular glands are contained in a special diverticulum 
of the proventriculus. In Tantalus ibis MITCHELL 2 has 
described a remarkable divergence from the usual structure 
of the proventriculus. In this bird the glandular areas are 
two, as in other storks. Above these is a row of crypts, 
which are partly glandular and partly lymphatic, and are 
believed to be organs for the absorption of water. Among 
the Steganopodes and in other birds the proventriculus is 
much larger than the gizzard, which follows. In certain 
tanagers this state of affairs culminates in the apparent 
absence of the gizzard as a distinct structure (see below). 
LUND and FOKBES have mentioned a number of tanagers in 
which this occurs. The gizzard is more muscular in grain- 
eating and in some other birds than it is in flesh- and fish- 
eating birds. It is strong and hard and lenticular in form in 
the Galli, Ralli, &c., bag-like and soft-walled in the heron, 
&c. The lining of the gizzard undergoes a remarkable 
modification in certain pigeons (q.v.), where it may be even 

1 ' On the Proventricular Crypts of Pseudotantalus ibis,' P. Z. S. 1895, p. 271. 

* A comprehensive work upon this organ is that of CAZIN, Ann. Sci. Nat. (7), 
iv. 1887, p. 177. See also the same, ' Structure et Mecanisme clu Gesier des 
Oiseaux,' Bull. Soc. Philom. 1888, p. 19. 


The intestine of birds varies much in proportional as well 
as (naturally) in actual length. In the systematic part of 
this work a number of actual measurements will be found ; 
from these it is obvious that on the whole purely frugivorous 
birds have a short gut, while fish- and grain-eating birds have 
a long gut. To compare, for example, two birds of roughly 
the same size but of different feeding habits, the touraco 
and the common pigeon, we find in the former a gut of 
42 c.rn., and in the latter of 108-132 c.m. As the gut is 
always longer than the abdominal cavity in which it lies, it 
has to be thrown into folds in order to find room. 

In the embryo chick the gut is straight and is supported 
by a continuous dorsal mesentery of equal vertical diameter 
throughout. The coiling is both lateral, which results in 
lateral foldings of the mesentery, and vertical, which results 
in unequal growths of the mesentery. It only affects the 
middle part of the alimentary tract, the oesophagus and 
stomach on the one hand, and the rectum on the other, or 
at least a part of it, retaining the original straight condition. 
The lateral foldings give rise to secondary connections 
between different regions of the mesentery, and tend to 
obscure the course of the gut ; but it is easy, by carefully 
removing the entire intestine to distinguish these secondary 
mesenteries from the primary sheet binding the gut to the 
dorsal body wall. 

When the body walls of a series of birds are removed, and 
the disposition of the intestines thus shown examined, they 
have been found to present great differences. These have 
been studied and described by GADOW in two memoirs, 1 and 
the main results extracted for the account of the digestive 
system in Newton's ' Dictionary of Birds.' It is mainly 
from the latter work that the abstract here given is drawn. 

In a goose, for example, the main disposition of the 
intestinal folds is in a longitudinal direction ; they run 
parallel with each other in a direction roughly coinciding 

1 ' Versuch einer vergleichenden Anatomie des Verdauungssystemes cler 
Vogel,' Jen. Zcitschr. xiii. 1879, pp. 92, 339 ; ' On the Taxonomic Value of the 
Intestinal Convolutions in Birds,' P. Z. S. 1889, p. 303. 


with the long axis of the body. On the other hand the 
intestines of a gull, seen when the body wall is cut through 
and without any other disturbance, have a watch spring -like 
arrangement. In Platalea the coils are parallel, but mainly 
at right angles to the long axis of the body. 

These and other variations have been mapped out by 
GADOW into seven principal schemes, which are represented 








' / 











/ ,' 








' 1 





a, isoccelous ; 6, auticoelous ; c, antiperiecelous ; rf, isoperiooelous ; e, cj'cloccelous : 
/,</, plagioccelons ; /(, telogj'rous ; P, pylorus. 

]ST.B. The ascending branches are dotted. 

in the annexed cuts. These are, of course, not accurate 
pictures of the actual course of the gut, but diagrams ' of the 
principal relative positions ' of the intestinal loops ; and it 
must be further explained, in relation to another mode of 
mapping the intestine that we shall refer to immediately, 
that the diagrams, mainly if not entirely, concern the lateral 
foldings of the mesentery that have been already mentioned ; 
they are representations, in fact, of the relations of the 
folding of the gut to the body cavity and not to the medial 
line of attachment of the mesentery ; nor is any attention 


paid to fixed points in the intestine, such as the cseca or the 
vitelline duct. They express, however, an interesting series 
of facts. The general term of orthoccelous is applied to 
those cases where the folds are as a rule parallel to each 
other and in the long axis of the body. When they form 
spirals the general term of cycloccelous is applied to them 
by GADOW. 

The prevailing number of loops is four, of which the first, 
the duodenal (which contains the pancreas), is a loop which 
rarely undergoes additional twisting. The orthoccelous con- 
dition may be regarded as the starting point. The cyclo- 
coelous arrangement, as will be seen by the figure, is derived 
from the orthocoelous by the conversion into one spiral of 
the second and third loops. In all the Passeres the cyclo- 
coelous arrangement is arrived at by a spiral twisting of the 
middle or second loop only (there being but three loops) ; 
this kind of gut has been termed by GADOW mesogyrous. 
The Limicolae, which have a spiral formed by the second and 
third loops, are also, of course, mesogyrous ; but it is clear that 
a similar state of affairs has been arrived at independently. 
Finally, there is the telogyrous condition, in which merely the 
end of a given loop or loops becomes twisted into a spiral, 
the rest remaining straight. This is shown in the last of 
the series of figures on p. 24. In such cases, as in the one 
figured, the duodenal loop rarely, but still occasionally, under- 
goes a twisting. The plagiocoelous condition is an irregular 
twisting of the ends or of parts of the loops of an orthocoelous 
gut. These varied arrangements of the gut may be recog- 
nised in most birds ; but there are a few exceptions of which 
note must be taken. In certain fruit-eating birds, such as 
Carpopliaga, liliamphastos, the gut is so short and wide that 
the number of loops is reduced, and the arrangement quite 
undecipherable. On the other hand the extremely lengthened 
gut of the fish-eating 'J'ninlion produces an equal confusion. 

Since GADOW' s description of the coils of the intestinal 
canal in birds the subject has been studied from another point 
of view by CHALMEES MITCHELL.' GADOW considers only 

1 'On the Intestinal Tract of Birds,' P. Z. S. is'.Hi, p. 13(5. 



the way in which the folds of the gut are packed away in the 
body cavity. MITCHELL describes the actual coiling of the gut 
itself. In its simplest condition the gut of any animal, as is 
shown in their embryos, is a straight tube, passing from the 
stomach to the cloaca, supported by a continuous dorsal mesen- 
tery, the ventral mesentery being in nearly all vertebrates 

defective so far as the intes- 
tinal region is concerned. 
Thi,s simple condition is, 
however, not retained in any 
existing bird ; in all the length 
of the tube is to some extent, 
generally to a large extent, 
longer than the body. The 
alligator (fig. 8) offers the 
ideally simplest condition of 
a coiled intestine, where ad- 
ditional length is achieved 
without any complications 
of the gut, merely by its 
8.-Alligator Mississippi* ; bein thrown into a series of 

ALIMENTARY TRACT (AFTER CHALMERS folds, of which all are more 

or less alike. So simple a 

condition as this does not occur in any known bird. But 
there is more than one type in which this arrangement is 
retained w r ith but little modification. It is a significant fact 
that the most primitive arrangement of the folds of the 
intestine, judged from the crocodilian standpoint, than which 
we have none other more riearty approximating to the 
probable reptilian ancestor of birds, is found in birds which 
other considerations lead us to assign a low position in the 
avian series. In the accompanying drawing (fig. 9) of the 
screamer, for example, we have a gut which is but slightly 
advanced from that of the crocodile. The greater part of 
the small intestine shows the same series of undifferentiated 
folds, only the duodenal loop (not missing as a specialised 
fold in any bird) being separated off from the general coiling. 
The large intestine, however, differs from the short and 



straight large intestine of the crocodile by its convoluted 
course. So too with the gallinaceous bird (fig. 10), where 

FIG. 9. Cliauna cliacaria ; ALIMENTARY TEACT. 

s, proveutriculus : g. glandular tract : </. duodenal loop ; l-l, large loop of small 
intestine ; y, vitelline dtift : c, caeca ; Li, large intestine ; p.v, portal vein ; r.v, 
rectal vein. (After CHALMKRS MITCHELL.) 

FK;. 10. Argus gic/finti'iix, CHICK; INTESTINAL 


FIG. 11. Gasuarius ; INTESTINAL TRACT 

the duodenal loop is again the only specialised region of the 


The ostrich and the cassowary present but little modifi- 
cation of the same 
primitive type of gut. 
In the former we have 
a rather more com- 
plicated duodenal loop, 
which is furnished with 
a small subsidiary fold. 
Then, too, the large in- 
testine has the same 
remarkable folded con- 
dition seen, but to a 
less extent, in Chauna. 
Casuarius is simpler 
than Struthio. In all 
these figures (taken 
from Mr. MITCHELL'S 
paper) the course of 
the principal blood 
vessels is shown ; only 
the veins, as the arteries 
were found invariably 
to accompany the veins. 
It will be observed that 
the principal feeder of 
the portal vein runs 
directly across the main 
mesenteric fold, and 
ends near to the vitel- 
line duct, the rudiment 
of the yolk sac ; from 
the junction of this 
with the main stem of 
the mesenteric vein 

arise two other vessels ; one of these supplies the duodenal 

loop, the other the large intestine ; there is sometimes a 

FKI. 12. Struthio camelus; INTESTINAL 

x, short-circuiting vessel cut across. (After 



' short-circuiting ' connection, as MITCHELL terms it, 
between the duodenal and the main mesenteric stem. 

Fi(i. 13. Haliadus albicilla ; IXTSETINAL TRACT. 
x as in fig. 12. (After CHALMERS MITCHELL.) 

FIG. 14. Am uniniium ; I.\TI:STINAL TKACT. 
., as in fig. 12. (After CHAI.MKKS .M m 11 KLL.) 

In other groups of birds this simple state of affairs is further 
evolved, but in some there are remains of the primitive 


folded intestine for a greater or less extent ; thus in the 
penguin a good deal of the intestine retains the early 
system of folding. But this bird is remarkable for the 
extraordinary complications of the duodenal loop, which has 
become enormously lengthened, and, in Eudyptes, thrown 
into a series of secondary loops, which in Aptenodytes 
are arranged in a spiral a convergent resemblance, thinks 
MITCHELL, to Haliaetus (see above, fig. 13). Platalea, 
which, possessing as it does the complete muscle formula of 
the leg, must be regarded as a primitive type among the 
storks, has the greater part of the small intestine disposed 
in the primitive fashion ; the duodenal loop is, of course, 
distinct, and there is also a well-marked long loop just 
before the large intestine. Haliaetus albicilla, to which 
reference has just been made, is less divergent from the 
primitive form than many other birds, and Lanis marinus 
has preserved considerable traces of the same. Among the 
most specialised birds from the present point of view are 
the parrots (see fig. 14) . 

In Ara, at any rate, there is nothing left of the original 
primitive folds of the intestine ; the whole of its course is 
disposed into six specialised loops. The owls and the Capri- 
mulgidse appear, according to MITCHELL, to have preserved 
much of the primitive short, convoluted loops of the lower 
birds, and so, though to a less extent, has Corythaix. The 
very short intestine of the Passeres, represented here by Parus 
major, is not so easy to understand, but its general appearance 
is that of a primitive gut. Some further details will be found 
in the succeeding chapters, under the different families to 
which they refer. 

The caeca lie at the commencement of the large intes- 
tine, which indeed can be, as a rule, only differentiated from 
the small intestine by the break thus caused. This is, how- 
ever, not invariably the case, for in Stntthio (see fig. 12, p. 28) 
there is a distinct break between the two sections of the 
gut, quite independent of the caeca, which can be readily 
seen from an inspection of the figure cited. The caeca are 
among the most variable organs of birds. Not only are they 


sometimes completely absent, as in the majority of that 

large assemblage of birds the picarians, but when present 

they show every degree of relative size. In the Passeres, and 

in some other birds, the two caeca are the merest nipples, 

which cannot be believed to serve any function. They are 

in the same way reduced in size 

in the hawks and storks. On the 

other hand, in the gallinaceous 

birds, in the Limicolae, and in some 

others the caeca are large tubular 

diverticula of the gut, the length 

being sometimes to be measured 

by inches. Among the owls the 

caeca are large, and have the i''i<- i ; j. I'am* major. 

additional peculiarity of being (fiESSteo^S 

swollen at the blind extremity. 

The caeca are most complicated in the ostrich, the screamers, 

and the tinamou Calodromas, under the descriptions of which 

birds will be found an account of this organ. The herons 

are remarkable for the fact that one of the two casca has 

disappeared, the remaining one being but small. 

The liver is invariably composed of two lobes, of which 
the left often shows a more or less distinctly marked 
secondary division. The size of the lobes varies greatly, as 
does their relative size. Thus in some birds the liver lobes 
are quite hidden by the sternum ; in others again they 
descend some w r ay down below the shelter of that bone and 
are apparent when the muscular walls of the abdomen are 
cut through or removed. The two lobes are occasionally 
equal or subequal in size ; more generally there is a dis- 
crepancy, the right or left, as the case may be, being the 
larger, sometimes very much the larger. The two lobes of 
the liver are commonly firmly attached to each other by a 
bridge of hepatic tissue. In Chauna they are nearly separate, 
being only united by a very narrow isthmus of liver substance. 
The liver sometimes (e.g. mRhynchotus rufescens) has two or 
three small vessels, belonging to the portal system, entering 
its substance at the free edge, a state of affairs which has a 



very lizard-like appearance. The relative sizes of the 
liver lobes appear to be of no importance systematically. 


<!./!., gall blail'lev ; //.</., o/., bile durts ; p t pancreas. 

Numerous livers are figured by GADOW in Bronn's ' Thier- 


The gall bladder is an organ which is not invariably 
present in birds. It is even sometimes present and some- 
times absent in the same family (e.g. parrots). As a rule 
this vessel is of a rounded or oval contour and is embedded 
on the surface of the right lobe of the liver. The Picidse, 
Capitonidse, and Khamphastidae are remarkable for the 
extraordinarily elongated gall bladder, which reaches a long 
way down the abdominal cavity ; this is described more 
fully below. The penguin has an almost equally elongated 


FIG. 17. DUODENUM OF Syn-liaptes. 

v.f, gall bladder ; d.c, cystic duct ; dJi, 
hepatic duct ;, d. pi, pancreatic 
ducts. (After BUAXDT.) 

rhaptes (AFTER BRANDT). 

gall bladder. The position of the apertures of the cystic 
and hepatic ducts upon the small intestine varies. The 
ostrich is remarkable for the fact that the single duct opens 
practically into the stomach. 

The pancreas lies in the fold of mesentery that unites 
the two arms of the duodenal loop. It is commonly more or 
less distinctly composed of two parts, and in relation to this 
there are two pancreatic ducts which pour its contents into 
the duodenum. Apparently, however, no value can be 




attached to either the form of the gland or the number and 
position of the orifices of its ducts. In Syrrhaptes para- 
doxes, for instance, both of the arrangements figured in the 
accompanying cuts have been found by BKANDT, who inves- 
tigated the structure of the bird. In one of them both 
ducts open close to each other and to the cystic duct on the 
ascending part of the duodenal loop ; in the other the cystic 
and hepatic ducts were on opposite sides of the duodenal 
loop, and in common with each opened a single pancreatic 

Rhea americuna. 

hc.e, Jie, bile ducts ; pl,p2, pancreatic 
ducts. (After GADOW.) 

LOOP OF Rli. Darwini. 

f-./ii; bile ducts; pl,p2, pan- 
creatic ducts. (After GADOW.) 

duct. This latter arrangement was found by GADOW in 
Pterocles. In two species of Rhea the relative positions of 
the pancreatic and bile ducts were as is shown in the figures. 
In the owl Photodilus badius I found that the cystic duct 
opened near to the summit of the ascending arm of the 
duodenal loop ; below this opened the hepatic duct, and 
some way below this again, and near together, the two pan- 
creatic ducts. A good many details upon this subject will 
be found in GADOW'S paper on the digestive organs of birds. 
The cloaca of birds is the terminal chamber of the 
alimentary canal, which also receives the urinary and genital 
ducts, and is provided with an appendix of unknown function, 


the so-called bursa Fabricii. GADOW, in a recent work ! upon 
this region of the alimentary canal, recognises three cham- 
bers in the cloaca. Above, and separated by a constriction 
from it, is the coprodseum, into which the rectum opens ; 
this is divided by a constriction from the middle chamber, 
or urodaeum, which receives the genital and urinary ducts ; 


, rectum ; b, orifices of ureters ; //, genital papillae ; c, fold separating coprodieum from 
urodaeum : </, fold separuthifr urodseum from proctodaeum ; e, opening of /, bursa 
Fulirieii. (After FORBES.) 

then follows the proctodseum, of which the bitrsa Fabricii 
is a diverticulum. 

The bursa Fabricii has been chiefly investigated by 
FORBES 2 and WENCKEBACH. 3 It is a dorsal diverticulum of 
the proctodseum, and therefore has nothing to do with the 

1 ' Remarks on the Cloaca and on the Copulatory Organs of the Amniota,' 
Phil. Trans, vol. clxxviii. p. 5. 

- ' On the Bursa Fabricii in Birds,' P. Z. S. 1877, p. 304. 

3 ' De Ontwikkeling en de Bouw der Bursa Fabricii,' Inaucj. Diss. Leyden 
1888. See also E. RETTERET, ' Contribution a 1'Etude du Cloaque,' &c., J. de 
VAnat. xxi. 1885, p. 369. 

D 2 



ventral bladder of other vertebrates. It is largest in young 
birds, and often becomes obliterated in older birds. The 
general relations of the bursa to the cloaca are shown in 
the two accompanying figures. The organ contains a 
quantity of lymphatic follicles, and presents us with two 
types. In most birds it is a diverticulum opening by a 
narrow neck into the proctodseum ; but in the struthious 
birds (in the young at any rate) it is not constricted at its 
orifice into the proctodneum, and the boundaries of the two 
are therefore indistinct. The structure and arrangement of 

'R, coproclajuni ; C, urodfeum ; D, proctodseum ; B, bursa ; d, ureters. (After FORBES.) 

the follicles and of the bursa generally have led WENCKEMANN 
to certain classificatory conclusions. 

Reproductive and Renal Organs 

The kidneys are so unimportant from the point of view 
of the present book that they can be dismissed in a few 
words. Each kidney is a looulated organ lying in the pelvic 
region, so closely in contact with the adjacent bones that 
they are marked by grooves upon the dorsal surface. In 
some hornbills each kidney is divided into an anterior and 
a posterior piece, which are perfectly separated. A ureter 
runs from each kidney to the urodseum. 

The reproductive organs consist of a pair of testes in the 
male, and of one, rarely two, ovaries in the female. Corre- 
sponding to the single ovary (the left) is a single oviduct, 
the right one remaining rudimentary. BALLOWITZ ' has 

1 ' Untersuchungen iiber die Struktur der Spermatozoen,' *c., Arch. Mikr. 
Anat. xxxii. 402. 


figured the spermatozoa of many birds, whence it appears that 
their form is often characteristic, and may be of systematic 
use. A penis is not present in all birds ; it exists in the 
Struthiones, the Anseres, Tinami, Herodiones, Galli, and 
(EcUcncmus. It is a paired organ that is to say, it is 
composed of two incompletely joined halves with a longi- 
tudinal groove. 

The Ccelom 

Birds differ both from reptiles and mammals in the com- 
plication of the subdivisions of their body cavity. The 
subject is one that is far from being thoroughly worked out, 
but enough information has been collected to allow of a 
certain amount of definite statement and of comparison 
with other animals. 

When a bird is dissected in the usual way from the 
ventral surface, the abdominal cavity, or at least the cavity 
containing what are generally termed the abdominal viscera 
i.e. liver, intestines, &c. is seen to be divided into two by a 
toughish septum, which varies in extent according to the 
bird dissected. This membrane, to which attention has 
been directed by WELDON ' under the name of ' pseudepi- 
ploon,' has been investigated in a number of birds by me. 2 
Its relations in the common fowl have been described with 
the aid of diagrammatic representations of sections by 

This membrane, believed by MALL 4 to be the actual 
homologue of the omentum of the mammal, is more or less 
horizontal in direction, so that it may be conveniently 
termed, without prejudice to its homologies, the ' horizontal 
septum.' This horizontal septum is attached to the ventral 
body wall, to the oblique septa (of which see description 
later), and to the gizzard, which viscus appears to lie withui 

1 In his memoir upon the anatomy of the storks and flamingo in P. Z. S. 
1883, p. 638. 

- ' Notes on the Visceral Anatomy of Birds,' P. Z. S. 1885, p. 836. 

:i ' On the Subdivision of the Body Cavity in Lizards, Crocodiles, and Birds,' 
P. Z. S. 1889, p. 452. 

1 Joitrn. Morpli. 1891. 


the thick septum, in a cavity formed by the splitting of its 
layers. Anteriorly the horizontal septum, passing forwards, 
lies beneath the liver, coming into relations on each side 
with a cavity which will be referred to later as the ' pulmo- 
hepatic recess.' 

If, therefore, the abdominal walls of the bird have been 
cut through anteriorly to the attachment of the horizontal 
septum to the abdominal walls, the only abdominal viscera 
exposed will be the gizzard and the liver lobes. These 
latter are separated from each other by the median vertical 
'falciform ligament,'' which is continued backwards to divide 
the cavity into right and left halves. 

If, on the other hand, the abdominal walls of the bird 
have been cut through posteriorly to the attachment of the 
horizontal septum to the abdominal walls, the viscera 
exposed will be the intestines and kidneys and not the 

In some birds for instance, in the duck, and in many 
charadriiform birds the horizontal septum is so short 
behind the gizzard that the latter is closely attached to the 
abdominal parietes by what looks at first sight almost like a 
pathological adhesion, due to peritonitis. On the other 
hand in many storks, in Chauna, Cariama, struthious and 
other birds, the horizontal septum is very extensive, reaching 
back to the immediate neighbourhood of the cloaca. Various 
intermediate stages are offered by other birds. 

The Oblique Septa. Eeference has been already made to 
these structures, which are present in all birds, and concerning 
whose homologies there is some divergence of opinion. Their 
structure and relations are as follows : On either side of the 
body is a tough fibrous sheet of membrane, which runs an 
oblique course (hence HUXLEY'S name 1 of oblique septum'), 
entirely enclosing and shutting off from the abdominal cavities 
(dorsal and ventral) the lungs and air sacs, with an excep- 
tion to be noted immediately. These oblique septa have, 
as HUXLEY pointed out, a tent-like arrangement, coming 
into contact with the median septum in front of the heart, 

1 On the Respiratory Organs of Apteryx,' P. Z. S. 1882, p. 560. 



thence diverging to be attached ventrally to the sternum 
along two lines, one on each side, set obliquely to the median 


-' jy 

V.I.I. M '' 


/- \', air sacs ; 7V, trachea ; </./, dorsal margin ; v.n.l, ventral margin of lungs ; 1-3, dissepiments 
between air sacs ; Ao, aorta ; c.a, coeliac artery ; in. u, mesenteric ; m, muscular fibres ono.s, 
oblique septum ; F^S, vertical median septum : L.c, longus colli; A', kidney. (After HUXLEY.) 


* t 


\ \ 



irr * 







attachment of the falciform ligament. Dorsally they are 
attached to the parietes. As a general rule the abdominal 
air sac is the only one of the posterior air sacs (see fig. 23) 
which is not enclosed within the oblique septa ; in all birds, 
so far as is known, except the Apteryx, the wall of this air 
sac ' has been apparently driven out, like a hernial sac, 
between the peritoneum and the parietes, and projects into 
the abdominal cavity.' In Apteryx the air sac in question 
is completely enclosed by the oblique septum. Another 
exception to the statement made above as to the completely 



rt, as a free fold ; 6, falciform ligament. 

dissepirnental nature of the oblique septa occurs in a few 
birds (e.g. emu and Cariama), in which the posterior end 
of the oblique septum, though firmly attached to the dorso- 
lateral parietes, is not so attached ventrally, but projects 
into the abdominal cavity as a free fold. In these cases the 
free fold is double (see fig. 25), the inner half being conti- 
nuous with the horizontal septum. To the possible signi- 
ficance of this fact we propose to return later. 

The oblique septa are, as has already been stated, mem- 
branous, but they are occasionally and partially invaded or 
covered by muscular tissue. HUXLEY speaks of ' unstriped 



muscular fibres ' in the oblique septum of the duck ; and in 
the puffin } (Fratercula arctica) and the penguin (Eudyptes, 
Eudyptula minor and Spheniscus demersus) the posterior 


L.L,R.L, lobes of liver; L, luugs; A.S, horizontal septum ; O.S, oblique septum. 

FIG. 27. SIMILAR DIAGRAM OF CROW (Cori'iis capellanus). 

, rudiments of sternal attachment of oblique septum. Other letters as in tig. -ii. 

part of the oblique septum is largely covered with a thickish 
layer of muscular fibres, which FILHOL 2 (their describe) 1 in 

1 BEDDARD, 'Notes on the Visceral Anatomy of Birds,' II. 'On the 
ratory Organs in certain Diving Birds,' P. Z. S. 1888, p. '2-VJ. 

- ' Sur la Constitution clu Diaphragme des Kinlijiitca.' Hull. Sc. J'lnlani. (7), 
vi. p. 2a"). 



Eiidyptes) has termed ' muscle diaphragmatique transverse.' 
These muscles are, however, striated ; but the duck is not 
the only bird with unstriated fibres in the oblique septum, 
for these also occur in the toucan. 

In all birds, with the exception of certain passerines 
possibly of the entire group of passerines the oblique septa 
have the structure and relations that have been thus briefly 
described. In passerines l they have undergone what appears 
to be a modification. The oblique septa of each side, instead 


fit, gizzard ; L, liver; O.S, oblique septum. The liver is covered by a membrane 
continuous with the oblique septa. 

of being attached independently to the sternum, become 
fused with the falciform ligament in the middle line, and 
form a horizontal sheet of membrane covering over the two 
lobes of the liver. The original (?) attachments of the 
oblique septa are not, however, in these birds entirely lost ; 
a much fenestrated membrane sometimes, indeed, reduced 
to a thread or two remains to remind the anatomist of the 

1 BEDDAKD, ' On the Oblique Septa in the Passerines, and in some other 
Birds,' P. Z. S. 1890, p. 225. 


more prevalent conditions. In the rook, however (fig. 28), 
they are completely preserved. But the attachment of the 
falciform ligament to the sternum in the median line is lost. 
The cutting off of two lateral sections of the body cavity 
by the oblique septa, and the division of the remainder by 
the horizontal septum, do not, however, exhaust the sub- 
divisions of that space. The liver lobes are attached, as is 
so common among reptiles, to the oblique septa, in the 
neighbourhood of the lungs, by what may be termed the 
pulmo-hepatic ligaments. In making the comparison with 
reptiles we assume for a moment the correctness of BUTLEK'S 
contention that the oblique septa are in reality a portion of 
the pulmonary aponeurosis, a view which will require a 
careful re-examination. This ligament assists in closing 
the pulmo-hepatic recess, which is really an extension forward 
of the abdominal cavity, as shown by MALL'S instructive 
figures of casts of the perivisceral cavity of birds. They are 
narrow cavities, one on each side of the body, walled and 
floored by the ligament mentioned, and by the oblique and 
horizontal septa. The aperture of entrance has been com- 
pared to the foramen of Winslow of the mammal, and so 
named. In some birds there are no further complications of 
the thoraco-abdominal coelom, but of others there are still 
a few facts to relate before dealing with the homologies of 
the various spaces and membranes. In the Australian pas- 
serine StrutJiidea cinerea the liver lobes are each partitioned 
into two by a transverse septum, which runs from the falci- 
form ligament in the middle line to the oblique septum on 
either side. This septum, which is clear and transparent, 
does not actually divide the liver lobes ; it arches over each 
with a free crescentic margin. In some other birds similar 
septa are present, with very nearly the same relations. In 
hornbills, cuckoos, and owls at any rate in some species 
of each family -the two liver lobes are each completely 
shut off from the ventral section of the abdominal cavity 
(the ' subomental space,' as it has been termed) by delicate 
partitions, of which one only, the left, is present in some 
other birds, e.g. Chrysotis Gitildingi. The 



is occasionally presented by birds in which the horizontal 
septum is very short, and therefore almost vertical in direc- 
tion. But in the birds mentioned the septa are perfectly 


rj, gizzard ; a, baud of muscular fibre connecting it with oblique septum ; b, umbilical 
veiu ; //, liver seen through fibrous partition mentioned in the text. 


distinct from, and present in addition to, the horizontal 

Homologies of Oblique and Horizontal Septa. Sir RICHAKD 
OWEN in the year 1838 described the oblique septa of the 
Apteryx 1 as 'a well-developed diaphragm.' His figure, 
indeed, is highly suggestive of the mammalian diaphragm, 
more so than are (in our opinion) the actual structures 
observable in the dissection of the bird. OWEN anxinin'<l 
that the structures corresponded, contenting himself with 
indicating the principal differences between the avian dia- 
phragm and the mammalian, and pointing out how Apteryx 
was nearer to mammals than any other bird. This assumption 
was undoubtedly based upon current opinions of the day, for 
in the ' Le9ons d'Anatomie Comparee ' of CUVIER (ed. '2, 
vol. vii. p. 21) such a comparison appears to be drawn. As 
OWEN justly observed, the imperforateness of the oblique 
septum of the Apteryx is more mammalian than in other 
birds, even struthious, where the abdominal air sacs project 
beyond it ; the obliquity of its direction, too, is paralleled in 
the dugong and manatee, and it is furthermore less oblique 
than it is in other birds. 

SAPPEY and MILNE-EDWARDS also use variants of the 
word ' diaphragm ' to describe what we term in the present 
work, following HUXLEY, the oblique septa. But for the two 
first-named observers the costo-pulmonary (see under descrip- 
tion of lungs, below) muscles also form part of the diaphragm. 
The older observers, impressed with certain resemblances 
(such as warm-bloodedness) between birds and mammals, 
regarded them as more nearly akin than is at present the 
belief. HUXLEY devoted much work to the demonstration 
of the nearer relationship between birds and reptiles a 
relationship which is now generally held. In redescribing 
the respiratory organs of Apteryx HUXLEY pointed out 
their thoroughly ornithic character, and remarked that ' in 
this, as in all other cases, the meaning of ornithic peculiari- 
ties of structure is to be sought not in mammals, but in 
reptiles.' It is on a priori grounds likely enough that there 

1 ' On the Anatomy of the Southern Apteryx,' Trans. Zool. Soc. ii. 276. 


are structures in birds comparable to the diaphragm of 
mammals; but in that case the likeness would be due to the 
derivation of both from a common form, perhaps low down 
in the reptilian series. At any rate certain anatomical facts 
forbid a precise comparison of the entire oblique septa of the 
bird to the mammalian diaphragm, though, as will be shown 
in the sequel, there appear to be a few genuine resemblances 
which can hardly be got over. HUXLEY emphasised the fact 
that in the bird the heart lies behind the so-called diaphragm, 
which is moreover not supplied by a phrenic nerve. If com- 
parable to any structure in the mammal, it is with the medias- 
tinum that they should be homologised. As to the phrenic 
nerve, several considerations present themselves; it is, it is 
true, a specialised nerve, but it is spinal in origin. Now in birds 
it is also branches of spinal nerves which supply the oblique 
septa. In the mammal it may be that the pulling out of the 
phrenic nerve may be due to a cause similar to that which 
has produced the looping of the recurrent laryngeal nerve, 
When nerves are drawn out by a change in position, or by an 
elongation of the structures which they supply, there is at 
least a tendency for their roots of origin, of many, to fuse 
into a single nerve ; witness, for instance, the limb nerves 
arising from the anterior and posterior plexuses. The fact 
that the spinal nerves which form the limb plexuses are not 
always exactly the same has not led to any very serious 
belief in the serial homology only of the fore limbs in any 
two vertebrates which show these differences. The lungs, 
and consequently the diaphragm (assuming for a moment its 
correspondence with the mammalian diaphragm), are further 
back in birds ; hence their different nerve supply. If we 
look upon the posterior portion of the oblique septa, which is 
alone, be it observed, muscular, as the homologue of the 
lateral parts of the mammalian diaphragm, the rest being 
absent, no great violence to the mutual relations of the 
different structures concerned will have been done. In any 
case it may well be that both the mammalian diaphragm 
and the avian have been derived from some such reptilian 
structure as is to be seen in the crocodile. 


Professor HUXLEY ' has made such a direct comparison. 
' As in birds, the liver [of crocodiles] lies between the 
stomach and the pericardium, and has a peculiar peritoneal 
investment, shut off from the great 'sac of the abdomen ; and, 
as in the ostrich, the whole circumference of the stomach is 
united by fibrous tissue with the parietes. A fibrous expan- 
sion extends from the vertebral column over the anterior face 
of the stomach, the liver, and the dorsal and front aspect of 
the pericardium to the sternum and the parietes of the 
thorax, separating the thoraco-abdominal space into a respi- 
ratory and a cardio-abdominal cavity, and representing the 
oblique septum of the bird.' Further on we read : ' A broad 
thin muscle arises, on each side, from the anterior margin of 
the pubes ; and its fibres pass forwards, diverging as they go, 
to be inserted into the ventral face of the posterior part of 
the pericardium, and into the ventral and lateral parts of the 
fibrous capsule of the stomach, passing between that organ 
anVl the adherent posterior face of the liver, and being 
inserted into the fibrous aponeurosis which covers the anterior 
surface of the stomach and represents the oblique septum.' 

Professor HUXLEY seems, according to BUTLER, to have 
included the ' omentum ' with the oblique septa in his com- 
parison with the fibrous expansion and the accompanying 
muscle of the crocodile. I have already pointed out that 
' the entire fibrous expansion which arises from the vertebral 
column, and extends over the anterior face of the stomach, 
liver, &c., in the crocodile represents both the oblique septa 
and the omentum in the bird.' A justification for this 
opinion is to be seen in the dissection of an emu and one or 
two other birds. We have occasionally observed that where 
the posterior part of the oblique septa is free from the 
abdominal walls, ending, in fact, in & free edge within the 
abdominal cavity, this edge is really continuous with the 
horizontal septum, as show r n in the cut (fig. 25). The 
oblique septum is thus merely a fold of the horizontal septum ; 
they form one continuous structure. As to the muscles of 
the crocodile mentioned in the quotation just made from 
^H ' 'On .l/i/i' //.'.' loc. cit. p. 568. 


Professor HUXLEY, we have already referred to the existence 
in birds of what may be considered their homologue, and we 
have only to add that in a hornbill the left-hand portion of the 
horizontal septum was muscular that, at any rate, a strong 
band of muscle bound the gizzard to the left oblique septum. 

Circulatory System 

The bird's heart is very uniform in structure ; there are 
very few and but slight differences in any part of the heart 
between the most and the least specialised forms. It is, 
however, in certain particulars equally distinctive in structure, 
and differs in a number of well-marked points from the heart 
of either reptile or mammal. As might be expected, the 
reptile which shows the nearest approximation in the anatomy 
of its heart to the bird is the crocodile, while the Monotre- 
mata are the mammals which on the other side occupy a 
corresponding position. 

As with the mammalia the heart is completely separated 
into four chambers ; in the bird the heart has perhaps more 
of an elongated form than in the mammal, the apex (which, 
as in the mammal, is formed by the left ventricle alone) being 
rather more pointed than in the heart of any mammal. In 
a transverse section through the ventricular walls a notable 
difference in the relative dimensions of the right and left 
ventricles for the two types is apparent. It will be noted 
that in the bird the cavity of the right ventricle is, as it were, 
partially wrapped round that of the left, and is in consequence 
of a decidedly crescentic form. The cavity of the right 
ventricle of the mammal's heart is more oval in form, and is 
not wrapped round that of the left. In this particular the 
Monotremata stand midway between the bird and the higher 

The interest of the structure of the bird's heart, however, 
largely, for reasons of comparative anatomy, centres in that 
of the valve which guards the orifice from the auricle. The 
interior of that ventricle has fairly smooth walls, a sculp- 
turing so conspicuous in the mammalian ventricle being 


almost entirely absent. The valve itself, represented in fig. 
30, really consists of two parts, which are distinguished by 
the insertion of a large papillary muscle, which ties the entire 
valve to the free wall of the ventricle. It is only rarely that 
any other representative of the generally numerous papillary 
muscles and chordae tendineae of the mammalian heart occur, 
but occasionally a few muscular threads in addition to the 
single papillary muscle are to be found. Their existence has 
been noted, for instance, in Aptcnjx an strolls and in a few 
other birds. It has been usually held that the muscular 
right auriculo- ventricular valve of the bird's heart represents 



papillary muscle ; 6, c. valve. (After 

FIG. 31. HEART OF Aj>fi'i-i/.>\ INTERIOR 

X, Hap of ventricular wall removed with 
muscle. (After LANKKSTKII.) 

only one half of the complete valve of the mammalian and 
crocodilian hearts. In these last-mentioned animals the 
entire circumference of the ostium, which leads from the 
auricle into the ventricle, is surrounded by the valve, which 
thus forms a complete collar. There is, however, an excep- 
tion in the case of the Monotremata, where the septalflap of 
the valve (i.e. that lying on the side of the ostium which 
abuts upon the interventricular septum) is partially or 
entirely absent. On a careful comparison, however, between 



the bird's heart and that of the crocodile it appears that this 
is not the case. If the hearts of the two animals be 
laid side by side in a corresponding position, it will be 
seen that the crocodile's heart valve is furnished with a 
muscle which seems comparable to that lettered a in the 
bird's heart. And, furthermore, on the septal side of this 
muscle the fibres which in the bird constitute that half 
of the valve have a direction which is quite different from 
that of the fibres in the other and larger half. Finally, 
while the larger half of the valve is never, so far as is 
known, fibrous in character, the lesser half occasionally 
appears to be so wholly or partially. Thus there are some 
grounds for thinking that the bird's right auriculo-ventricular 
valve is composed of a complete outer half and of a smaller 
septal half, presenting, therefore, less difference in this one 
particular from the monotrematous than from the crocodilian 
heart. 1 

We can, therefore, derive the bird's heart as regards this 
valve from a heart like that of the crocodile, in which the 
septal flap has for the most part disappeared. But in one 
bird at any rate there appear to be traces of a still further 
retention of the septal half of the valve. GEGENBAUK, who 
some years since wrote an exhaustive paper upon the verte- 
brate heart, 2 made the following remarks about the heart of 
the condor, which in translation run as follows :- 

' Only in the heart of Sarcorhamphus do I find a peculi- 
arity which has interest in this connection. From the 
anterior origin of the muscular valve on the septum ventricu- 
lorum a fold runs backwards, which is formed by a thicken- 
ing of the endocardium. The fold runs obliquely backwards 
and downwards, and crosses in its direction the margin of 
the muscular va]ve. The course of this fold corresponds to 
the line of origin of the membranous valvular flap of the 
crocodile ; I think it reasonable, therefore, to regard it as a 

1 See for a fuller account BEDDARD and MITCHELL, ' On the Alligator's Heart,' 
P. Z. S. 1895, p. 342. 

- ' Zur vergleichenden Anatomie des Herzens,' Jen. Zeitschr. I860 ; ' Notes 
on the Anatomy of the Condor,' P. Z. S. 1890, p. 142. 


remnant of the structure which is further developed in the 

I have, since that sentence was written, examined 
the heart of a condor, in which was found along a line corre- 
sponding to where the nap would be, were it present, ' a 
series of tiny yellowish spots and vesicles . . . probably 
pathological,' but perhaps, like other pathological structures, 
associated with a rudimentary structure. With this excep- 
tion no trace has ever been found of a septal flap other than 
the small flap already described. 

The left ventricle of the bird's heart has an auriculo- 
ventricular valve, which is completely membranous, and is 
tied to the parietes of the ventricle by tendinous threads 
attached to papillary muscles. 

There is one more structure occasionally present in the 
right ventricle of the bird to which we must direct attention 
before leaving the matter. The late Professor ROLLESTON, 
in his Hunteriaii lecture, described and figured in the heart 
of the cassow r ary a muscular pillar uniting the free and 
fixed walls of that ventricle, to which he gave the name of 
moderator band. This structure occurs in a few r other birds 
-for example, in Clinnga Burmeisteri, where it has been 
figured. In the latter bird, however, there are two muscular 
bridges, which run in the same direction. One of them is also 
connected with the muscle tying the auriculo-ventricular valve 
to the free wall of the ventricle. This may conceivably be a 
rudiment of the septal half of the valve lying to the right side 
of the heart. In any case these moderator bands, which 
are also found in deer and in other running animals, seem 
to be, according to ROLLESTON'S suggestion, a mechanism 
for increasing the power of the ventricle to contract, and 
thus ensuring a more rapid and regular flow of blood into 
the lungs. It is characteristic, where it occurs, of running 
animals. 1 

1 For other facts about the avian heart see A. SABATIEK, ' Etudes sur le 
Cceur,' &c., Ann. Sci. Nat. (5), xviii. 1873, art. No. 4 ; F. R. GASCH, ' Beitriige z. 
vergleichenden Anatomie des Herzens,' Arch. f. Naturg. liv. 1888, p. 119 ; 
C. ROSE, 'Beitrage zur vergleichenden Anatomie des Herzens der Wirbelthiere,' 
Morph. J.B. xvi. 1890, p. 27. 

E 2 



The arterial system of birds 1 is chiefly remarkable for the 
large number of the different arrangements of the carotids. 


i:c, J.c, carotids ; r.s, l.s, subclaviaus ; r.i, l.i, 
right and left innominate ; a, aorta ; ?i, 
its origin. (This aiid five following figs, 
after UAIIROD.) 






FIG. 35. CAROTIDS OF Cacatua. 

Many writers, especially NITZSCH, among the earlier anato- 
mists, have drawn attention to some of these variations. 

1 L. A. NEUGEBATJER, ' Systema Venosum Avium,' Nov. Act. Acad. Nat. Cur. 
xxi. 1845, p. 517 ; EATHKE, ' Uber die Caroticlen . . . cler Vogel,' Arch. f. Anat. u. 
Phys. 1850, p. 184, and ' Bemerk. liber die Entstehung, &c., der gemeinsch. 
Carotis,' ibid. 1858, p. 315 ; GARROD, ' On the Carotid Arteries of Birds,' P. Z. S. 
1873, p. 457 ; C. H. WADE, ' Notes on the Venous System of Birds,' J. Linn 
Soc. xii. 1876, p. 531 ; F. HOCHSTETTER, ' Beitriige zur Entwicklungsgeschichte 
des Venensystems,' &c., Morph. J.B. xiii. 1888, p. 575, and ' tiber den 
Ursprung der Subclavia d. Vogel,' ibid. xvi. 1890, p. 484. 



But the whole matter was described at considerable length 
by GAKROD, who had more abundant material to work upon, 
but who, nevertheless, left for his successor FORBES one out 
of the eight known types to describe. The most prevalent 
type is that illustrated in fig. 32. It characterises a large 
number of birds. The two carotids are of equal size, and 
run up the neck for the latter part of their course in the 
hypapophysial canal. A modification of this (fig. 33) is seen 
in the common bittern and other birds, where the carotids 
are of equal size, but fuse into one trunk early in their 
course. In Phoenicopterus the right (fig. 34) and in Cacatna 





sulphured (fig. 35) the left of the two trunks, which are later 
fused together, are very much the smaller. This state of 
affairs leads to the condition shown in Passeres, where the 
left carotid alone is present (fig. 36) . Quite exceptionally, 
and only seen in two species of the bustard, genus Eupodotis, 
the right carotid alone is present. 

A very curious modification of the carotids is seen (ex- 
ceptionally, according to FURBRINGER) in the hornbill, 
Bucorvus. Here the two carotids are entirely superficial, 
running up the neck in company with the vagus nerves. In 
this case, as OTTLEY discovered, the true carotids are reduced 
to the condition of white imperforate cords, the developed 


carotids being the equivalents of the comes vagi nervi of 
other birds. FOEBES thought that the same might be the 
case with Leptosoma, where the carotids are unusually 
small, and apparently bound together, not absolutely fused, 
as also in Opisthocomus. In certain parrots the left carotid 
artery is superficial, while the right runs in the ordinary 
way within the vertebral canal. This is illustrated in the 
accompanying figure (fig. 37). A final variation has been 
observed by FOEBES in the passerine Orthoiiys, where the 
left carotid, as in all passerines, is alone present ; but it runs 
superficially, and there is no deep right carotid, as in the 
parrots, just referred to. These facts, striking though they 
are, are unfortunately of but little value in classification, or 
at least their value is not understood. We may, however, 
accept FOEBES'S statement that ' no passerine bird has ever 
yet been found with more than a left carotid, and no pigeon, 
duck, or bird of prey without two normally placed ones.' 

In all birds, as is well known, the right aortic arch ! has 
alone persisted. It is, however, a commonplace of the text- 
books to mention the fact that the Eaptores have often a 
ligamentous rudiment of the left. There are occasionally 
(perhaps individual) remains of the left arch more conspicu- 
ous. Thus I have found a considerable tract of the left arch 
capable of being injected, and measuring quite an inch in 
length, in Spizcetus and in the hornbill Aceros. Cariama, 
I may mention, has (at least sometimes) a ligament repre- 
senting the otherwise aborted left aortic arch. 

A variation in the thigh arteries has been noted by 
GAEEOD, who found that in some birds the ischiadic, in 
others the femoral, w r as the most important. In Passeres the 
neotropical Clamatores were termed by him Heteromeri, 
since the femoral was the principal artery ; other Passeres 
(including, however, the neotropical and clamatorial genus 
Rupicola) Homoeomeri, from their ischiadic artery. The fact 

' J. Y. MACKAY, ' The Development of the Branchial Arterial Arches in 
Birds,' &c., Phil. Trans, clxxix. 1889, p. Ill ; J. F. VAN BEMMELEX, ' Die Visceral- 
taschen u. Aortenbogen bei Keptilien u. Vogeln,' Zoul. Anz. ix. 1880, pp. 5'2(>, 


that in different species of the same genus (Centropus) the 
same variation occurs tends to throw considerable doubt 
upon the value of the character, an observation that is fre- 
quently and unfortunately necessary to make in describing 
the anatomy of birds. 

The descending aorta gives off three branches, which 
supply the alimentary canal. These are, in order of origin, 
the coeliac, the superior, and the inferior mesenteric. These 
arteries do not, however, supply certain definite regions of 
the gut. Thus in Bernicla rnbidiceps the left caecum 
receives its blood both from the coeliac and from the superior 
mesenteric. The right caecum, however, appears to be 
supplied by the superior mesenteric alone. The coeliac is 
mainly concerned with the blood supply of the stomach and 
the liver. In Porphyrio the two caeca are provided with 
blood from both coeliac and superior mesenteric. A dissection 
of Platalea leucorliodia showed that roughly the coeliac was 
concerned with the blood supply of the duodenal loop and 
of the posterior part of the intestine ; the mesenteric, on the 
other hand, supplied the anterior part of the intestine, ex- 
cepting the duodenal loop. 

In a few birds (e.g. in Buceros and Haliaetus, GADOW) 
the coeliac and superior mesenteric arise from a common 

The posterior mesenteric supplies the end of the alimen- 
tary tract. The aorta also gives off two spermatic arteries, 
and on either side a crural and an ischiadic. 

We shall deal briefly with the venous system, since it has not 
been up to the present largely used for systematic purposes. 
This is doubtless due in a great measure to our imperfect 
knowledge of the variations that occur. As in the lower 
mammalia, there are two venae cavse superiores. These 
are formed by the union on each side of a jugular, a sub- 
clavian, and a vertebral. The two jugulars are often of very 
unequal size ; the right is usually stronger than the left, but 
the two veins are connected in the neighbourhood of the 
head by a transverse branch. Sometimes the left jugular 
may absolutely disappear. Connected with the subclavian 


are several veins which extend on to the pectoral and even 
on to the abdominal region of the body. Most important 
on physiological grounds of these is the abdomino-pectoral, 
which on each side of the body collects blood from the 
pectoral and abdominal regions, and forms in the female 
during the breeding season a network of vessels with the 
corresponding arteries. 

The vena cava inferior is composed of the two hepatic 
trunks and of an unpaired median portion, the main stein of 
the vena cava. The latter traverses the right lobe of the 
liver, and in its transit receives several smaller twigs from 
the liver. At the commencement of the kidney the vena 
cava divides into, or rather is composed of, the two venae 
iliacae. It has been pointed out that in diving birds the 
part of the vena cava which traverses the liver is wider than 
in other birds, a state of affairs which is paralleled in certain 
aquatic mammals. 

The two common iliacs divide each of them into two veins, 
of which the first to be given off is termed the vena iliaca 
externa. This divides at once in the pigeon (according to 
JOURDAN) into the femoral and into a trunk which runs 
along the kidney, and after giving off the sciatic and 
numerous branches to the kidney substance receives the 
hypogastric from the pelvic region, and then joins its fellow 
in the middle line ; at the point of junction of the two 
iliacae externae a median coccygeal is received, and a median 
mesenteric from in front. The other branch of the iliaca 
communis is purely renal. It results from what has been 
said that blood entering the kidney from any of the 
branches of the common iliacs may traverse the kidney 
substance before reaching the heart via the vena cava 
posterior. The suprarenal bodies also have their portal 
system. The body of each side receives a branch from the 
rib region and from the branch of the iliac which runs 
embedded in the substance of the kidney. 

The existence of a renal portal system in birds is 
therefore possible, but not certain, on the anatomical facts 
available ; but the liver portal system, as in all other verte- 


brates, is quite certain. All the blood from the alimentary 
canal, however, need not reach the heart via the liver. 

In Acer'os iiipalensis the vena cava inferior receives three 
veins from the liver, the abdominal vein (see below), and a 
smaller twig which is compounded of branches from the 
oesophagus. Moreover the blood from the posterior part of 
the intestine, at any rate, may reach the system of the vena 
cava inferior via the mesenteric vein, which, as already 
stated, enters the iliacs at their point of junction in the 
middle line. 

In addition to the main portal trunk a number of small 
veins (five on the left side in Chauna chavaria) may enter the 
liver lobes separately, a state of affairs which is precisely 
that found in lizards. 

The umbilical vein, which is the equivalent of a part at 
least of the anterior abdominal vein of the lower vertebrata, 
is found in very various conditions of degeneration among 
birds. In some it appears to be fairly well developed ; in 
others it is practically absent altogether. In a specimen of 
the hornbill Aceros nipalensis it was as well developed as 
I have ever found it in birds. The vein arose near to the 
posterior end of the abdominal cavity as a double vessel ; 
further forward the two halves joined to form a single 
vessel. The vein is supported by the falciform ligament, 
and the upper of the two component vessels receives, not 
very far from the junction of the two, a recurrent vessel 
from the inside of the sternum. It may be that the 
recurrent nature of this vessel is one among many hints of 
the shortening of the sternum among birds. The anterior 
abdominal trunk does not enter the liver, but joins the 
hepatic vessels, and its blood is conveyed straight to the 

In Platalea leucorhodia I could find no trace of an 
anterior abdominal vein in the falciform ligament. In Gni* 
monachus it was very small ; in a large number of other 
birds of different families the vein was present, and large, e.g. 
Crax globicera, Spizaetus coronatus, Serpentarius, Bucorvus y 
Chauna chavaria, Botaurus stellaris, Bernicla brenta, &c. 


Respiratory System 

Trachea. As a general rule the trachea is a straight tube 
passing into the thorax, where it bifurcates into the two 
bronchi. It is composed throughout of rings, which are 

cartilaginous, or may be 
wholly or partially ossified. 
The rings are generally 

simple rings, which are like 
each other and quite com- 
plete, excepting just at the 
bifurcation, where it is com- 
mon for them to be modi- 
fied in connection with the 
formation of the syrinx. 

This is especially marked 
in the tracheophone Passeres 
and in the Ciconiidae, to the 
accounts of which families (as 
well as to below, ' Syrinx _') 
reference must be made for 
the facts. Other modifica- 
tions of some of the last 
tracheal rings are to be seen 
in the cassowary, where the 
last few are incomplete 
behind, as in the mammals 
and in the bird of Paradise, 
Seleucides (see fig. 38), where 
the membranous interspaces 
between the rings become 

FIG. 38. WINDPIPE OF Seleucides nigra. 

4-11, tracheai rings; in, third bronchial ; ^, largely increased, and the 

rings ossified at both sides, 
but not in the middle, in a peculiar fashion. 

A peculiarity of the trachea, seen in representatives of 
some most diverse groups of birds, is its looping. This is, 
of course, suggestive of the similar looping of the trachea in 



crocodiles and chelonia. The occurrence of this state of 
affairs is not a character of any one group, but it is found 


FIG. 39. BHKAST REGION OF Manucodia. 
Tr, trachea ; P.M, pectoral muscle ; Fern, femur. 

sporadically, as it were, in members of quite different groups. 
The facts have been collected into a general account of the 


matter by FORBES. 1 There are various grades of this 
lengthening of the trachea. In certain species of Phon/j- 
gama and Manucodia among the Oscines the loops, which 
vary in complication, lie, as is shown in the figure of 
Manucodia (fig. 39), beneath the skin. Many of the curassows, 
a few Scolopacidae (Rhynchtea australis and E. capensis), the 
duck Anseranas melanoleuca, have a convoluted trachea of 
the same kind. In the male of Tetrao urogallus the loop is 
present, but is in the cervical region, not in the thoracic 
and abdominal, as in the types just referred to. 

The Syrinx. The voice organ of birds, usually termed 
the syrinx, is, as is well known, situated at the bifurcation 
of the two bronchi. Its complexity varies greatly, though 
it cannot be said that a complex voice organ necessarily 
implies an elaboration of sound-producing power. Some of 
the singing birds and the parrots, whose voices are capable 
of emitting a great variety of tones, have, it is true, a much 
specialised syrinx. But, on the other hand, there are other 
passerines which have just as complicated a syrinx, but can, 
like the raven, by no means vie with some of the starling 
tribe, for example, in range of sound. Then, too, some of 
the singing passerines have syringes which are much simpler 
than those of others which sing as well and no better. 
It is, however, true that in the least differentiated forms of 
syrinx the bird has but one or two notes. The ostrich, for 
example, which has one of the simplest syringes, can roar, 
but possesses no variety of sound. The Apteryx, whose 
syrinx is about on the same level of organisation, appears 
to be absolutely mute. The sounds of the emu are due 
not to its simple syrinx, but, chiefly at any rate, to a 
throat pouch, to which due reference will be made later. 

The syrinx of birds, as has been said, varies considerably 
in structure. Many of the variations will be treated of in 
the systematic part of this book, since they are more of 
systematic than of morphological interest. In this place, 
however, the leading modifications of the organ will be 

1 ' On the Convoluted Trachea of two Species of Manucode, with Bemarks on 
similar Structures in other Birds,' P. Z. S. 1882, p. 347. 



shortly described. The syrinx is an organ special to birds ; 

there are no hints of it in any reptile. In reptiles there is 

no modification at the bifurcation of the bronchi ; the tube 

simply branches, and there are two sets of cartilaginous 

rings where there was but one. In birds the case is 

different, and it may be convenient to commence with what 

may be regarded as the typical avian syrinx, which has been 

termed the ' tracheo-broi/cJiial,' since the end of the trachea 

and the beginning of the 

bronchi take a share in its 

formation. It is fair to term 

this the typical syrinx, since 

it is found in the majority of the 

groups of birds ; it occurs, for 

instance, in such diverse families 

as Passerines, Ardeidae, Eallidae, 

Struthiones, Picariae, &c. The 

accompanying cuts illustrate 

this form of syrinx in a number 

of birds, and from an inspection 

of them the principal features 

in the organisation of this form 

of syrinx may be gathered. 

At the end of the trachea there 

is usually a certain amount 

,, -, . ,, . . FIG. 40. SYIIINX OF Indicato , 

of modification of the tracheal ENLARGED. 

rillgS, which may be more O1 1 ". first bronchial semi-rings. (After GAKKOD.) 

less marked, and may be in different directions. It is not 
necessary to particularise here, and we can select fig. 40 
to illustrate one example of this modification, which consists 
in a complete fusion of the last few rings of the trachea. 
The bronchi are formed at first of the short semi-rings, the 
wide interspace being occupied by membrane, the tympani- 
form membrane, which closes them internally ; the extent 
of this membrane varies, and below it the bronchial semi- 
rings become more closely applied sometimes, indeed, 
becoming complete rings. The tympaniform membrane of 
each bronchus is separated from its fellow by a cartilaginous 



or bony bar, which runs across the base of the trachea, 
arising from the last tracheal ring anteriorly, and attached 
to the penultimate or antepenultimate ring posteriorly. 
This bar is called the pessulus. It is shown in fig. 41, 
which represents a syrinx seen from below. When the 
syrinx is cut open it may be seen that this bony bar bears a 
tough semilunar membrane, directed upwards between the 
bronchi ; the voice is due to the vibrations of this mem- 
brana semilunaris. 

Another external feature of the typical tracheo-bronchial 
syrinx is the presence of a pair of mus- 
cles which arise some way up the trachea 
and are inserted on to an early semi- 
ring of the bronchial series, one on each 
side, or even to the last or nearly the last 
of the tracheal series. From this starting 
point we can follow the modification of 

FIG. 41. SYRINX OF Cymbirlu/uclt/ts, FIG. 42. SYRINX OF Balceniceps. 
< 5, FROM BEHIND (AFTER FoKT.Ks). ', ligamentous rudiment of intrinsic 


the syrinx in a number of directions in the way of com- 
plication, in the way of simplification, and in a direction of 
alteration which can hardly be termed either complication or 

The first sign of simplification is the disappearance of 
the intrinsic syringeal muscles, which in many forms have 
completely disappeared. The disappearance is not neces- 
sarily associated with any other changes in the general 
structure of the organ. Occasionally, as in Bal&niceps (see 


fig. 42), the former presence of the muscle is testified to by 
a thin ligament, which occupies the position that the muscle 
would occupy were it present. 

The syrinx of the hoatzin shows an intermediate stage ; 
the muscle is absent for the lower part of its course, but 
present above ; it is represented below by a fibrous band. It 
seems from what w r e know of the relation of muscle to 
tendon generally that this change is in the direction indi- 
cated and not in the converse direction. The fibrous band of 
the syrinx of Balceniceps has, so to speak, been muscle ; it is 
not on its way to become muscle. The ostrich affords an 
example of a further degeneration of the syrinx, or a reten- 
tion of a very primitive and unspecialised syrinx, according 
as we view the facts. The syrinx of this bird has been care- 
fully described and figured by FORBES/ whose words we will 
quote. ' The trachea inferior to the insertion of the sterno- 
tracheales slightly narrows, having above the antepenultimate 
ring a diameter of about one inch. The tracheal rings are here, 
as elsewhere, entire simple rings of an average depth of about 
15 inch, and are separated only by very slight interannular 
intervals. The trachea is slightly compressed and posteriorly 
carinated for about the last seven rings. The last ring but 
four is somewhat produced downwards in the middle line 
both anteriorly and posteriorly ; it is in consequence nar- 
rower laterally than elsewhere. The antepenultimate ring- 
presents the same features more strongly developed. In 
two of the four specimens examined it sent down a small 
prssuliform process of cartilage in the middle line behind, 
filling the chink left between the posterior extremities of 
the last two (incomplete) rings. The penultimate ring is 
narrower and more cylindrical than its predecessors ; it is 
also wider transversely, and incomplete behind in the middle 
line, its extremities, however, being closely approximated to 
each other. The last tracheal ring is still wider transversely, 
and more cylindrical ; and it, too, is incomplete posteriorly 
to a greater extent than its predecessor ; viewed from the 

' ' On the Conformation of the Thoracic End of the Trachea in the "Eatite " 
Birds,' P. Z. S. 1881, p. 778. 



side it is convex upwards, as are its immediate predecessors 
in a less degree. The interannular intervals between all 
these rings are, when undisturbed, mere chinks filled up 

FIG. 43. SYRINX OF titridhio. 

a, last, <M, penultimate, ooo antepenultimate tracheal rings : I, section of wall of wind- 
pipe to show vocal chord ; 1-7, tracheal, I, II, bronchial rings. (After FOHUKS. ) 

by dense fibrous and elastic tissue. There is no trace of a 
pessulus, though the last tracheal ring is slightly produced 


downwards in front. The first bronchial semi-ring on each 
side is narrow and cylindrical, strongest anteriorly, and 
somewhat attenuated posteriorly. It is separated only by a 


narrow interval from the last tracheal ring. The second and 
third rings are similar, but are more slender and lengthy ; 
they are convex downwards, but very slightly so ; hence the 
interannular intervals are slight here also. Their anterior 
ends are very slight, inturned, impinging but to a small 
extent on the membrana tympaniforniis, which completes 
the bronchial tubes internally, and, in consequence of the 
absence of any three-way piece, passes continuously from 
one bronchus to another, so closing the tracheal tube in- 
feriorly. The fourth, fifth, and succeeding bronchial rings 
are similar in character, but their ends, which tend to be 
dilated posteriorly, are successively more and more incurved 
to about the tenth. Nowhere are the bronchial rings com- 
plete. There is at most only a trace of a membrana semi- 
lunaris, in the form of a feeble, scarcely raised antero-pos- 
teriorly directed fold of mucous membrane.' 

This syrinx, therefore, differs from the more typical 
tracheo-bronchial syrinx in, at any rate, three essentials 
(1) the absence of tracheo-bronchial muscles ; ('2) in the 
slight amount of specialisation of the last rings of the trachea ; 
and (3) in the absence of a pessulus. The only distinguish- 
ing feature of the syrinx which is present is the membrana 
tympaniformis. But the presence of this, and of the rudi- 
mentary membrana semilunaris, fully justified FORBES in 
contradicting the assertion, prevalent at the time when he 
wrote, that the Struthiones had no ' lower larynx,' an assertion, 
indeed, which could not possibly be made with the syrinx of 
Ehca, a quite typically tracheo-bronchial one, in existence. 
Still, it is undoubted that the syrinx of the ostrich is in a 
very simple condition, and hardly deserves the name. In 
the stork tribe we have a series of stages in the degeneration 
of the syrinx. In Abdiinia splienorlujnclia, as in other 
storks, there are no intrinsic syringeal muscles, but the 
membrana tympaniformis is well developed and of consider- 
able extent. In Xenorhynclius the membrana tympaniformis 
is almost, but not quite, obliterated, and, finally, in Ciconin 
the bronchial rings are rings, and not semi-rings ; there is in 
them no trace of a membrana tympaniformis ; but in all the 




pessulus is present. The syrinx of the stork, indeed, and of 
the American vultures very nearly approaches what we 
should on a priori grounds regard as the original form of 
the syrinx. 

In the other direction the syrinx may be further increased 
in complication ; this is brought about by an hypertrophy of 
the intrinsic muscles. The simplest case is that of the 
plover, Vanellus cayenuensis, figured and described by 
GAEEOD,' whose figure is here reproduced. It will not need 

FIG. 45. SYRINX OF Yanclla cayennensis, FROM IN FRONT (AFTER GARROD). 

much description ; the principal change is in the enormously 
thickened pair of muscles. 

This modification of the syrinx, however, is seen at its 
extreme in certain passerines and in the parrots. Here we 
meet with a multiplication of the intrinsic muscles, which 
may exist to the number of three or four pairs. A syrinx of 
this kind, when found in the Passeres, is frequently termed 
oscinine, the group of Passeres exhibiting the character being 
the Oscines, a term, however, which is not now used in the 
classification of the group. 

1 'On the Trachea of the Tantalus locitlator and of Yanellus caijcnnensis,' 
P. 7.. .S. 1878, p. 625. 



In Menura superba, which has been described by GAEROD, 
there are three pairs of muscles, which are attached to 
different bronchia] semi-rings ; the posterior pair of muscles 
are attached to a ring below that which bears the insertion 
of the anterior pair, while the remaining middle pair are 
inserted higher up again. 

We now come to the consideration of those modifications 
of the syrinx which will be spoken of neither as degenerations 
nor complications. They are parallel modifications, which 


can in all cases be traced to the typical tracheo-bronchial 
syrinx, though whether they have originated from it is, of 
course, a matter of question. To the two varieties of syrinx 
which we briefly refer to here the names tracheal and 
bronchial syrinx have been given, implying the fact that 
the modification of the windpipe has taken place in the 
one case mainly or entirely in the trachea, and in the other 
mainly or entirely in the bronchi. 

The tracheal syrinx is distinctive of a group of Passeres 
which have been on that account called the Tracheophonse ; 



but a syrinx that presents some of the same characteristic 
modifications distinguishes the stork tribe. The principal 

modification of this type of 
syrinx is that a large number 
of tracheal rings are altered 
in character, in a way which 
will be pointed out in detail 
for one or two forms. The 
tracheal syrinx of the Passeres 
was investigated by JOHANNE s 
MiJLLER, 1 but GAEROD has 
added many new facts of 
importance to our knowledge 
of this kind of syrinx. 

In Hijlactes megapodius 
there are nine tracheal rings 
which are very much thinner 
than their predecessors ; on 
the anterior side, however, 
there are twenty-three which 
are thus altered. Two or three 
of the anterior bronchial 
semi-rings are modified and 
ossified ; from the second of 
these on each side is a ridge 
of cartilage, the processus 
vocalis, which extends up to 
the twelfth tracheal ring 
(from the bottom) . As a rule 
these tracheophone Passeres 
possess intrinsic muscles to 
FIG. 47 TRACHEA OF Tantalus locu- their syringes, but the rule is 

lator. a. FROM FRONT, b. FROM . w itv.rmi- pvr>pnfinn<5 of 
SIDE. (AFTEB GABHOD.) 11Ot wltno1 BXCept )11S, 

which Conopopliaga is one, 

as well as Hylactes, already described ; others, such as 
Furnarius, have the intrinsic muscles. The same form of 
syrinx seems to exist in the storks. In Tantalus loculator, 

1 In Abh. Berlin. Akad. 1845, p. 367. 





for instance (see fig. 47), the lowest seventy-eight rings of the 
trachea are modified through being thinner than those else- 
where, and this portion of the tube is of a greater calibre than 
that above. In Cicoina alba the lowest twenty-nine rings are 
thus changed in structure, and ' there is a small prolongation 
upwards of the lateral portions of the three lowermost 
tracheal rings, which forms a consolidated triangular process 
on each side, overlapping the next few rings and looking 
extremely like the rudiment of the similarly situated proces- 
sus vocales of the passerine 
tracheophone syrinx, which 
resemblance is increased by 
the thinness of the neigh- 
aiid by their 
from before 

The bronchial syrinx is 
seen in its most extreme de- 
velopment in Steatornis and 
in Crotophaga, where it was 
originally described by MUL- 
LER ; but other cuckoos and 
goatsuckers, as has been 
shown by me, 1 possess also 
a syrinx which may be 
called bronchial ; further- 
more, as WUNDEELICH has 
shown, 2 the ow r l tribe resem- 
ble the goatsuckers in this 

being flattened 

FIG. 48. SYRINX OF Steatornis. FRONT 

respect, while there are in- 
dications of the bronchial syrinx in certain petrels. 

The fullest description of the syrinx of Steatornis, which 
we take as a type of the perfectly formed bronchial syrinx, 

1 ' On the Syrinx and other Points in the Anatomy of the Gaprimulgidse,' 
P. Z. ,S'. 1886, p. 147 ; ' On the Structural Characters and Classification of the 
Cuckoos,' P. Z. S. 1888, p. 168. 

- 'Beitriige zur vergleichenden Anatomie und Entwickelungsgeschichte des 
unteren Kehlkopfs der Vf^el,' Nor. Act. Acad. Leop. Cces. 1884. 


is contained in a paper upon the general anatomy of this 
bird by GARROD. 1 From that paper we borrow the descrip- 
tion as well as the illustration. It will be seen from that 
drawing (fig. 48) that the trachea of the bird bifurcates, as 
does the trachea of a mammal, without any modification of 
the rings, either tracheal or bronchial. The latter are at first 
complete rings ; it is not until the thirteenth or fourteenth 
the exact position appears to vary that the syrinx appears ; 
here the rings cease to be complete rings, and are semi-rings- , 
their inner ends being completed by membrane, the mem- 
brana tympaniformis. To the first of these modified semi- 
rings is attached in the case of either bronchus the intrinsic 
muscle of the syrinx. 

The transition between this purely bronchial syrinx and 
the more usual tracheo-bronchial syrinx is afforded by 
various genera of cuckoos and goatsuckers (of which a par- 
ticular description will be found later), in which the mem- 
brana tympaniformis is placed, as in Steatornis, far down 
the bronchus, but which have also a sheet of membrane 
forming a continuation of the membrana upwards to the 
trachea, which is due to the non-closure internally of the 
earlier bronchial semi-rings ; this latter gets more and more 
limited in various genera until we have the purely tracheo- 
bronchial syrinx, in which the wide membrana tympani- 
formis commences at once on the bifurcation of the 

The syrinx has undoubtedly some value as a test of 
affinity. As to the Passeres, it is, as FURBRINGER has 
remarked, a ' classical ' object for the determination of 
relationships. In other families too it is of importance. From 
a more general standpoint, however, apparently but little 
reliance can be placed on the modifications of this so variable 
organ. An approximation to the reptilian condition in the 
absence of any special modification at the bifurcation of the 
bronchi is seen in some of the struthious birds and in the 
American vultures. It is not clear, however, that this 
simplicity is not a case of the reduction rather than of the 

1 ' On some Points in the Anatomy of ,S7w/omis,' P. Z. ,V. 1873, p. 526. 


retention of a primitive character. In special cases the 
form of the syrinx seems to be of not little value as a mark 
of affinity. The peculiar syrinx of the storks, for example, 
distinguishes them from their near allies the herons. The 
stork-like syrinx of Tantalus is one of the many reasons for 
placing it with that family. The peculiar form of syrinx 
termed the ' bronchial syrinx ' may seem to some to militate 
against the value of this organ as a test of affinity ; but, on 
consideration, the fact that both owls and goatsuckers 
possess it will not seem extraordinary in view of their other 
resemblances, while the cuckoos are perhaps not so widely 
remote from those two families as they have been placed. 
We may have here a clue to the relationship of these three 
groups of birds. The complicated (as regards musculature) 
syrinx of the parrots is so far an indication of affinity with 
certain Passeres. The systematic position of the parrots is 
by no means clearly defined, and therefore this indication 
of a possible affinity must not be ignored. 

The Lungs and Air Sacs. The lungs of birds never 
depend freely in the ccelom, as is the case with most 
reptiles. They are closely fixed to the parietes, and covered 
with a thin and transparent aponeurosis, which is the 
peritoneum. So closely are they adpressed to the body 
walls that when they are carefully removed by dissection an 
impress of the ribs is to be seen upon their dorsal and 
lateral surfaces. An approach to this peculiar position of 
the lungs in Aves is to be seen in the crocodilia and to a 
less extent in the Chelonia and Monitor lizards. In these 
animals the lungs are bound down to the parietes, and do 
not hang freely, as in the Lacertilia generally and in the 
mammalia. The lungs in birds occupy the space between 
the first rib in front and the anterior end of the kidney 
behind. They nearly meet in the middle line. Seen from 
below, when left undisturbed in the body, only the adjacent 
structures being cleared away, the lungs present two facets, 
an anterior and a posterior ; the latter is divided from the 
former by a ridge which does not divide the lung into two 
equal halves. The anterior is considerably the smaller. 


When the lung is thus bared it is seen to be provided 
with a number of conspicuous orifices where the covering 
aponeurosis is deficient ; these are termed the ostia, and 
they lead into the air sacs. Their number is variable in 
correspondence with the variability in the number of air 
sacs. The free surface of the lung is supplied with bands 
of muscle, which have been termed ' diaphragm/ but which 
are called by HUXLEY l costo-pulmonary muscles. These 
muscles arise from the ribs and spread out over the aponeu- 
rosis covering the lung ; they are, as a rule, extensive, 
extending but a short distance from their origin. The 
number of these costo-pulmonary muscles varies much 
among birds ; but little attention has been hitherto paid to 
them. For instance, among the storks the muscles in 
question are reduced to a minimum ; there are only two 
pairs at the anterior end of the lung, which arise not from 
ribs, but from the end of the windpipe. In herons, on the 
other hand, and in the emu, all the ribs bordering upon 
the lungs give off fasciculi of fibres, which in the emu are 
of considerable thickness. Each bronchus enters the lung 
at a little distance from its anterior end, and sometimes, as 
in the condor, the cartilaginous rings cease some little way 
before it enters. The bronchus dilates somewhat when it 
has entered the lung, and from the posterior end of this 
dilatation a tube is continued backwards, which opens into 
the posterior or abdominal air sac. This trunk is termed 
by HUXLEY the mesobronchium. Further on in its course 
the mesobronchium gives off another branch, which opens 
into the posterior intermediate air sacs (a description of the 
air sacs will be found a little further on). From the 
dilatation of the mesobronchium, the vestibule, arise four 
other tubes, which are called the entobronchia by HUXLEY 
(whose nomenclature is adopted throughout in the present 
section). The first curves forward and gives off several 
branches, one of which opens into the prsebronchial air sac, 
while the main trunk is continued into the subbronchial 
air sac. The second entobronchium passes dorsally and 

1 ' On the Respiratory Organs of Apteryx,' P. Z. S. 1882. 


ramifies, a wide branch descending to the subbroiichial 
ostium. The third entobronchium runs backwards and 
gives off a number of branches. Close to its origin from 
the bronchus it opens into the anterior intermediate air sac 
by the anterior intermediate ostium. The fourth ento- 
bronchium runs parallel with this ; it gives off branches 
from its ventral wall, but ends caecally. In addition to the 
entobronchia there are the ectobronchia. These are six or 
seven branches given off laterally and dorsally from the 
mesobronchium. These various bronchia are in communi- 
cation with each other, so that the substance of the lung is 
a meshwork. 

As a rule there are ten air sacs in birds, which are 
arranged in five pairs, five, in fact, arising from each lung. 
In front of the windpipe are the praebronchial air sacs ; 
below the trachea are the subbroiichial air sacs ; the oblique 
septa, which have been described elsewhere (p. 38), enclose, 
in the duck and all other birds except the Apteryx, two air 
sacs, the anterior and posterior intermediate sacs. The 
abdominal air sacs lie among the intestines, and are fed by 
an ostium which is at the extreme posterior end of the lung ; 
they have been, as HUXLEY has expressed it, pushed out 
from the space enclosed by the oblique septa like a hernia. 
In Apteryx, quite exceptionally, these air sacs are not so 
pushed out, but lie within the area enclosed by the oblique 
septa. The only differences that have been noticed in birds, 
apart from those that have been already mentioned, appear 
to consist in the number of the intermediate air sacs and in 
the condition of the praebronchial. WELDON ' has described 
the breaking up of the pripbronchial in the storks into a 
number of sacs, at least five in number, and the complete 
fusion of the subbroiichial sacs into a single one. The 
breaking up of the praebroiichial sacs is carried to a more 
complete extent in Chauna. In some birds there are three 
instead of two intermediate air sacs. I have observed this 
in Podargns. In many Accipitres the abdominal air sacs are 

1 ' On some Points in the Anatomy of PJicenicopterus and its Allies,' P. Z. S. 
1883, p. 640. 


peculiar in that one of the walls of the sacs has got to be 
firmly adherent to the ventral parietes, and the walls of the 
two sacs enclose between them the intestines, which have 
thus the appearance of being enclosed in a special compart- 
ment of the coelom. In one or two Accipitres there is the 
same subdivision of the intermediate air sacs that I have 
referred to in Podargus. 

These air sacs communicate with subsidiary spaces lying 
among the viscera, between the muscles, in the skin 
(particularly in Cliautia and several Steganopodes), and 
with the bones. The skull, however, is aerated by a set of 
spaces which are not connected with the trachea and lungs, 
but with the Eustachian tubes and the nasal chambers. 

The literature relating to the lungs and air sacs is large. 
In addition to the memoirs already quoted the following 
bear upon the matter : 

F. BIGNON, ' Sur les Cellules Aeriennes du Crane,' &c., 
C. R. Soc. Biol, 1887, p. 36. 

Idem, ' Recherches sur les Cavites Aeriennes Cervico- 
cephaliques chez les Psitacides,' Bull. Soc, Zool. France, 
xiii. 1888, p. 180. 

Idem, 'Note sur les Reservoirs Aeriens de TUrubu 
(Catliartes atra),' C. E'. Soc. Biol, (9), i. 1889, p. 39. 

Idem, ' Contribution a 1'Etude de la Pneumaticite,' &c., 
Mem. Soc. Zool. France, ii. 1889, p. 260. 

MILNE-EDWARDS, ' Observations sur 1'Appareil Respira- 
toire de quelques Oiseaux,' Ann. Sci. Nat. (5), iii. 1865, 
p. 137, and ibid, 1867, p. 12. 

Idem, ' Sur les Sacs Respiratoires du Calao rhinoceros,' 
C. E. 1885, p. 833. 

E. FICALBI, ' Alcune ricerche sulla struttura istologica 
delle sacche aerifere,' &c.,Atti Soc. Tosc. Sci. Nat. (vi. 1885, 
p. 249). 

H. FILHOL, ' Sur la Constitution du Diaphragme des 
Eudyptes; Bull. Soc. Philom. (7), vi. 1882, p. 235. 

N. GUILLOT, ' Memoire sur 1'Appareil de la Respiration 
dans les Oiseaux,' Ann. Sci. Nat. (3), v. 1846, p. 25. 

E. SELENKA, ' Beitrage zur Entwicklungsgeshcichte d. 


Luftsacke des Huhns,' Zeitsch. wiss. Zool. xvi. I860, p. 

H. STBASSER, ' Die Luftsacke der Vogel,' Morpli. J.13. iii. 
1877, p. 179. 

G. BOCHE, ' Prolongements Intra-abdorninaux des Beser- 
voirs Cervicaux chez 1'Autruche,' Bull. Soc. Philom. (8), i. 

1889, p. 111. 

Idem, ' Sur 1'Appareil Aerifere des Oiseaux,' ibid. (8), ii. 

1890, p. 5. 

Idem, ' Contribution a 1'Etude de 1'Anatomie Comparee 
des Reservoirs Aeriens d'Origine Pulmonaire chez les 
Oiseaux,' Ann. Sci. Nat. (1), xi. 1891. 

H. BOULAET, ' Note sur un Systeme particulier de Sacs 
Aeriens observes chez quelques Oiseaux,' Jonrn. de VAnat. 
et Plujs. xviii. 1882, p. 407. 

Idem, ' Note sur les Sacs Aeriens Cervicaux du Tantale,' 
Bull. Soc. Zool. Fr. 1885, p. 348. 

Other references are contained in the paper of HUXLEY 
upon the lungs and air sacs of Apteryx, already quoted. 

Muscular Anatomy 1 

The most general feature of the muscles of birds is the 
great length of their tendons of insertion ; the tendency of 

1 For memoirs dealing with the muscular anatomy of several types see 
CARUS, ' Erliiuterungstafeln zur vergl. Anat.,' Leipzig, 1826 (Astur, Falco, 
Cypse/its) ; P. HARTING, ' Observations sur 1'Etendue relative des Ailes et le 
Poids des Muscles Pectoraux,' &c., Arch. Nterl. Sci. Exact, et Nat. iv. 1809, 
p. 33 ; S. HAUGHTON, ' On the Comparative Myology of Certain Birds,' P. B. 
Irish Ac. 18157, p. 524 (Falco, Grus, Anas) ; G. JAEGER, ' Das Os humero-scapu- 
lare der Vogel,' J.B. K. Akad. Wiss. xxiii. 1857, p. 387; LEGAL and HEICHEL* 
' tJber die Beziehungen der Grosse der Flugmusk.,' etc., Her. schles. Ges. 
1879 ; H. J. MAGNUS, ' De musculis costarum sternique avium,' Diss. Inaug. 
Vratislaviae, 1867, and in Arch. f. Anat., 1869 ; H. PFEIFFER, ' Zur verglei- 
chendeii Anatomie des Schultergeriistes,' &c., Diss. Inaug. Giessen, 1854 ; 
J. J. PRECHTL, ' Untersuchungen iiber den Flug der Vogel, ' Q Vienna, 1846; 
QUENNERSTEDT, ' Studier i Foglarnas Anatomi,' Lund's Univ. Arsskr. ix. 1872, 
p. 4 ; N. KUDINGER, 'Die Muskeln der vorderen Extreniitaten,' &c., Nat. 1V/7/. 
Holland. Maatsch. Wet., 1868; C. J. SUNDEVALL, in A'. Vet. Akad. forh. 
1843, p. 303, and Furh. Skandin. Naturf. 1851 ; G. ALIX, 'Sur les Muscles 
Flechisseurs des Orteils,' &c., Bull. Soc. Philom. xi. 1874, p. 28, and Essai 
sur VApparell Locomotcur des Oiseaux. Paris, 1874. 


this is to mass the fleshy and heavy parts of the muscle 
about the centre of gravity of the body, a desideratum for a 
flying animal. This peculiarity of the muscular system is 
especially well seen in the muscles of the leg. The muscular 
system of birds is remarkably constant for the species, the 
number of variations being apparently, comparatively speak- 
ing, but slight. It is true that in but few cases has a large 
number of individuals been carefully dissected ; but of a 
good many species, on the other hand, have three, four, or 
even more individuals been dissected from the point of view 
of the relations or presence of a particular muscle or muscles. 
The muscular system too is apt to be very constant for a 
given genus or even a larger division. A glance at the 
systematic part of this work will show how trifling are the 
variations even between families in some cases. All these 
facts lead to the inference that the muscular system in birds 
is of very considerable value for classificatory purposes. 
GAEROD, FORBES, and FURBRINGER are the three anatomists 
who have laid greatest w T eight upon the muscular system as 
an index of affinity. It is, thinks Professor FURBRINGER, the 
muscles of the anterior extremity which have the greatest 
value of any part of the muscular system. The wing is an 
organ which is used in much the same way by all birds 
in which it is properly developed. On the other hand the 
uses of the muscles of the leg are manifold ; we have hopping 
birds, climbing birds, perching birds, swimming birds, &c. 
&c. Nevertheless GADOW is inclined to think (with GTARROD) 
that they are the most important. The existing knowledge 
of the muscles of birds is mainly confined to the muscles of 
the leg and of the fore limb, a knowledge which we owe 
almost entirely to GARROD and FURBRINGER, many other 
anatomists having, of course, filled up many details. Less is 
known about the muscles of the head, neck, trunk, and hyoid 


It is curious, indeed, how very few birds have been 
at all thoroughly dissected. Apart from the detailed account 
of Apteryx by Sir KICHARD OWEN, and of less comprehensive 
memoirs by COUES on the diver, by MORRISON WATSON 


on the penguin, we have only two recent memoirs which 
contain anything like a complete account of the muscular 
structure of a given type. These are the book upon Com us 
corax by SHUFELDT and a paper by CHALMERS MITCHELL 
and myself upon Palamedea. The most comprehensive 
general account of bird muscles is unquestionably that of 
GADOW in Bronn's ' Thierreich.' I shall base the following 
account of avian musculature largely upon the last-men- 
tioned work, adding to it only such details as were inacces- 
sible to GADOW at the time of its publication. The muscles 
known to vary will naturally be treated at greater length 
than those of whose comparative structure but little is 
registered. GADOW allows altogether 112 separate muscles 
and sets of muscles like those of the ribs, arranged in a 
serially homologous row. Some of these are, however, 
divided again. Of these, so far as we know at present, the 
following are of the greatest systematic importance, as present- 
ing really considerable variations even to disappearance : 

Glutceus maximus, gl. anterior. 

Obturator internus. 

Femorocaudal and accessory femorocaudal. 


Semitendinosus and accessory semitendinosus. 

Sleeps femoris. 


Flexo res prof undi hallucis et digitonim. 


Tibialis anticus. 

Pectoralis primus, p. secundus. 




An conceits. 

E.rjit/i/sor seen nda riorum. 

Cucullaris propatagia Us . 

The value of muscles in classification has been highly rated by 
many ornithologists, especially, of course, by GAEEOD, FOKBES, and 
FUBBRINGER. It is, however, only a comparatively small number 


of the total series of muscles in the body that can be trusted much 
as evidence of affinity. The ambiens is unquestionably of value as 
it is found or not found, as the case may be, through whole groups 
whose mode of progression when walking or climbing is as different 
as can be. Its total absence from all picarian and passerine birds 
is a fact upon which I comment elsewhere. There are very few 
groups in which the ambiens may be present or absent, and in 
those cases it is often reasonable to separate as distinct families 
the genera which have it from those which have it not. This 
cannot, perhaps, be done in every case. Some storks, such as 
Abdimia, have no ambiens, while the majority have it. There are 
auks with and auks without this muscle. The same may be said 
of petrels, parrots, and pigeons. Rlu/nclwps, the only larine bird 
without an ambiens, may be, perhaps, rightly elevated to the dis- 
tinction of a separate family. These examples, however, are so 
few that they may be compared to such singular exceptions as the 
absence of the odontophore in the nudibranch Doriopsis, which 
does not in the opinion of any one invalidate the great importance 
of that structure in arranging the mollusca. In estimating the 
value of the ambiens the facts of its total or apparently partial 
suppression, referred to below, must be borne in mind. The 
entire absence of all trace of the muscle in the owls shows that 
they are not necessarily to be placed in the neighbourhood of the 
parrots, in which the muscle, when absent, has left traces behind. 

Muscles of the Fore Limb 

Pectoralis Primus. This muscle consists of two parts, 
the thoracic part, arising from the sternum, and an abdominal 
portion, arising from the pelvis. The latter portion, well 
developed in lower vertebrates, is slight in birds, and is often 
completely absent. The pectoralis thoracicus arises from 
the sternum, the clavicles, and intermediate membranes ; it is 
inserted on to the humerus. In ratite birds there is no origin 
from clavicles, but, on the other hand, an origin from cora- 
coids not present in carinates. There is frequently an 
intimate connection between the pectoral near its insertion 
and the tendon of origin of biceps. The pectoralis is frequently 
divided into two portions, the mode of division being twofold. 
In Apteryx the coracoidal portion is separate from the sternal, 
a state of affairs which recalls some of the lower vertebrates ; 


in others the pectoral is divided into a superficial and deeper 
layer. In many ' Ciconiiformes ' this is the case. The 
pectoral, in all birds except the ratites, gives off one or two 
branches to the patagium. The branch has been termed 
the pectoralis propatagialix. There are either two separate 
muscles split off from the surface of the pectoral (as in Nisus), 
the tendon of one going to the tendon of the tensor patagii 
longus, that of the other to the tensor patagii brevis ; or 
(Podargits] there is but one muscle which divides into two 
tendons ; or the origins of the two tendons are separate, one 
of them commencing with a special muscle, the other arising 
as a tendon from the surface of the pectoralis ; or both may 
be tendinous in origin. Finally, there is in tinamous and 
gallinaceous birds (some) a special ' propatagialis posticus,' 
joining with its tendon that of the other muscle. The 
pectoralis abdominalis, totally absent in nearly all ratites, in 
storks, various hawks, &c., is divisible into two parts, of 
which one or other is sometimes wanting. The pars posterior 
springs from the pelvis and adjacent fascia ; it ends in front 
freely or comes into more or less close relations with the pars 
anterior. In Anseres, for example, the two form one con- 
tinuous band of muscle, their boundaries being simply marked 
by a slight tendinous inscription. The pars anterior arises 
from the skin close to the termination of the last, or is, as 
already mentioned, continuous with it ; it generally ends upon 
the humerus, near or in common with the insertion of the 
main part of the pectoral. In a few birds (quite remote 
from each other in the system, Pelecanus, Chauna, Cathartes) 
the terminal tendon is lost in the axillary region ; a more 
remarkable modification, possibly of classificatory importance, 
is described later in Crypturus. In a variety of birds there 
is a slip from this muscle to the metapatagium. 

Latissim/us Dorsi. This muscle is divided by FURBRIN- 
GEK into three sections - 

(1) L. d. (ulterior. 

(2) L. d. posterior. 

(3) L. d. metapatagialis mid dorso-cutaneus. 

The first-named muscle is totally wanting in Apterijx and 


Alcedo bengalensis. 1 It arises in other birds from the spinal 
processes of a varying number of cervical and dorsal vertebrae. 
The narrowest area of origin is seen in Alcedo, Maerochires, 
and various passerines, where it embraces at most a single 
vertebra. In other birds it may arise from as many as four 
and a half vertebrae. The broad fleshy or tendinous, or 
partly fleshy and partly tendinous, insertion varies in width. 

The second division is entirely wanting in Otis, Pterocles, 
many passerines, &c. It arises from the spinous processes of 
posterior dorsal vertebrae, ilium, and even adjacent ribs. Its 
origin is usually widely separated from that of anterior ; but 
there are various intermediate conditions which culminate in 
the cuckoos, Maerochires, and some other birds, where the 
two muscles form one, with, however (save in Cypselus), 
separate insertions. The insertion of this muscle is by a 
slender tendon in front of that of the latissiinus dorsi 

The latissimi dorsi metapatagialis and dorso-cittaneus 
are two slips running to the metapatagium and the neck 
region of the skin respectively. They are not often both 
present, but are in Apteryx, Charadridse, Alcidas, and some 
gallinaceous birds. They are both absent in ratites (excl. 
Apteryx), Maerochires, Colii, Bucerotidae, &c. The dorso- 
cutaneus is the rarest, and apart from the instances mentioned 
is found only in the Cracidae, piciform birds, and passerines. 

Cucullaris. This is an extensive muscle occupying the 
greater part of the neck. The only muscle superficial to it 
is the spliinctor colli. It has two main divisions. The 
pars cranialis arises from the region of the occipital and the 
squamosal ; in many birds (of the most diverse orders) a 
branch is given off from this (the dorso-cutaneus) , which 
ends on the back below the spinal pterylon, whose feathers 
it raises. The main part of the muscle ends upon the 
clavicle, or sternum, or ligaments in the neighbourhood. 
In some birds a part of the fibres end upon the fascia 
covering the pectoralis primus, and in those with a crop 
some of the deeper fibres come into relation with that 

1 Not in .1. ispidri. 


organ, forming a leva-tor in-fjluviei. A portion of the cucullaris 
also directs itself towards the patagium, and in most Passeres 
and in parrots, Pici, and Upnpa forms a special cucullaris 
propatagialis, joining the tensor longus tendon. The pars 
cervicalis of the cucullaris arises from the dorsal edge of the 
neck, and is inserted near or in common with the other part. 
In many birds (e.g. Anseres) a slip is given off from this 
which supplies the humeral pterylon. It is termed the cucul- 
laris dorso-cutaneus. 

Rliomboideus Externus. This muscle arises tendinously, 
the width of the tendinous part being about the same as 
that of the muscular part, from the last cervical and from 
the dorsal vertebra ; it is inserted fleshily along the whole 
length of the scapula. The muscle varies but slightly ; the 
origin is more or less extensive, and the vertebra from 
which it arises are not always the same. Its insertion is 
not always limited to the scapula, but sometimes extends on 
to the furcula. In Casuarius and Apterijx the muscle arises 
from the ribs. 

Elioniboideus Profundus. This muscle also arises ten- 
dinously from the neural spines of the dorsal vertebrae, or 
from both dorsal and cervical. Its origin sometimes extends 
as far back as to the ilium. It is inserted into the 
scapula below the last. In Casuarius and Apteryx this 
muscle arises from the ribs. The rhomboideus profundus is 
occasionally divided into two distinct parts ; in Megalama 
there are three distinct divisions. 

Serratus Super ficialis. This muscle is divided by FUB- 
BRIXGER into three parts, of which two are always present, 
while the third is sometimes absent. This is the -pars 
metapatagialis. The pars anterior arises from one or more 
ribs on the boundary line between the cervical and dorsal series. 
It is attached to the scapula along the ventral border, but, quite 
exceptionally, in Bliea on to the dorsal border. The pars 
posterior springs from a varying number of dorsal ribs, and 
in several birds (e.g. Rliampliastos) it, with the pars anterior, 
which can hardly be separated as a distinct muscle, springs 
from a considerable number of ribs five in the case referred to. 


It has generally a broadish insertion on to the scapula, but 
in Meiglyptes, many Passeres, c., it is attached merely to the 
extremity of that bone. The pars metapatagialis is absent 
in ratites (except Apteryx), hummingbirds, and a few others. 
It springs from 1-4 ribs, and is inserted on to the rneta- 

Serratux Profundus. This muscle is highly developed 
in Struthio and Casuarius, less so in other struthious birds 
and in the Carinates. It arises from a variable number of 
cervical and dorsal ribs, and it passes backwards (in the 
contrary direction, therefore, to the serratus superficialis) to 
be inserted on to the scapula. 

Patagialis. 1 This muscle, concerned with the folding of 
the patagial membrane, is present in all birds except the 
struthious. It arises from the clavicle and from the tip of 
the scapula, is sometimes divided into two muscles from the 
start, and sometimes arises as a single muscle, which 
immediately divides into two tendons, the tensor patagii 
longus and the tensor patagii brevis. Exceptionally the 
former may be absent. In a specimen of Crex pratensis 
the representative of tensor patagii longus was found by 
GAREOD to be simply the biceps slip, a muscle that will be 
treated of presently. The size and importance of this 
muscle vary considerably ; it is largest in the parrots, 
where, indeed, it is uncertain whether a part of the deltoid 
has not been converted to a similar function. This matter, 
however, is dealt with under the description of that family. 
The tensor patagii longus always ends in a single tendon 
which runs along the anterior margin of the patagium and is 
inserted on to the metacarpal. It is usual for the middle 
part of this tendon to be of a more fibroid character and of 
a yellowish colour, contrasting with the steely and typically 
tendinous aspect of the other tendons arising from the 
tensores. Very commonly the entire tensor muscle is 
reinforced by a tendinous or muscular slip from the pecto- 
ralis, and sometimes there is a separate slip to each of the 

1 G. BUCKET, 'Premiere Note suv 1'Appareil Tenseur,' etc., C. R, Soc. Biol. 
1888, p. 328. 


two tensores. It is also very general for one of the two 
tendons, or for both before their division, to be attached by a 
tendinous slip to the deltoid crest of the humerus. The 
tensores patagii are of considerable use in classification. 
But it must be admitted that they are apt to vary greatly 
from genus to germs. The variations chiefly concern the 
more or less complicated condition of the tendons of the 
brevis. The simplest condition is seen, e.g., in Rhamphastos 
Cuvieri, where the tendon is single and is attached below to 
the tendon of origin of the extensor metacarpi radialis. 

A further degree of complication is seen in, e.g., a cuckoo, 
where the single tendon gives off, near to the fore arm, a slip 
running obliquely wristwards, which is attached to the tendon 
of the extensor metacarpi radialis. In the limicolous birds the 
main tendon is usually divided from the first into two, of 
which the anterior has the wristward slip, already referred 
to ; in those birds and many others there is the further 
complication of a band of tendinous fibres which arise at the 
junction of the wristward slip with the fore arm, and pass 
obliquely forwards and upwards to be inserted on to the 
tendon of the longus. This slip is termed, in the following 
pages, the ' patagial fan ; ' it is frequently of a fanlike form. 
The tendon of the tensor patagii brevis has not always the 
regular form that it has in the types that have been already 
selected for illustration. In the tinamou, Eliyncliotus, for 
instance, the tendon is a diffuse fascia spreading out over 
the greater part of the patagial membrane ; in other birds, 
e.g. storks, it is a broad, rather diffuse band, as a rule with 
a thicker edge or edges. A peculiar condition of the 
tendons of the brevis characterises the auks, some gulls, and 
at least one limicolous bird. In them (see below) one or 
two delicate tendons arise from the longus tendon near to 
the insertion thereon of the patagial fan, and run obliquely 
backwards and downwards to be attached on to the dorsal 
surface of the fore arm the reverse side, that is to say, to 
that to which the other tendons which together make up 
the tensor patagii brevis are attached. 

In some birds, e.g. certain passerines, the tendon of the 



tensor longus is reinforced by a muscular slip ending in a 
tendon which is derived from the cucullaris muscle. Another 
muscle which is also related to the patagialis in an analogous 
way has been termed the ' biceps slip ' (q.v.) 

Anconceus Longus. This muscle arises from the neck of 
the scapula alone, by a head which is entirely tendinous or 
partly muscular, or in addition from the edge of the scapula, a 
little further away from its junction with the coracoid, by a 
tendinous head. It is inserted bv a broad tendon on to the 

7 ' P.L 

FIG. 49. TENSOKES PATAGII OF Phcswicoptenis. 

T.I'.L, tens', ir patagii longus ; T.l'.B, tensor patagii brevis : Bi, biceps 
.-lip: E. ('. A', extensor carpi radialis. (Alter VELDON ) 

elbow joint. Sometimes there is an accessory head from 
the humerus, which in this case arises, as a rule, in common 
with the tendon of insertion of the posterior latissimus 
dorsi. In Palamedea this head is double, the two halves 
being united by cross tendinous threads. On the whole the 
humeral head is characteristic of GAREOD'S Homalogonata* 
and not of the Anomalogonatae ; but there are exceptions on 
both sides. The breadth of the humeral head varies greatly ; 
it is sometimes reduced to a thin thread. 


Triceps. This muscle springs from the greater part of 
the humerus fleshily by two heads, of which one that from 
the tuberculum minus is often tendinous. The name 
triceps, be it observed, has been given to the muscle on the 
understanding that the last-described muscle is a part of it. 
Their tendons of insertion join. 

Expansor Secundarionun. 1 This extraordinary muscle 
appears to be partly a skin and partly a skeletal muscle. 
A bundle of non-striated fibres arises near the secondary 
feathers of the arm and ends in a tendon. This is occa- 
sionally reinforced by a band of striated fibres arising 
from beginning of ulna. The long tendon is inserted in 
various ways. The typical condition (termed by GAEEOD 
' ciconiine ') is for it to be inserted into the middle of a liga- 
ment running from the scapulo-coracoid to the sterno- 
coracoid articulation. Other modifications occur among the 
gallinaceous birds (q.v.), &c. The muscle is totally absent 
in Struthiones, Sphenisci, Alcse, Psittaci, the majority of Pico- 
Passeres, and in a few species of groups where it is usually 

Sterno-coracoideiis. This muscle, wanting only in the 
Macrochires, runs from the anterior lateral border of the 
sternum to the adjoining region of the coracoid. The muscle 
shows every stage between a single muscle and a completely 
double one. It is double, for example, in Casuarius. In 
Struthio, Chauna, and some other birds where there is but 
one sterno-coracoid, it is the homologue (according to FUR- 
BEINGER) of the deeper section of the double muscle. 

Scapulo-humeralis Anterior.- Buns from the beginning 
of the post-glenoidal region of the scapula to the beginning 
of the dorsal surface of the humerus. It is a muscle which is 
frequently absent. FURBRINGER failed to find it in Struthiones, 
Sphenisci, Fregata, Chauna, Columbae, Pterodes, Chwnga, 
Bucorvus, &c. 

Scapulo-humeralis Posterior. Contrary to the last this 
is a large muscle and is never absent. It arises from the 

1 GAREOD, 'On the Anatomy of Chauna di'i-Liana,' &c., P. Z. S. 1876 
p. 193, &c. 


hinder part of the scapula and runs to the humerus, where 
it is inserted on to the tuberculum mediale. 

Coraco-brachialis Extends. This muscle, which is 
relatively larger in the struthious birds, runs from the 
coracoid to the beginning of the ' planum bicipitale ' of the 
hurnerus, where it is generally covered by the tendon of the 
biceps. It is noteworthy that among carinate birds this 
muscle is largest in the tinamous, which thus approach the 
ostrich tribe. It is most reduced in the Passeres, in some 
of which, indeed, it has actually disappeared. 

Coraco-brachialis Interims. Springs from the coracoid 
and often from neighbouring parts of sternum. It is inserted 
on to the median tubercle of the humerus. 

Pectoralis II. This muscle arises from the ventral 
surface of the sternum, from the coracoid, and from 
the coraco-clavicular membrane. It is inserted by a long 
tendon of attachment to the lateral tubercle of the humerus. 
The muscle is small in ratites, large in carinates. 

Deltoides Major. Arises from the acromion and the 
dorsal part of the clavicle, and is inserted on to the deltoid 
crest of the humerus. The muscle and the length of its 
attachment vary much in size. It is large in Accipitres, 
Passeres, &c., small in Alcse, Psittaci, &c. It appears, indeed, 
to be absent in Psittacula. 

Deltoides Minor. This is a small muscle passing from 
the neighbourhood of the foramen triosseum. It is absent in 
Phacthornis and (occasionally) in Cypselus. It is also 
absent in Struthiones. 

Biceps. This muscle consists typically of two heads, as 
its name denotes. The longer of these arises from the 
coracoid by a long tendon. The second head arises also 
tendinously from the head of the humerus. The insertion 
of the muscle is double, on to the radius and the ulna. The 
division commences at a varying distance from the actual 

In the penguins this muscle is totally absent. In 
Colymbus, Pelecanoides, Thalassiarche, and other petrels, in 
some Alcidfe, the coracoid head alone is present, the humeral 


head being in some of these birds entirely diverted to form 
the biceps slip to the patagialis (q.v.) The coracoidal head, 
when it exists alone, may be divided into two quite separate 
muscles, uniting only at their very origin. This is the case 
with certain Alcidae. This division of the coracoidal half of 
the muscle also exists in the Laridae and in certain Limicolae, 
where there is a humeral head present also. In the Stega- 
nopodes both heads are present, but the humeral head after 
its attachment to humerus is continued on to the coracoid. 
A. trace of this arrangement is apparently left in some birds 
(e.g. Porpliyrio), where, though the humeral head arises from 
the humerus only, a ligament passes on from that part to 
the coracoid. 

Brachialis Inferior. A flat fleshy muscle arising from 
the distal part of the humerus, and inserted upon the ulna. 
In the penguins (where the biceps is absent) this muscle is 
particularly large, and is inserted on to the radius. 

PronatorSnblimis. This muscle springs tendinously from 
the inner condyle of the humerus and is inserted fleshily, 
and for a varying distance in various birds, upon the second 
third to second eighth of the radius. 

Pronator Pro/Hindus.- This muscle lies deep of the last, 
but has a similar origin and insertion. In the Katita3 this 
muscle and the last form a single muscle. 

Entepicondylo-ulnaris. This muscle, found apparently 
only in Galli and Tinami, arises in common with the pro- 
nator profundus, and is inserted on to the ulna. 

Ectepicondylo-radialis. This muscle arises tendinously 
from the outer condyle of the humerus and is inserted 
fleshily on to the radius. It appears to be wanting in the 
penguins, and to be largest in the Galli. 

Ectepicondylo-ulnaris. This muscle arises from the 
outer condyle of the humerus, and is inserted similarly to 
the last upon the ulna. Its insertion is fleshy, and in 
Palamedea it is larger than the last. 

Flexor Carpi Ulna-ris. This arises from the inner con- 
dyle of the humerus by a strong tendon, in which there is 
a well-marked sesamoid ; it runs down the inner side of the 


ulna to be inserted on to the great tuberosity of the ulnar 
carpal bone. A thin muscle arising from it passes into a 
tendon which is connected with the secondary feathers. 
The above refers to Palamcdea. In penguins the entire 
muscle is represented by a tendon only. 

Extensor Digitorum Corn-inn nis. Arises from the external 
condyle of humerus. It splits on the hand into two tendons, 
of which one is inserted on to the basis of the first phalanx 
of digit I., the other on to the corresponding phalanx of 
digit II. In Struthio the first of the two tendons is wanting. 
In the penguins the muscle is represented only by a tendon 
which is inserted on to the outer side of metacarpale II. and 
on to the basis of the first phalanx of that digit. 

Extensor Longus PoHicis. This muscle arises from the 
proximal region of both radius and ulna. The common 
tendon is inserted on to the origin of metacarpale I. 

Extensor Indicis Longus. This muscle is two-headed. 
The longer head arises from the radius from its middle two- 
thirds, bat sometimes also receives a few fibres from the 
ulna ; the second much shorter head springs either from the 
distal end of the radius or from the os carpi radiale, or 
finally from the basis of metacarpale II. ; the united tendons 
are inserted on to the head of the first and the basis of the 
second phalanx of digit II. The second head is absent in 
Fulica and in some other birds. 

Interosscus Dorsalis. This muscle arises fleshily from 
the opposed surfaces of metacarpals II. and III. The 
common tendon is inserted on to the base of the second 
phalanx of the second digit. 

Interosseus Palmar is. Has an origin from the same 
metacarpals as the last and is inserted on to the first phalanx 
of digit II. 

Ulni-metacarpalis Ventral is. It arises fleshily from the 
radial face of the last quarter of the ulna, and is inserted on 
to the head of the first metacarpal. 

Ulni-metacarpalis Dorsalis. SFhis springs by a tendon 
from the distal region of the ulna, and has an insertion upon 
the third metacarpal. 


Extensor Mc'tacar/ti Ulnaris. Springs from the external 
condyle of the humerus. There is generally a second head, 
which, instead of being tendinous, is fleshy and rises from 
the humerus a little below the first head. The tendons in 
which the two end do not join until a little before their 
insertion 011 to the base of the metacarpal of digit I. The 
degree of separation of the two heads differs considerably. 
In the penguins only one head is present. 

Flexor Digitoruin Sublimix. From the internal condyle 
of humerus to os carpi ulnare is a strong aponeurotic fascia, 
from the distal end of which springs the muscle in question, 
to be inserted on to phalanx I. of digit II. In Palamcdea Mr. 
MITCHELL and I traced the tendon to the base of the second 
phalanx of the same digit. In Psittacus and Columba the 
tendon has the same extension. In Struthio the muscle is 
entirely absent. 

Flexor Digitoruin Profundus. This springs from the 
middle and proximal third of the ulna, and is inserted on to 
the basis of the second phalanx of the second digit. In 
Corvus, &c., the muscle is two-headed, the two heads being 
separated by the insertion of the brachialis interims. In 
other birds the extent of the origin varies. 

Abductor Indicis. From metacarpal Il.tobasisof phalanx 
I. of digit II. In Palamcdea it also arises from the flexor 

Flexor Pollicis. From metacarpal I. to thumb phalanx. 
In Palamcdea its fleshy belly gives rise to a slip which passes 
to the abductor indicis. 

Adductor Pollicis. In Palamedca it arises from the meta- 
carpal just beyond the articulation of the thumb ; it ends 
in the ala spuria and not on the thumb bone. In some 
birds it has also a connection with the thumb bone. 

Extensor Pollicis Brevis. This muscle arises fleshily from 
the second metacarpal. It is inserted on to the phalanx of 
the thumb. 

Abductor Pollicis. This muscle arises from metacarpal I. 
and passes to phalanx of same digit. 


Flexor Digiti III. Arises from metacarpal III. and is 
inserted on to basis of phalanx I. of same digit. 

Flexor Metacarpi Radialis. This muscle arises from the 
outer condyle and is inserted on to the ulnar border of 
metacarpal II. or on to the beginning of metacarpal III. 

Muscles of the Hind Limb 

Sartorius. 1 This is a broad strap-shaped muscle arising 
from the ilium and from the fascia covering the glutaeus 
maximus ; it is inserted on to the ligament containing the 
patella, and on to the crest of the tibia. The muscle has an 
origin which sometimes extends further forwards, and is 
then overlapped by the latissimus dorsi. Sometimes, on 
the other hand, its insertion moves further back. In 
Phcenicopterus the muscle is divided into three distinct 

Glutceus Maximus? This often large muscle was used 
by GAEEOD 3 in his muscular classificatory scheme, and at 
first termed the tensor fascise. It has an origin which is 
sometimes entirely in front of the acetabulum, and some- 
times extends behind it. It arises tendinously from the 
fascia covering the glutaeus medius, and from the ridge of 
the ilium ; ite insertion is tendinous on to fascia covering 

GlutcEiis Anterior. 4 This muscle arises from the ridge of 
the ilium below the last, by which it is entirely covered ; it 
is inserted by a tendon on to the outer face of the thigh. 
The most remarkable modification which this muscle under- 
goes is its entire conversion into tendon in Bucorvus, &c., 
in which birds it comes to be merely a thigh ligament. 

Glutceus Medius. 5 This arises fleshily from the ilium, 
and is inserted by a short strong tendon on to head of 

1 ' Ilio-tibialis internus ' (GADOW). - ' Ilio-tibialis externus ' (GADOW). 

3 ' On certain Muscles of the Thigh of Birds,' Ac., P. Z. S. 1873, p. 626, and 
1874, p. 111. 

4 ' Ilio-femoralis externus ' (GADOW). 

^ ' Ilio-trochantericus medius et posterior ' (GADOW). 



Glutccns QuartuK. 1 This is a small muscle lying at its 
insertion between that just described and that about to be 

Glutccus Minimus.' 2 This muscle is also small, and 




arising from the ilium is inserted next to the last on to the 
neck of the femur. 

Pectineus.* This is a smallish muscle arising below the 

1 ' Ilio-trochantericus medius et posterior ' (GADOW). 
' Ilio-trochantericus anterior ' (GAI>\V). 
3 ' Ilio-femoralis internus' (GADOW). 


origin of the gluteeus quartus and inserted on to the inner 
face of the femur below the head. 

Vastus Externus. 1 This arises fleshily along the greater 
part of the outer side of the shaft of the femur ; it is fused 
with the crureus in front, and ends with it upon the patellar 

Crureus. 1 This is tendinous on the outer surface at its 
origin from the neck of the femur ; it also arises from a 
considerable part of the shaft of the femur and is inserted 
as has already been described. 

Vastus Interims. 1 This arises from the inner surface of 
the shaft of the femur and is inserted on to the tibia along 
with the tibial insertion of the sartorius. 

Obturator E.i-tcrnus? This is a small deep-lying muscle, 
which arises from the ischium or the ilium, or even partly 
from the pubis ; it is inserted on to the trochanter. 

Obturator Interims. 3 This muscle arises from the mem- 
brana obturatoria and from the ischium. It is inserted on 
to head of femur. GAKROD 4 laid some classificatory stress 
upon the shape of this muscle, oval or triangular. 

Gemelhis.* This is a small fleshy muscle, double or 
single, surrounding the tendon of insertion of the last. 

Semite)idinosus. t] T-n Palamedea thismuscle arises fleshily 
from the ilium behind the biceps ; it is half an inch broad, 
and, after being joined by the small accessory which springs 
from the femur near its distal end, sends a flat tendinous 
slip to the semimembraiiosus. The rest of the tendon of the 
muscle joins the middle head of the gastrocnemius. The 
muscle shows considerable variations in its attachments and 
size. It is completely absent in the owls, hawks, and swifts. 
The accessory head is absent in kingfishers, many Stegano- 
podes, Col/jinbn$, &c. 

1 ' Femori-tibialis ' (G.u>o\v). - ' Ischio-femoralis ' (GADOW). 

3 ' Obturator ' (GADOW). 

4 ' On the Anatomy of Chauna derbiana,' &c., P. Z. S. 187<>, p. 195. 

5 ' Accessorii musculi obturatoris ' ( 

6 ' Caud-ilio-fiexorius ' (GADOW). 


Femorocaudal .' This is another of the variable muscles 
of the thigh. Typically the muscle is two-headed, one head 
arising from the transverse processes of the caudal vertebrae, 
the other (termed by GARROD accessory femorocaudal) from 
the ilium. The two are inserted together upon the flexor 
side of the femur, as a rule by a longish tendon. 

The variations culminate in the entire absence of the 
muscle, which occurs in Chunga Burmeisteri and Leptoptilus. 
In some species for example, in most Passeres and picarian 
birds the caudal portion is alone present. In others e.g. in 
Serpentarius, Otis, Phcenicopterus, the iliac portion is alone 

Biceps Femoris?- -This muscle is covered externally by 
the glutseus maximus where this is present ; otherwise it is 
the most superficial of the flexors of the leg. It arises from 
the postacetabular region of the ilium, and ends in a strong, 
generally round, tendon, which passes through a sling of 
tendon which is derived from the femur independently, and 
from the same bone in common with one of the heads of the 
gastrocnemius, to be inserted on to the fibula. The principal 
variations of the muscle concern its more or less extensive 
origin. It has never been known to be absent. In Cory- 
thaix GADOW states that it is double. In the ostrich and in 
the ducks and swans there is the usual sling, but before 
entering the sling the biceps gives off a branch, which joins 
one of the heads of the gastrocnemius. In Fregata and in 
some swifts quite exceptional^ the sling is totally absent, 
but the muscle has the usual insertion. In certain auks the 
muscle gives off, before entering the biceps sling, a branch 
to the thigh superficially. In Podica seuegalensis the 
muscle divides into three branches. The first of these has 
a considerable superficial attachment to the outside of the 
leg ; the second is attached to the fibula below the attach- 
ment of the third insertion, which is the normal one, through 
a sling. In Heliornis only the first and third of these are 

1 ' Caud-ilio-femoralis ' (G.uxnv). z ' Ilio-fibularis ' (G.vnow). 


Semimembranosus, 1 Arises from the iscliium, sometimes 
trenching a little upon the pubis. It is inserted upon the 
tibia. The variations which it shows are mainly of size. 
In Plio&nicopterus it is two-headed, and in certain Falconidee 

i us 



completely double. It is often inserted in common with 
the semitendinosus. 

Adductors.' 1 - -These are, as a rule, two broad, flat, fleshy 
bands, which arise from the pubis, and from the iscliium, 
and are inserted upon the inner edge of the femur. The 
separation between the two parts is less in some birds than 

' Ischio-flexorius ' (GAM>\V). 

'-' ' Pub-ischio-femoralis ' (GADOW). 


in others. There is also occasionally (e.g. hornbills) an 
additional attachment to the gastrocnemius. 

Ambiens. This muscle, as is well known, is not present 
in all birds. Though the late Mr. GAREOD used it largely 
in his scheme of classification, its mere presence or absence 
is not an absolute guide to the systematic rank of the bird. 
Broadly speaking, it is present in all the birds which 
GAEEOD called homalogonatous, or normal-kneed, and 





1, rtexor longus hallucis : 2-4, flexores perforati ; 2', ",' tk-xores perforati et perforantes. 

it is absent in all the birds which were termed by him 
anomalogonatous, or abnormal-kneed. But there are excep- 
tions, at any rate on one side. Thus while the muscle is 
present in the storks generally it is absent from the nearly 
related herons, and, indeed, is absent in three storks, Xenu- 
rlnjnchus, Abdimia, and Dissura. When the muscle is 
present it has as a rule the relations described above ; but 
in a few birds it does not reach beyond the knee, thus 
showing, perhaps, an incipient disappearance. The import- 



ance of this muscle in classification has been much increased 
by MITCHELL'S interesting paper upon its exact relations 
to the flexors of the leg in a series of birds. 1 He has shown 
that in Baharica clirysopelargus the mass of muscle which 
forms the flexor perforatus arises from three distinct heads ; 
one of these is, in common with the flexor lorigus hallucis, 
from the intercondylar notch of the femur; the second is 

i.QLt/7; 2. 




Iff J? 



Fin. 53. LEG MUSCLES OF Opisthocomus (AFTER MITCHELL). 
II-IV, flexores in-rt'unii i. 

from the outer condyle of the same bone ; the third is from 
the tendon of the ambiens. This tendon divides into three, 
one for each of the three divisions of the flexor perforatus. 
The arrangement will be obvious from the accompanying 
cut. Apart from slight differences in detail the same 
arrangement was found to hold good for a few other birds 
provided with an ambiens. 

1 ' On the Perforated Flexor Muscles in some Birds,' P. Z. S. 1834, p. 495. 


In Nycticorax Garden i, which has no ambiens, there is a 
difference in the origin of the flexors in question which is of 
great interest. The two origins from the femur are as in the 
crane. But there being no ambiens there can be no origin 
from that muscle. Nevertheless the third head of the 
flexors is present in the shape (see fig. 53) of a broad ten- 
dinous band arising from the fibula, which soon divides into 
the three tendons to the three muscles, precisely as does the 
tendon of the ambiens. This is highly suggestive of the 
rudiment of an ambiens, a suggestion which is confirmed on 
referring to a specimen of Opisthocomus without a fully de- 
veloped ambiens (fig. 53). And as the herons are birds which 
are presumed to be really homalogonatous birds, though 
they have lost the ambiens, the fact is of additional interest ; 
particularly is this so when we compare the conditions 
obtaining in Nycticorax with those which characterise 
Corvus, a clearly anomalogonatous bird, none of whose near 
relatives possess an ambiens. In Corvus capellanus it was 
discovered that the flexors usually connected with the am- 
biens, or with its rudiment, had no origin from the fibula 
at all, and arose by only a single head from the femur. The 
same was practically the case with Bubo maximus, only 
that both femoral heads were present. Now the owls, 
formerly relegated to the Accipitres, are more generally 
looked upon as related to the picarian birds, forming, in 
fact, a section of the anomalogonatse of GAEEOD. The state 
of their ambiens is entirely confirmatory of this placing. 
We have some evidence, therefore, that there are degrees in 
the disappearance of the ambiens, which, so far as the few 
types that have been examined enable us to say, distinctly 
support the division of birds into the two great divisions of 

Peroneus Superficialis. Confined as a rule to the tibia 
in its origin, this muscle sometimes springs also from the 
fibula. The tendon of insertion, after giving off a branch 
to the tarsus, becomes attached to the tendon of the flexor 
perforatus digiti III. This muscle is occasionally completely 
absent ; this is the case with various Picopasseres, owls, &c. 



In Podiceps, according to GADOW, there is no branch to the 
flexor tendon, the tendon of insertion -ending at the ankle. 

Peroneus Profundus. This muscle arises from the lower 
part of the tibia, and is inserted on to the outer side of the 
ankle. It is completely absent in such birds as Ciconia, 
Otis, &.c. 

Gastrocnemius. This great muscle occupies the greater 
part of the back of the leg. It has three heads, of which 
the outer arises from the outer condyle of the femur, and 
from the ligament which supports the insertion of the 
biceps ; the inner head springs from the inner side of the 
head and neck of the tibia ; the middle head is confused at its 
origin with the insertion of the accessory femoro-caudal. 
This middle head appears to be wanting in Cypselus. It is 
inserted by a strong tendon to the tarso-metatarsus, and 
also, dividing, to the phalanges of the toes. 

Poplitei(s.~-This muscle passes between the tibia and 
the fibula in most birds, but has been found to be sometimes 
absent (e.g. Picus, &c.) 

TibiaMs Anticus. This muscle arises by two distinct 
heads. The first is entirely tendinous, and is from the 
external condyle of the femur. It forms a long and strong 
ligament, which runs over the knee ; the second head is 
fleshy, and is from the front part of the head of the tibia. 
Its long tendon of insertion is attached to the metatarsal. 
In certain birds (Chrysotis, Podargus, and owls) the tendon 
and even a portion of the muscle are double. 

Extensor Digitorum Communis. This muscle arises gene- 
rally from the front part of the tibia only, but sometimes its 
fibres of origin stray on to the patella and on to the fibula. 
The divisions of the tendon of the muscle are usually inserted 
on to several phalanges of the toes, which they supply. It 
is remarkable that the parrots are the only birds in which this 
muscle supplies the hallux as well as the other toes ; it is, 
therefore, in them, as GADOW remarks, truly an extensor 
commutiis. In other birds digits I., III., IV. (when present) 
are the only toes supplied. The common tendon divides in 
various ways ; in Grus virgo the tendon divides into two, 


and each of these again divides ; there are thus four tendons, 
of which the two middle ones supply the third toe. In 
PtilonorJii/neJii/x riolacciiH the tendon first gives off a branch 
to the second toe, and then divides for the third and fourth. 
In tthaniphastos carinatitx all three branches are given off at 
the same level. In Pharomacrus mocinno the extensor sup- 
plies only the two middle digits. In Scopus there is a slight 
variation of what is found in Gnis, Nothura, &c. The 
tendon divides into three, and the middle one again divides 
into two, both of which latter supply the middle digit. 
The superficial flexors of the foot consist of- 
Flexor Perforatus et Perforans Indicts (fig. 51). Arising 
from the outer condyle of the femur, and from the septum 
between itself and adjacent muscles ; its tendon is inserted 
on to base of second phalanx of its digit. 

Flexor Perforatus et Perforans Medii. Has two heads 
of origin, one as in last, the other from fibula. Its tendon 
perforates that of flexor perforates of the same digit, and is 
perforated by flexor profiuulns ; it is inserted on to base of 
second phalanx (in Palamedea, third in Ciconia} of its 

Flexor Perforatus. This muscle (see figs. 5'2, 53) arises 
from two heads, an inner head from the intercondylar notch 
and an outer head from the outer condyle of the femur. As 
will be seen from the annexed cuts, those portions of the 
muscle which supply digits III. and IV. have slips from both 
heads, but not that which supplies digit II. These muscles 
also arise either from the ambiens or from the fibula, as has 
been explained above under the description of the ambiens. 
The tendon to digit II. is inserted at the base of the first 
phalanx, that to digit III. is usually joined by a vinculum 
(absent in Opisthocomus, As to otus, and Ehijt idiceros plicat tin) 
to tendon of flexor perforates et perforans medii, and is 
inserted on to base of second phalanx of digit III. The third 
tendon has four slips of insertion, on to four proximal 
phalanges of digit IV. The descriptions of the insertion of 
these tendons applies to Ciconia iiigra. There are varia- 




Flexor Profundus. Arises from nearly whole hinder 
surface of tibia and fibula, and sometimes also by a head 
from the outer condyle of femur. 

Flexor Hallucis. Arises by a single head, or by two 
heads, from the outer- condyle of femur and from inter - 
condylar region. 

The tendons of the two last-described muscles are con- 


FIG. 54. Gallus bankiva (AFTER FIG. 55 Apteryx Mantelli (AFTER 

nected with each other in various ways, which have been 
described and illustrated by Mr. GARBOD.' In using his 
figures I accept the seven types admitted by Dr. GADOW. 2 

(1) In gallinaceous birds, pigeons, parrots, storks, &c., the 
tendons cross, and are united by a simple vinculum (fig. 54). 

1 ' On the Disposition of the Deep Plantar Tendons in different Birds, v 
P. Z. S. 1875, p. 339. 

2 In Newton's Dictionary of Birds, p. 617. 



(2) In Apteryx, &c., the vinculum is very strong, and 
forms the direct continuation of the tendon of the flexor 
hallucis ; the tendon to the hallux has the appearance of 
being a branch of this (fig. 55). 

(3) In many Accipitres the flexor hallucis divides into two 
parts at the lower end of metatarsus ; one of these goes to hal- 
lux, the other blends with the branch of the flexor communis 

FIG. 5G. Tinnimculus 
alaudarius (AFTER GARROI>). 

Fia. 57. Buceros rhinoceros 

which supplies digit II. ; there may be in addition a strong 
vinculum, uniting tendons before their splitting (fig. 56). 

(4) In Ehca, Phcenicopterus, &c., where hallux is small 
or absent, the two tendons are present, but fuse together, 
branching later to supply the digits present. 

(5) In Buceros, Podargus, Sarcorhamphus, &c., the two 
tendons fuse completely before supplying the digits (fig. 57). 

(6) In Megaleema, Rlmmphastidce, &c., the vinculum 
is present, but flexor digitorum supplies only digit III., 


the others being supplied by trifurcate flexor hallucis 
(fig. 58). 

(7) Both tendons present, but no vinculum at all in 
Passeres (fig. 59). 

Of these seven types there are naturally modifications 
and intermediate conditions, which will be described in the 
systematic part below. 

Short Flexors. Of these there are two series ; one set 



FIG. 58. Mcgalcsma (AFTEK 


arises from the bones of the metatarsus, the others from the 
long tendons. We shall commence with the first set. 

The flexor hallucis brevis is often (Palamedea, Aceros) 
composed of two distinct muscles with separate tendons 
and insertions on to the first digit. 

Flexor (adductor) dirjiti II. is inserted on to the median 
side of the base of first phalanx of digit II. 

Flexor (adductor) digiti IV. arises near to the last, and 
has a similar relation to the basal phalanx of digit IV. 


Flexor secundux (adductor) digiti IV. is a muscle which 
appears to be generally absent (e.g. Ciconia), and is not, 
according to GADOW, mentioned by authors. It is present 
in Rliea and Bucorvus. 

The last-named muscle may be the equivalent of the 
first of the two short flexors which arise from the deep long 
flexor tendons. Of these there are two. 

(1) MITCHELL has described in Opisthocomus a muscular 
slip which leaves the longus hallucis tendon and runs to the 
fourth digit. A similar muscle is present in Ardea cinerea ; 1 
I have found it in the hornbill, Ceratogymna elata (not in 
Acer os nipalensis) . 

(2) Another muscle (flexor brevis digiti III. of GADOW) 
arises from tendon of flexor prof undus in Opisthocomus, Rhea, 
<kc., and passes to third digit.' 2 

Of short extensor muscles there are at most six. 

The extensor hallucis, generally single, is two-headed 
in Pandion, and formed of two distinct muscles in Pala- 

The extensor proprius and extensor brevis digiti III. both 
supply the third digit. In Aceros and Bucorvus a single but 
two-headed muscle appears to represent both. 

The extensor (adductor) digiti II. is not always present. 

The same may be said of the extensor digiti IV. 

Muscles of the Neck and Trunk 

These muscles are, many of them, not easy to isolate and 
describe. There is, in consequence of this, some divergence 
in the published accounts. Furthermore, insufficient data 
have been collected for the estimation of the use of these 
muscles in classification. The account which follows is an 
almost verbatim transcript from a paper upon Palamedea 3 
by Mr. MITCHELL and myself. 

' It is suggested by MITCHELL that these muscles (which require further 
study) may ' throw light upon the origin of the very peculiar modes of distribu- 
tion of the hallucis tendon in some groups of birds, as it has been repeatedly 
shown that a tendon may be the homologue of a muscle.' 

- GARROP MS. 3 Quoted on p. 108. 


Biventer Cervicis. The two muscles are perfectly separate 
from each other. They arise tendinous from the spiuous 
process of the first dorsal vertebra. Then follow a tendon 
of an inch long, a belly of two inches, again a tendon of four 
inches, then another muscular belly of one and a half inch, 
which is inserted fleshy on to the occipital below the coin- 
plexus. In some kingfishers (q.v.) the two muscles are 
joined by a tendinous limb. 

Complexus. This muscle arises from the transverse pro- 
cesses of the third and fourth cervical vertebra, and from 
the fibres covering the inter trahsversarii of the same. It 
is inserted, separated from its fellow by a septum, on to the 
transverse ridge of the occipital. The muscle is entirely 

Longissimus Dorsi. It arises by a series of fleshy fibres 
from the front edge of the ilium, becomes tendinous in the 
middle, and then is inserted by fleshy fibres on to the lateral 
surface of the vertebral spine next in front ; the next anterior 
part arises tendinously from the spinous process of the most 
posterior uncovered dorsal vertebra, and is inserted on to 
the vertebra next in front : then follow two of precisely 
similar relations ; the next is carried on to the dorsal surface 
of the longissimus dorsi, as also is the last or most anterior 

Ilio-costalis. This complex muscle' lies laterally to the 
foregoing muscle ; it is fused at the edge with its fibres. It 
arises from the ilium and from the transverse process beside 
the attachment of the rib ; two similar slips in front of this 
arise from the transverse process and from the adjacent 
surface of the rib. The ends of the slips are inserted partly 
on to the surface of the ribs and partly pass on to the lateral 
musculature of the neck. 

Cervicalis Ascendens. This is the lateral muscle anterior 
to the ilio-costalis. It consists of five distinct slips arising 
from the transverse processes of vertebrae xvi.-xi., with the 
exception of xn.- The two posterior are inserted on to 
the vertebrae next in front ; the next two are inserted on to 
the surface of the oblique muscles next in front ; the last 


one on to the oblique muscle next but one in front. Behind 
these slips, which were obvious, there were indications of 
additional slips both in front and behind, but these were 
not sufficiently differentiated from the adjacent muscles for 
separate description. 

Lung its Cera ids. This median muscle arises from the 
forward continuation of the longissimus dorsi and from the 
median underlying part of the spinalis complex. 

Spinalis Complex. This system of muscles lies deeper 
than the foregoing. It is divisible into three parts. Part I. 
(sometimes called the spinalis dorsi) arises apparently only 
from the longissimus dorsi ; it gives off six fleshy bellies 
which increase in length from the posterior to the anterior ; 
they are inserted on to the upper posterior surface of the 
oblique processes of cervicals x.-xvi. In addition the 
superior fibres from these heads form a well-marked rounded 
muscular cord, which runs forward to form the longus colli 
posticits. Part II. consists of only four well-differentiated 
slender bellies ; these arise from the spinous processes of 
cervicals xm.-xv., and they are inserted on to a continuous 
longitudinal band, the posterior part of which sends slips 
to the three posterior branches of the spinalis dorsi, while 
the anterior end is inserted on to the oblique processes of 
cervicals x., XL, at the roots of the anterior two spinalis 
dorsi bellies. Part III. (longus colli posticits) arises from the 
sides of the spinous processes of cervicals II.-XL, and from 
part I. of the spinalis complex ; it is inserted by digitations 
which merge with the intervertebral muscles in front of its 
origins. It has been specially described and figured by 
GARROD for Plot its. 

Eectits Cap it is Posticits. It arises from the spinous 
process of atlas and axis ; its fibres spread out over the 
occipital under the complexus. 

Intertransver sales. These muscles are obvious all the 
way along from the ilium to the neck, running between the 
transverse processes of the vertebrae. 

Obliqui (Transverso-spinales}. They are clearly differen- 
tiated only from the last to the seventh cervical. They are 


large fleshy digitations arising from the transverse processes, 
and inserted on to the lateral face of the spinous processes 
next but one in front. 

Rectus Capitis Anticus Major. It arises all along the 
neck from the hypapophyses and from fascia ; about the 
middle of the neck it grades into the long us colli, from a slip 
of which it first arises about the level of the seventh vertebra. 
Its broad fleshy insertion is tendinous on the outside, is 
fused with its fellow in the middle line, and extends for 
about a quarter of an inch on the anterior outer edge of the 

Rectus Capitis Anticus Minor. This is a fleshy broad 
muscle underlying the preceding. Its origin is fleshy and 
continuous from first four vertebra?. It has a broad fleshy 
insertion to the extreme outer posterior face of the ridge 
behind the meatus auditorius. 

Long us Colli. It arises from the middle of the centrum 
of the second dorsal vertebra tendinously, and then by a 
series of tendons from each vertebra up to the overlap of the 
rectus capitis. It is inserted by a series of slips to the 
vertebrae in front of its origins. 

Intertuberculares. These are a series of short muscles 
forming the deepest layer of the neck musculature. 

Interappendiculares Costarum. The first arises from the 
end of the last free rib, and runs backwards and downwards 
to the lateral anterior process of the sternum ; the second 
from the junction of the sternal and costal parts of the first 
complete rib ; it shortly fuses with the third, which arises 
from the costal part of the next rib. These two are then 
inserted together. The fourth arises from, the third, fourth, 
and fifth costal ribs and from the space between them, and 
is inserted immediately behind the others. The posterior 
ones are smaller. 

Intercostales Externi. These are confined to the whole 
of the costal part ; the fibres run from above in front and 
downwards towards the caudal end. 

Inter costales Inter ni. These are confined to the lower 
half of the costal ribs, and are chiefly tendinous. 



Costi-sternales. Four slips arising tendinously from the 
sternal ribs, and inserted fleshy to the sternum. 

Costo-sternalis Extern us. This peculiar muscle, appa- 
rently found only in Palamedeidse, replaces physiologically 
the uncinate processes, as its broad ribbon-like belly runs 
diagonally across the outer surface of the ribs. It arises by 
a very thin flat tendon from the third, fourth, and fifth ribs, 
and from the interspaces 
between them. It is inserted 
to the costal edge of the 
sternum half an inch from the 
posterior end. 

Caudal Muscles 

The caudal muscles (in 
Palamedea) are illustrated in 
the accompanying figure. 

Levator Coccygis. This 
arises on each side from ilium, 
from lateral faces of spinous 
processes, and from trans- 
verse processes of caudal 
vertebrae. It is inserted on 
to membrane covering rec- 

Ilio-coccygeus. On each 
side there are two parts of 
this muscle, bol;h entirely 
fleshy. They arise from the FIG. 60. CAUDAL MUSCLES OF I'uiu- 
ilium and the ilio-sacral liga- ntfdea (AFTEB BEDI)ABr> AND MlTCHELL )- 
ment, and are inserted on to outer rectrix. 

Pubo-coccyyeiis Externus. This is the most posterior of 
the muscles of the tail. It arises from the pubis and is 
inserted on to external rectrix. 

Pubo-coccygeiis Interims. Arises in front of and below 
the last-mentioned. The origin extends also on to the 
ischium. The muscle is inserted on to last one or two caudal 


Depressor Coccyyis. This springs from the transverse 
process of the last sacral vertebra, and from the adjacent 
surface of the ilio-sacral ligament. It is inserted on to the 
transverse processes of the last three or four caudal vertebrae. 

Abdominal Muscles 

ObliquiLs Abdominis Externns. The muscle arises from 
the ribs and from their uncinate processes. It ends by an 
aponeurosis upon the pubis. 

Obliquus Abdominis Interims. Lies between the last 
muscle and the next. It passes from the pubis, extending 
on to ilium to the last true rib. A separate slip of this is 
described as the quadratus lumborum, running from the last 
false rib to the crest of the ilium. 

Transversus Abdominis. This is the deepest of the ab- 
dominal muscles. It springs from the pubis and preace- 
tabular ilium, and its aponeurosis ends in that of its fellow 
in the linea alba. 

Eectus Abdominis. Springs from last sternal rib and 
from sternum, and is attached to pubis. 

Transverso-analis. This passes across the abdomen in 
front of the cloacal aperture, and meets its fellow. It arises 
either from the pelvis or from the transverse processes of 
certain caudal vertebrae. 

Hyoidean Muscles 

Our knowledge of the muscles of the hyoidean apparatus 
and the neighbourhood is chiefly due to GiEBEL, 1 GADow. 2 
and MITCHELL. 3 I mainly follow the latter in his account 
:)f these muscles in Opisthocomus and in Palamedea. 


' Die Zunge cler Vogel,' &c., Zeitschr. f. d. ges. Katiirwiss. xi. 1858, p. 19. 
- BRONN'S Thierreich, ' Aves.' 

' A Contribution to the Anatomy of the Hoatzin (Opisthocomus cristatus),' 
P. Z. ,S'. 1816, p. 618 ; BEDUAHD and MITCHELL, ' On the Anatomy of Palamedea 
cornuta; P. Z. ,S'. 1894, p. 536. See also G. L. DUVKHNOY, ' Memoire sur 
quelques Particularites cles Organes cle la Deglutition de la Classe des Oiseaux,' 
Ac., M6m. Soc. Hist. Nat. Strasbourg, ii. 1835 ; J. KACZANDER, ' Beitriige zur 
Entwicklungsgeschichte cler Kaumuskulatur,' Mtli. Embr. last. Wien, 1883. 



The m.ylolujoid anterior is a sheet of muscles passing 
across between the rami of the mandible anteriorly. 

The iiujlolujoid posterior is composed of two layers, a 
deeper and a more superficial ; the latter is a broad sheet of 
muscle nearly reaching the mylohyoid anterior in front and 


1. _'. geniciliyoid : 3, 4, 5, mylobyoid ; 6, ceratoglossus ; 7, ceratohyoiil : 8, depressor 
inandilnilse ; Ol, mamlibular glaarl. 

covering the space between the rami of the mandible in the 
region where it is developed ; the deeper layer arises in 
common with the superficial layer from the rami ; it is a 
narrower muscle and runs forward. 


The genioliyoid is in two distinct portions ; the posterior 
division arises on each side from a ramus of the mandible ; 
it passes backwards and is wrapped round the ceratohyoid 
to its tip. The anterior portion also arises from the rami of 
the jaw ; it is inserted upon the ceratohyal partly under and 
partly distally to the insertion of the posterior division. 
This muscle is sometimes a single muscle, as, for example, 
in Palamedea. 

The genioglossus (entirely absent in Palamedea, &c.) is a 
slight muscle, springs from the middle line near the chin, 
and passes to the os entoglossum. 

The ceratoglossus is a strong muscle, arises from the outer 
side of the ceratohyal ; it ends in a tendon which is inserted 
along the side of the tongue almost to the tip. The muscle 
in Palamedea, &c., is divided into two parts. 

The ceratohyoid arises from the inner side of the cerato- 
hyal and is inserted on to the urohyal. 

The sternoJnjoid (in Palamedea) is a band of muscle which 
arises from the basihyal and entoglossus ; it spreads out 
over the thyroid cartilage and trachea. 

The liypoglossals are in Palamedea indistinguishable from 
the ceratoglossus. 

Muscles of the Head 

The temporalis (in Palamedea) is divided into two 
portions ; the superficial part arises from the whole temporal 
fossa and from the external and internal surfaces of the 
midtemporal process ; it runs to the outer upper surface of 
the lower jaw ; the deeper part is a pyramidal muscle ending 
in a stout tendon attached to the lower jaw ; in addition to 
this another portion, deeper still, runs across from the forward 
process of the quadrate to the inner side of the ramus, and 
a wide band of muscle bridges the interval between the inner 
edges of the forward process of the quadrate and the wall of 
the orbit behind the optic foramen. 

The pterygoid is divided into several layers which con- 
nect the lower jaw with the palatines and pterygoids. 


The depressor mandibiihc consists of two parts which 
have been termed digastric and hiventer. It connects the 
under surface of the occiput with the lower jaw. 

Osteology ! 

Vertebral Column. Highly characteristic of birds is the 
saddle-shaped (' heterocoelous ') form of the vertebral centra. 
The existence of this mode of articulation, though confined 
to birds, is not found everywhere and in all of the existing 
members of the order, and is not found at all in some of the 
extinct forms, Archceopteryx and Ichthyornis. As to exist- 
ing birds, the opisthocoslous form is frequently met with, but 
so irregularly as not to be of much use from a classificatory 
point of view. The existence of such vertebrae, pointing 
towards reptiles, may, however, be accepted as some indica- 
tion of an archaic position in the order. The matter has 
been lately summed up by the late Professor PARKER, 2 from 
his own investigations, and from those of others. 

The dorsal vertebrae have been found to be opisthocoolous 
in penguins and auks, in Limicolae (including Laridse), but 
not in petrels ; among the parrots PARKER met with this 
condition in several forms, where it was found to be com- 
bined with terminal epiphyses, a character which is just 
commencing in the lowest mammalia (OniitJwrJtynchus) and 

1 The following are a few works which deal with the general osteology of a 
number of forms. More special treatises will be referred to in their proper place : 
E. BLANCHARD, ' Recherches sur les Caracteres Osteologiques des Oiseanx,' 
&c., Ann. Sci. Nat. xi. 1860, p. 11 ; J. F. BRANDT, ' Beitrage zur Kenntniss der 
Naturgeschichte d. Vogel,' &c., Mem. Acacl. Sci. St. Petersb. (6), iii. 1840 ; 
EYTON, Osteologia Aviuin, London, 1858-81 ; A. B. MEYER, Abbilchtngen 
von Vogelskeleten, Dresden, 1879-96 ; MILNE EDWARDS and GRANDIDIER, Hist. 
Nat., d~c., de Madagascar, ' Oiseanx,' Paris, 1879-85 ; C. L. NITZSCH, Osteogra- 
jiltixclie Beitriicje, etc., Leipsic, 1811 ; P. GERVAIS, 'Description Osteologique de 
1'Hoazin, du Kamiehi,' &c., Zool. in Voy. de Castdnau, Paris, 1855. Besides 
FURBBINGEH, UntersucJiuiigcn zur Morphologic r. Syst. d. Vogel, and GADOW, 
4 Aves,' in Bronn's Klasscnv. Ordniingen dcs Thicrreichs, for brief and largely 
osteological definitions of birds see SEEBOHM, Classification of Birds, London, 
1889, and SHARPE, Osteological Catalogue of College of Surgeons 
London, 1891. 

2 ' On the Vertebral Chain of Birds,' Proc. Boy. S<>c. xliii. 1888, p. 465. 


entirely exceptional among birds. Steatornis is another bird 
with opisthocoelous vertebrae. So too are the cormorants and 
darters. Another form of vertebral articulation met with 
in reptiles namely, the procoelous articulation also exists 
in birds. In all birds, of course, the atlas is procoelous, 
articulating with the convex occipital condyle. In many 
birds ' the last two movable joints in the caudal series 
become proccelous.' 

The biconcave form of vertebra? characterises the extinct 
Archceopteryx, 1 and the gull-like Iclithyornis, called by its 
name on this very account. The concavities, however, 
according to FURBRINGER, seem rather to have been produced 
by the maceration out of a plug of cartilage than to have 
characterised the unaltered vertebra. It is doubtful, in fact, 
whether the spaces left in the fossil vertebrae were filled 
during life with copious remains of the notochord, as in 

The vertebral column of birds can be distinguished into 
four series, as in the higher vertebrates generally. It is 
customary to regard as cervical those vertebrae which either 
have no movable ribs or, if they have, do not become con- 
nected through their intermediary with the sternum. The 
rib-bearing vertebra are the dorsal series, while those which 
articulate with the pelvis are usually termed sacral. But 
it seems better to reserve the term ' sacral ' for the two 
vertebras which, in the chick, bear the ilium. 

The number of true sacrals is not, however, always two. 
J. J. PARKER 2 describes in the young Apteryjc three vertebrae, 
which abut upon the ilium, and are the only ones in which 
there are separate rib-like ossifications at the ends of the 
transverse processes. These vertebrae, which are quite con- 
spicuous in the adult (fig. 62), are regarded as the true 
sacrals. There are also three in the ostrich (fig. 63). In 
other birds (e.g. Larus, Chioms) there is apparently only 
one sacral vertebra. 

Behind the sacrum are the caudal vertebrae. Arclmo- 
pteryx is unique among birds for its long tail, composed of 

1 Not certainly. - ' Development of Aptcryx,' Phil. Trans. 1891. 



separate vertebra*. In all other birds the tail is short and does 
not extend far beyond the sacrum. In the majority of carinate 
birds the terminal vertebrae are fused together into the highly 
characteristic ploughshare bone (urostyle or pygostyle). 1 


il, ilium ; />, pubis ; i, ischium ; lp, prepubie process. 

There is, however, a closer correspondence between 
the tail of Arcliaopteryx and that of the carinate bird 


8, 9, in, sacral vertebras ; p, parapophyses ; rf, diapophyses. 

than might be assumed from the last-mentioned differ- 
ences. The first four caudal vertebrae of Archceopteryx have 
strong transverse processes, which are weaker, but present, 
on the fifth, which thus affords a transition to the remaining 
sixteen, upon which there are no such processes. In the 

1 W. MARSHALL, ' Untersuchungen liber den Vogelschwanz,' Ned. Arch. f. 
Zool. i. 1873, p. l'J4. 


same way the free caudals of carinate birds have transverse 
processes, which are at most faintly represented upon the 

fused posterior set of caudals which form the pygostyle. 
Four of the posterior caudals of Arcliceopteryx have fine 


splints of bone lying on one side, which have been compared 
to the ossifications in tendons found among the ptero- 
dactyles. They may conceivably be misplaced chevron 

The pygostyle varies much in the degree of its develop- 
ment. It is weakest in various aquatic birds, such as the 
auks and grebes, where it is thin and narrow ; in more 
purely flying birds it is very thick at the base, and is turned 
upwards instead of, as in the auks, carrying on the line of 
the tail. In the grebe there is really no more definite a 
ploughshare bone than in the ostrich. The number of 
vertebrae which are fused together to form the urostyle 
varies. In the ostrich MARSHALL finds four, five in the 
grebe and hornbill, six in the duck and in Eurylcemus. 

The total number of vertebrae ' in the column varies 
greatly; the extremes are something like thirty-nine and sixty- 
four (reckoning the urostyle as one). The greatest number 
characterises the ratites, and the smallest some of the higher 
arboreal birds. Archceopteryx had only about fifty vertebrae. 
While, therefore, it may be generally true to put down as 
older types those with the largest number of vertebrae, it is 
evident that on this view ArchccopteryxTfm.^i be regarded as 
a parallel branch to the existing birds, and not as their 
ancestor. The number of vertebrae, though it may perhaps 
be considered from this general point of view, is not of the 
faintest use for the systematic arrangement of existing forms. 
The number varies so extremely that among the Gruidae 
Professor PARKER found no two alike. Rather more fixed, 
but still subject to variation among the species of a genus, 
are the cervical vertebrae ; and some account will be taken 
in the pages which follow of this fixedness. The results 
must, however, be tempered by the reflection that while the 
common swan has twenty-five the black-necked swan has 

Between the successive centra are the ' intervertebral 

1 GIEBEL ('Die Wirbelzahlen am Vogelskelet,' Zcitschr. f. d. ges. Nat. xviii. 
18GG, p. 20) gives a long list ; see also ' Der letzte Schwanzwh bel des 
Vogelskeletes,' ibid. vi. 1855. 

I 2 


discs,' the ' intercentra ' or ' basiveiitral ' elements. These 
are, as has been shown, 1 originally the portions of each 
vertebra with which the ribs articulate, from which they 
are outgrowths. But as the ribs come to articulate with the 
centra these structures degenerate. In the development of 
Aptery.T T. J. PARKER found a postoccipital and a post- 
atlantal intercentrum, and two in the caudal region, which 
ossify so as to retain their independence in the adult 

Intercentra in the caudal region of the bird's vertebral 
column are by no means so rare as might be inferred from 

some published statements upon 

"\ --sss^^Tv ^ ie ma ^ter. They are especially 

conspicuous among the Limi- 
colae and the nearly allied auks, 
and in most water birds. In 
Numenius femoralis, for exam- 
ple, there are three small osseous 
nodules lying between caudal 
j^'^^S&&& 1 * vertebrae 1-5. Behind these 

FIG. 65. LAST Two VERTEBRA OF are a series of hypapophyses, 

STKUTHIO (ATTEK MIVAI.T). which ^ & contmuatlon of 

us, neural spines ; rf, osseous bridge. 

the same series, but much more 
pronounced and ankylosed to the vertebrae. 

They exist also in the duck tribe. In Biziura lobata 
there are three distinct intercentra in the form of largish 
nodules. I have found intercentra also in Palamedeae, Tubi- 
nares, Steganopodes, Colymbi, Herodiones, Opistlwcomus. 

These free intercentra are rare among the Pico-passeres, 
but in a few of them are present. Thus in Tcccus there is 
a distinct intercentrum lying between the last free caudal 

The hawk tribe have not these bonelets as distinct struc- 

In the cuckoos, parrots, Ralli, Otides, Columbae, and the 

1 This matter of the composition of the vertebra has been recently gone into 
by G.VDOW (on the 'Evolution of the Vertebral Column of Amphibia and Amniota,' 
Phil. Trans. IH'.Mi, p. 1), who quotes previous literature. 


tinamous I have not seen in the adult skeleton any free 
intercentra ; nor in the Grues, excepting Chitnga. 

Further details on this matter will be found in the paper 
cited below. 1 

Though free intercentra are by no means universal 
among recent birds, hypophyses of the last caudals are 
almost so. That these latter are derived from intercentra, 
and are, therefore, not comparable to the hypapophyses of the 
cervicals and dorsals, is clear from such cases where the gra- 
dual transition between free intercentra and fixed hypapophy- 
ses is shown. In reptiles the intercentra are in the tail region 
constantly in the form of chevron bones, which are V-shaped, 
articulating with the vertebral column by the free ends of 
the V. This form of the hypophyses of the caudal region is 
not so common as a simply bifid condition, but does oc- 
casionally occur. I have seen it, for example, in Tubinares, 
Accipitres, and Cuculi. 

The first vertebra of the cervical series is called the 
atlas (see fig. 66) ; it is a ring on bone, of which the greater 
part of the ' centrum ' is formed by the projecting odontoid 
process (see fig. 67), the rest being formed by a pair of 
intercentra. In the hornbills the atlas is fused with the 
following axis vertebra. 2 Generally the atlas has not what 
the succeeding vertebrae have, a vertebrarterial canal, but 
this is sometimes present (see under ' Kibs,' p. 119). The 
odontoid process sometimes notches the lower part of the 
atlas, and sometimes perforates it. These two conditions 
are illustrated by figs. 66 and 68. It sometimes happens that 
the neural arch of the atlas is incomplete, e.g. CJuuti/u, 
Colius, Pandion. As a rule it is perfect. 

In the cervical vertebrae the chief facts which appear to 
be of systematic importance are the relations to each other 
of the paired processes, to which MIVART has applied the 
name of catapophyses. These are sometimes inconspicu- 
ous processes of the transverse processes on the under 
side. Very often the last one or two pairs of them closely 

1 BEDDAED, 'Note upon Intercentra,' &c., P. Z. S. lsi7, p. 4>'<~>. 
- In a specimen of Chunga I have found the same fusion. 


approach each other in the middle line, as in Psophia. 
Sometimes a number of these processes unite to form a 
canal ; this occurs in the Steganopodes, and most Herodiones 
(but not in Scopus umbretta) ; but the classificatory signifi- 
cance of the fact is marred by the occurrence of a similar 
canal similarly formed in some Picida3, and in the case of 
one vertebra in the parrot, Eclectics polychlorus. 

It is sometimes the case that the last of the catapophyses 
is consolidated into a thick process, which is bifid at the 
extremity ; this process forms a transition to the following 
haemapophyses (or hypapophyses) . These latter are un- 




in', articular surface ; 
, vertebrarterial 
caual ; lip. liyper- 

FIG. 67. Axis OF EMU (AFTER 

o, odontoid process : ., neural spine ; 
:, anterior zygapophyses ; pi, 
pleural lamella ; pc, articular sur- 
face ; Aw. hypapopliysis : lip, kyper- 

AS m FIG. 66. 

paired median processes, which commence upon the cervical 
vertebrae, and extend for a variable distance back along the 
dorsal vertebrae. They are very feeble, and sometimes 
limited to the cervical region, in the Herodiones. They are 
most highly developed in Sphenisci, Colymbi, Alcse, and some 
Anseres, being in these cases continued to the end of the 
dorsal series, and even being found upon some of the lumbar 
vertebrae. In many cases these processes are flattened out 
at the free end like an inverted T, or are trifid at the same 
place. This is seen to be due to the gradual shifting in 
position of a posterior set of catapophyses, which at first are 
at the sides of, and far from, the haemapophyses, but 


nearer and nearer, until at length they mount upon the 
hsemapophysis itself and pass to its very end. Details of the 
formation of the vertebra 3 will be found in the systematic 
part of this book. 1 

Ribs.' 2 The ribs of birds vary greatly in number. There 
are as a rule three series of ribs to be distinguished. The 
last cervical vertebrae, more or fewer of them, are furnished 
with short ribs which do not reach the sternum. Behind 
these are, again, a variable number of true ribs, which do 
reach and articulate with the sternum. These true ribs 
consist of the vertebral portion, which articulates with the 
vertebra, and of a sternal portion, which is articulated with 
the vertebral half of the rib above and with the sternum 
below; it is bent at an angle with the vertebral portion. 
Attached to, originally separate from, and sometimes per- 
manently separate from, the vertebral half of the rib is the 
uncinate process, of which there are a variable number. 
These processes are absent 3 in Arcliceopteryx and in the 
Palamedese only. Behind the true ribs, which articulate 
with the sternum, are a variable number, in all degrees of 

1 The relationship of the so-called catapophyses to the unpaired ha?mapo- 
physes varies, and suggests what has been advanced on other grounds an 
occasional excalation of vertebra?. Without wishing to commit myself to a 
belief in the actual dropping out of a vertebra from the middle of the series, I 
may mention some of the facts which may be regarded as pointing in this 
direction. In the grebe sEchmophorus the catapophyses form on certain 
vertebra a complete ventral canal for the carotids. The summit of the arch 
thus formed gradually acquires a median dorsal process. This increases, and 
the catapophyses finally end in the obliteration of the canal which they sur 
round, and a solid arch is formed ; the hypapophysis of the succeeding vertebra 
is single and no longer retains traces of its evolution from a ring of bone sur- 
mounted by a process. In other cases the catapophyses suddenly end and the 
hypapophyses begin without such intermediate stages. An intermediate stage 
is seen in certain types where the catapophyses end suddenly, but the first 
hypapophysis is double, either formed of two clearly fused pieces or with merely 
a bifid spine. These latter cases suggest the dropping out of one or more 
vertebra?, effecting the transition between the paired catapophyses and the un- 
paired hypapophyses. 

'' In all birds except Arcliceoptcryx the ribs are two-headed with a capitulum 
and tuberculum. 

3 They have been often said to be absent in Dinornis, but they are not. 

W. BEHREXS, Untersuchuiigen fiber den Proccssus uncinatus der VHt/cl nnd 
Crocodile, Inaug. Diss., Gottingen, 1880. 


degeneration, of floating lumbar ribs. Morphologically 
equivalent with ribs are the processes firmly ankylosed to 
the cervical vertebrae, which form a canal for the vertebral 
artery ; these are, as a rule, absent from the atlas, but are 
present on that vertebra in the Anseres, Opisthocomus, 1 
Triponax Feddeni, Dromccus' 2 (fig. 66). Arcliceopteryx is 
alone in possessing the abdominal ribs of the crocodiles and 
other reptiles. 

The Shoulder Girdle. 3 The shoulder girdle of birds consists 
of at any rate three separate elements the scapula above ; 
the coracoid, articulating with the sternum ; and the clavicles, 
generally united into a U-shaped piece. Of these the first 
two are preformed in cartilage, the last in membrane. 

The scapula is a thinnish sword-shaped bone which 
is attached by muscles to the ribs and to the vertebrae, 
and lies in a direction, as a rule, nearly parallel to the 
long axis of the body. The scapula does not show great 
variability of form among birds ; the most considerable 
variation is to be seen in the penguins, where the bone is, 
comparatively speaking, of enormous width. A free supra- 
scapula has been noted by PARKER in Opisthocomus. The 
coracoids articulate on the one hand with the scapula, and 
on the other with the sternum, where they are received into 
grooves on its anterior margin. There are some variations 
in the way in which these grooves are arranged : in some 
birds the two coracoids at their insertion are not in contact 
at all ; in others they are in contact ; and finally they may 
overlap, as in reptiles. 

The coracoid has in many birds a procoracoid process, 
which is believed to be the equivalent of the procoracoid of 
reptiles. This is especially prominent in the ostrich, but is 
present in a large number of other birds, though more 
reduced in extent. But its large or small size is so capricious 

Stated by PAEKEE to be absent. 

MIVAET figures a canal on one side of the atlas of the ostrich. 

'' A. SABATIEE, Comparaison des Ceintiircs ct dcs Membres Ant&rieurs ct 
Postcrlcurs dans la Seric dcs Verti'bres, Montpellier, 1880 ; PARKEB, quoted 
below. See also LUHDEE in J. f. O. 1871, p. 321. 


in its relations to other structural similarities and dissimilari- 
ties that the fact is not of great use in classification. 

The same remark may be made about the foramen coni- 
coidcum perhaps, which, again, is found in many birds and 
absent from others. As to the morphological significance of 
this foramen, which transmits a nerve twig to the pectoralis 
secundus, it may perhaps be regarded as the boundary 
between the coracoid and the procoracoid. 

In the course of the development of the common fowl, 
according to Miss LINDSAY, whose figures are here reproduced, 
there is a very considerable trace of the procoracoid. The 
three elements of the shoulder girdle are perfectly distinct 
from each other in the young embryo, but become fused (the 
scapula and the coracoid), again to get separate in the older 
chick. This temporary fusion may be significant of the 
struthionic condition to be described later. The intermediate 
piece is, it will be noticed, triangular in form, the elongated 
aspect of the adult coracoid being acquired later. Miss 
LINDSAY is of opinion that this change of form is to be cor- 
related with the disappearance of the anterior section of the 
bone, as indicated in the accompanying diagram, the disap- 
pearing (shaded) part being the equivalent of the procora- 

The interrelationship of the scapula to the coracoid offers 
facts of some importance. In the ostrich tribe the two bones 
are firmly ankylosed ; this is not the case with the young, 
but it is plainly the case with the adult. In carinate birds, 
on the other hand, there is not ankylosis, but a close union 
by means of fibro-cartilage. It appears, however, that in 
Didus (exceptionally '?) there is an actual synostosis, which of 
course bears out the suggestion that the synostosis of the 
ratite birds has something to do with their loss of the power 
of flight. In the ratite birds and in Hesperornis the scapula 
and the coracoid are nearly in the same straight line, the 
angle in Aptenjx varying from 150 to 122, whereas in the 
carinates the two bones are at right angles or at an even 
acute angle. That this is not a morphological distinction, 
but is distinctly related to the development of the shoulder 



muscles, has been clearly pointed out by T. J. PAEKEE.' He 
discovered an approximation to the struthious condition in 
several of the flightless rails and other birds. But the state 
of affairs which characterises the Tubinares warns us against 
placing too much reliance upon this apparently sound gene- 
ralisation ; for in them we are informed by FOEBES that 
' the angle it ' (the scapula) ' forms with the coracoid varies 
much in different genera, being most acute in Pelecanoides, 
whilst in the Oceanitidge it is hardly if at all less than a right 




i 1 /, clavicle ; j>c, procoracoid ; cor, coracoid ; sc, scapula. 1-3, fifth day ; 4, sixth day ; 

5, late on sixth day. 

angle.' The widest angle in a carinate bird is 106, so there 
is a difference of only 16 between extremes of carinates and 

The two clavicles - sometimes spoken of collectively as 
the furcula vary much in their degree of development. They 
are totally absent in the Apteryx. In the emu and in certain 
parrots they are distinct and smallish bones which do not 
come into contact with each other ; but in the majority of 
birds they form a single U- or V-shaped bone. 

The furcula varies in the expansion or non-expansion of 
the base to form a circular hypocleidium. In some birds the 

1 ' On Notornis,' in Tr. N. Zealand Inst. xiv. 1882. 

- A. WEITZEL, ' Die Furcula : eiu Beitrag zur Osteologie cler Vogel,' Ze-itsrlti-. 
f. d. ges. Naturw. xxv. 1865, p. 317. 


furcula is connected with a ligament (e.g. Psophia) by articu- 
lation (e.g. Herodiones) or by direct synostosis with the carina 
sterni. In the gallinaceous birds the furcula does not come 
into near relations with the carina, and in Opisthocomus 
quite exceptionally the bone is ankylosed with the spina 
sterni on the one hand, and with the coracoids, so firmly 
that ' no trace of the primitive distinctness of the bones is 
discernible.' There are also considerable variations in the 
degree of the connection between the furcula and the coracoid 
and scapula. 

In a few birds the ends of the clavicles where they arti- 
culate, or at least are connected, with the coracoid and scapula 
have a process, the acrocoracoid process of the clavicle. This 
is seen, for instance, in the Anseres, and the rudiment of such 
a process in the flamingo appears at first sight to be a duck- 
like character in that bird. But the same process is also 
developed, and to a great extent, in the Alcedmidae, a fact 
which must be borne in mind before coming to any such 

FURBRINGER has devoted some space to describing and 
illustrating the relations at their articulations between the 
clavicle, coracoid, and acrocoracoid. The two extremes may 
be seen in Phahicrocorax and Psophia ; in the former the 
clavicle articulates with the acrocoracoid only, and does not 
reach the scapula ; in the latter, where the procoracoid is 
well developed, the clavicle comes into contact with all 
three. 1 Further details will be found under the description 
of the different groups. 

The Fore Limb. The fore limb is present in all birds except 
most Dinornithidae, where up to the present no trace of one 
has been discovered. In Hesperornis only the humerus 
appears to exist ; in Apteryx, Dromaus, and Casuarius there 
is but one finger. With these exceptions the wing of birds 
consists of a humerus, radius and ulna, carpus, metacarpals, 
and three fingers (with sometimes a rudiment of a fourth) ; 
even Archaopteryx has not been definitely shown to possess 

1 A. TSCHAX, Recherchcs sur VExtr&mitf AnUrienre dcs Oiseaux et des 
Reptiles, Diss. Inaug., Geneva, 1889. 


more than the typical three fingers (see, however, below). 
The relative length of the arm varies much in birds ; it is 
longest in the flying gulls, terns, c., whence the name 
applied to the former of Longipennes. In the struthious 
birds it is the shortest, and in many running birds the wing 
is reduced in length. There is too variation in the relative 
lengths of the humerus, fore arm, and hand. In the divers, 
for instance, the upper section of the arm is the longest, in the 
gulls the fore arm, and in the penguins the hand. The 
length of the hand in the Macrochires is so great that it 
equals that of the humerus and fore arm together. The 
exact reverse is seen in the ratites, where the length of the 
humerus is greater than that of the rest of the wing. PYE- 
CR.AFT has brought out the interesting fact that during the 
growth of Opisthocomus the proportions of the different sec- 
tions of the w y ing alter. 

A study of the relative lengths of the different parts of 
the arm shows that a reduction of the wing, and a consequent 
decay of its powers as an organ of flight, do not invariably 
follow the same path. In the ostrich the middle segment is 
the shortest, in the cassowary the hand. 

The length of the humerus, the exact form and degree of 
development of the deltoid ridge, and of the tubercles for 
the insertion of muscles, furnish systematists with reliable 
points for the identification of genera and species. So much 
of our knowledge of extinct birds depends upon fragments of 
this and others of the ' long ' bones that the value of slight 
characters of this description has been thoroughly appraised. 
A glance, for instance, at LYDEKKER'S recently published 
' Catalogue of the Fossil Birds in the British Museum ' will 
reveal the importance of the power of discriminating species 
by such slight indications, which furnish the student of 
affinities between families or genera with nothing tangible. 

The radius and the ulna are always separate bones, of 
which the ulna is the longer ; it is frequently marked on its 
outer surface with tubercles, to which the quill feathers 
are attached. The most striking modification of the radius 
is seen in the Parridse (fig. 70, p. 125), where it is prolonged 


1 I'O 

on the outer side into a strong flat process, the upper surface 
of which is slightly grooved for the reception of the tendon 
of the extensor metacarpi radialis muscle. 

The carpus of recent birds consists only of two separate 
bones. But in the embryo there are six separate cartilages. 
The two bones which persist are looked upon by ZEHNTNER 
as ulnare + intermedium and radiale + centrale. The three 
distal carpals, according to PAEKER, fuse with their three 
metacarpals. In the emu, according to PARKER, there are 
no carpal elements either in the young or adult, in Casuarius 
galeatux there is a small ulnare. 

Though no bird has more than three more or less com- 

Fir;. 70. RADIUS AND ULNA or Metopidiits (AFTER FORBES). 

plete digits, there is commonly a trace of a fourth meta- 
carpal, found by ZEHXTXER in the development of the swift, 
and by STUDER in the development of the penguin. PARKER 
was convinced of the existence of a prepollex and of inter- 
calary digits, but WIEDERSHEIM looks with no favour upon 
this broadening of the hand. 

The metacarpals are free only in Archceopteryx and 
Gastornis ; in all other birds they are partially fused. The 
formula for the phalanges in Arcliceopteryx is I. 2, II. 3, 
III. 4. In some other birds it is I. 1, II. 2, III. 1, with the 
exception of the ostrich, 1 Numenius, and the embryo duck,' 2 
where digit III. has a small additional phalanx. In other- 
birds the formula is I. 2, II. 3, III. 1, and in the apteryx 
digit II. shows three phalanges in the course of its develop- 

1 PARKER, ' On the Structure and Development of the Wing in the Common 
Fowl,' Phil. Trans. 1888, p. 385, where previous literature is quoted. 

2 BAUR, Science, vol. v. p. 355. 


ment, of which one (apparently the middle one) is not 
recognisable in the adult. 

LEIGHTOX has contributed to this question with a study 
of the development of the wing of Sterna. 1 He finds here 
too a rudimentary fourth digit, which in the first stage, 
which he figures, is as long or nearly as long as the first 
digit. A rudimentary metacarpal even persists attached to 
the side of the last metacarpal in birds just before hatching. 
In the carpus there are never more than four distinct 
cartilages ; there are in the first place a radiale and ulnare, 


1, phalanx 1 (/'/<!) arid rudimentary phalanx 2 (7V<2) of digit III. ; c, connective tissue. 
2, another specimen with phalanges ankylosed. 3, distal part of digit III. of 
embryonic manus (4). 

and distally two cartilages, of which one appears to represent 
the combined distalia of the two first digits, and the second 
that of the third digit. In birds, just before hatching, all 
the distalia have fused into one mass. The cartilages 
lettered respectively radiale and ulnare in the figures are 
thought, however, by the author to be really radiale + 
intermedium and ulnare + centrale ; and in support of this 
view is the partial separation between the two supposed 
elements of each, which is, however, never carried very 

As to the homologies of the digits in the adult with those 

1 ' The Development of the Wing of Sterna Wilsonii,' Tufts Coll. Studies, 
1894. Previous literature is here quoted. 


of the reptilian hand, LEIGHTON leans to the view that 
the supposed pollex is really the index. In putting forward 
this opinion he rests first of all upon the fact that the radial 
artery is absent, thus indicating a reduction of the radial 
side of the hand ; the second argument is derived from the 
fact that in animals with a reduced manus the first digit 
is the first to go, and then is followed by the last ; thus in 
Orohi/>p/tK there are four digits, the first having disappeared, 
while in Protoliippus the fifth has vanished. In addition 
this view is moreover strengthened by a consideration of the 
most reduced manus that occurs in birds ; in Apteryx and 
Casuarius the reduction has similarly occurred on both sides 
of the large persisting digit, which is thus to be regarded as 
No. III. 

Sternum. The sternum in its most complicated condition 
consists of the following regions (see fig. 72) : Anteriorly 
it ends in a moderately narrow extremity which is known 
as the manubrium sterni or rostrum. On either side of this 
is a forwardly directed process, the costal processes or 
(interior lateral processes. In the middle of the sternum, 
and forming the great projecting keel, is the lophosteon or 
carina sterni, or keel. The sternum ends in a median 
process behind (sometimes, but wrongly, called the xiphoid 
process), to which are appended two processes on each side, 
which may be termed middle and external xiphoid processes, 
or these may be termed, for reasons which will appear later, 
the posterior lateral process and the accessory process. The 
nomenclature first used in the preceding brief descriptions 
is that of HUXLEY ; the second set of terms which will be 
used throughout in the descriptions which follow are those 
used by Miss LINDSAY in her paper upon the development 
of the avian sternum. 1 

The sternum is subject to much modification among 
birds, of which the principal varieties will be now described. 
The birds which show perhaps the greatest difference from 
the gallinaceous type, selected for the above description, are 
the ratite birds. In them there is no keel developed, hence 

1 ' On the Avian Sternum,' P. Z. S. 1885, p. 684. 


the name ratite (raft-like), or at most, as in Eliea, a slight 
protuberance, which, however, as will be pointed out imme- 
diately, is not really comparable to the keel of the carinate 
birds. There are, however, other birds, such as the extinct 
Cnemiornis and the living 
Stringops, in which the 
keel is absent, its ab- 
sence being associated 
with the loss of the 



FIG. 72 STERNUM OF Loplio- 
pliorus impeyanus (AFTER HUXLEY). 

x, rostrum ; cj>, anterior lateral process ; 
ji.t.n, posterior lateral process ; e..i; /'../, 
its inner and outer divisions ; l.o. cariua. 

FIG. 73. STERNUM OF Podica 
senegalensis (AFTEK BEDDABD). 

cl, clavicle ; co, - coracoid ; cl, x, articula- 
tion of clavicle. 

power of flight. In the singular Opistliocomus the anterior 
part of the keel is, as it were, cut away (the enormous crop 
resting here), the posterior region being retained. The four 
posterior lateral processes of the sternum figured above are 
not always present in birds. The extremest modification 
is as seen in the goose and the crane, where the posterior 


margin of the bone is entire, without any processes at all. 
In passerine birds generally, and in some others also, there 
is but a single pair of these processes ; while, finally, by 
excessive growth of the parts concerned the processes have 
joined and converted the notches into foramina. The con- 
verse course of events has been suggested i.e. that deficient 
ossification leads to the fenestrated condition, whence to 
the posterior notches is an easy step. Development, how- 
ever, shows that the former view is the more correct. The 
diversities in the form of the sternum undoubtedly must 




crt, anterior lateral process ; c, grooves for eoracoids ; /, elevation in centre ; ?.r, posterior 
end ; w, lateral view showing articulation of rilis. 

have some relation to the muscles which are inserted on to 
and take their rise from the margins of the bone. Thus, as 
already mentioned, the flat sternum of the Ratitse is associated 
with the slight development of the pectorales muscles and 
the consequent loss of capacity for flight. It has been 
ingeniously suggested that the relative development of the 
posterior lateral processes of the sternum has possibly an 
analogous explanation. The muscles that are attached 
thereto are mainly the pectorals and the abdominals. Now r 
the pull of these two is in an opposite direction. The 
tendency of the action of the pectorals would be to 



straighten the posterior margin of the sternum, while 
that of the abdominals would he to pull it out. perhaps 
irregularly. Hopping and walking birds might therefore 
be expected to have a more notched sternum than purely 
flying -birds ; that there is some relation of this kind seems 
possible when w r e contrast the sternum of the running 
gallinaceous bird with that of the essentially aerial eagle. 
Moreover, since the pull of the abdominal muscles is in two 
directions, one antero-posterior (recti), the other oblique (the 
obliqui), we might expect to find what we actually do find, a 
direction of the xiphoid processes which corresponds with 
the resultant of these two forces, as is indicated in the 
annexed diagram. There are a few other modifications in 
the shape of the sternum which have been made use of for 
systematic purposes, besides the keel and the notches, or 
excavations of the posterior border. The rostrum of the 
bone is sometimes very pronounced, and sometimes practi- 
cally absent altogether. According to its position, more 
dorsally or more ventrally, the process has been called by 
FURBRINGER spina externa, or spina interim sterni. In 
the gallinaceous birds the two are combined in a vertically 
compressed plate of bone which arises both from the lower 
and from the upper side of the sternum. In the passerines 
and in the todies, and a few of the allies of these groups of 
birds, the anterior process of the sternum is more or less 
distinctly bifurcate. 

The sternum of birds arises, as does that of other verte- 
brates, in the first place between the ends of the ribs which 
fuse together. Birds invariably have a few ' floating ' ribs 
at both ends of the sternum which are no longer connected 
with it, this connection being often lost ontogenetically. 
There is, in fact, usually a shortening of the sternum during 
development. The keel arises from the conjoined edges of 
the two sets of fused ribs ; it is not preformed separately as 
a median piece. This seems to settle in the negative an 
earlier view that the carina sterni was the surviving repre- 
sentative of the interclavicle of the reptiles, a view which 
commended itself to more than one anatomist of distinction, 


and appeared to be strengthened by the occasional connec- 
tion by ligament and bone with the hypocleidium. 

The most recent modification of this view is put forward 
by PARKER, who has shown in Opisthocomus a needle-shaped 
splint of bone lying upon the keel, and therefore independent 
of it. 

The. development of the sternum throws a light upon the 
homologies of its different parts in different birds, and in 
other vertebrates. 

It is plain in the first place that the spina externa and 
the spina interna have nothing whatever to do with the 
manubrium sterni of the mammal ; for they are (in the bird) 
secondary outgrowths, and not, as in the mammal, part of the 
primitive sternum formed by concrescence of the ribs. The 
same holds good of the posterior median region of the bone, 
which is a secondary outgrowth, and can therefore have no 
relations with the xiphoid process of the mammalian ster- 
num ; what does correspond to the latter are the posterior 
lateral processes of the avian sternum. 1 

Pelvis. 2 The pelvis consists of three pairs of bones, the 
ilium, ischium, and pubis. In the young embryo these 
bones form a continuous sheet of cartilage, but are all 
separate distally ; the ilium is directed in an antero-posterior 

1 The literature of the sternum is large, and is to a considerable extent to be 
found under the several groups. Memoirs of a wider scope are W. K. PARKER, 
' A Monograph on the Structure and Development of the Shoulder Girdle and 
Sternum in the Vertebrata,' Ruy Soc. Publications, 1868 ; L'HERMINIEK, 
' Recherches sur la Marche d'Ossifications,' &c., M&m. Ac. Sci. 1830. The 
history of the development of knowledge concerning the ossification of the 
sternum and the classificatory results therefrom is treated by NEWTON in Diet. 
Birds, ' Introduction.' Miss LINDSAY'S paper, already quoted, contains references 
to the chief memoirs upon the subject. See also li. DIECK, DC Sterno Avium, 
Diss. Inaug., Hal, 1867. 

'-' C. GEGENBAUR, ' Beitrage z. Kenntniss des Beckens der Vogel,' Jen. 
Zi'itsclir. vi. p. 157 ; MEHNERT, ' Untersuchungen liber die Entwicklung des 
Os pelvis d. Vogel,' Morph. J.B. xiii. 1888, p. 259 ; A. JOHNSON, ' On the 
Development of the Pelvis Girdle, etc., in the Chick,' Quart. J. Micr. Sci. 1883, 
p. 399 ; B. HAIJ, Jemfiii-ande Stxdicr ofvcr Foglarncs Backen, Lund, 1887, and 
' Morphologisk Byggnoden af Ilium,' &c., Act. Lund. Univ. xxii. 1887, p. 1 
G. BAUR, ' Bemerkungen liber das Becken d. Vogel v. Dinosaurier,' Morph. J.B. 
x. 1885, p. 613 ; A. BUNGE, Untersiichnngeti zur Entwicklungsgeschichtc des 
Beckcngiirtcls, &c., Diss. Inaug., Dorpat, 1880. 

K 2 


direction ; the ischium and the pnbis look downwards and 
slightly backwards ; at the end of the pubis, near to where 
it comes into contact with the iliac portion of the cartilage, 
is a forwardly directed process, the prepubic process. This 
primitive state of affairs has been most nearly preserved in 
Apteryx and Dinornis ; in these birds the pubis and ischium 
are free from each other distally and from the ilium ; their 
direction is, however, more backwards than in the embryo, 
and the prepubic process is relatively smaller. In all other 
birds the pubis and the ischium lie in a line more parallel 

U, ilium ; p, pubis ; ps, interobturator process ; lp, pectiueal process. 

with the ilium, and there is a greater or less connection 
between the several bones. This is seen in a less developed 
condition in the struthious birds and tinamous than in any 
others. In the tinamous, in fact, the pubis and ischium are 
quite free from each other distally, and from the ilium. In 
the ostrich the pubes unite in a ventral symphysis ; in Ehea 
there is a remarkable modification induced by the meeting 
of the ischia. The details of the struthious pelvis will be 
found described under the description of that group. In 
carinate birds the ischium is for the greater part of its extent 
fused with the ilium, a foramen only the ischiadic foramen 
-being left anteriorly. The pubes join to a less extent with 
the ischia, and are sometimes (e.g. Colymbus) quite free 


from them. The prepubic or pectineal process is large in 
the strathious birds, in the tinamous, and in a few carinates, 
such as Geococcyx. It has received much attention as the 
possible equivalent of the reptilian pubis, the bird's so-called 
pubis being in that case the homologue of the posterior pubic 
process of the reptile. The dinosaurians seemed at one time 
to have been the means of solving the questions involved ; 
for in some of them there is a backwardly directed pubis, 
not quite so bent as in the bird, from the anterior and upper 
end of which a stout bone, considered to be the homologue 
of the pectineal process, is directed forwards. This latter, 
however, is clearly a part of the pubis, while the pectineal 
process is at least not always a product of the pubis, being 
sometimes purely iliac in origin, sometimes partly pubic and 
partly iliac. It may be that the missing prepubic process of 
the dinosaurian pelvis is represented by the remarkable 
bones ankylosed to the ostrich's pelvis anteriorly, and con- 
tinued forwards by a cartilaginous tract, which were dis- 
covered by DARWIN and GARROD. This will reduce the 
pectineal process to the level of a mere projection of the 
pelvis of no particular significance save as a secure hold for 
the important ambiens muscle which is there attached. In 
any case it is safe to assert that wherever that process is pre- 
sent and long the ambiens is also present and well developed. 

Among carinate birds the pelvis does not show a great 
variability of form. The pelvis is either broader or narrower, 
being excessively narrow in the Colymbi. The proportions 
of the preacetabular and the postacetabular regions also 
differ, as do their relative breadths. 

It is noteworthy that the pubis is sometimes defective in 
the middle, appearing then to consist of a proximal and 
distal portion unconnected in the dry skeleton. The fact 
also that but few muscles arise from the pubis seems to show 
that it is in a condition of degeneration. 

In ArcJiCBOpteryx alone are the elements of the pelvis 
not ankylosed together. 

The acetabulum of all birds except Dromccus is perfo- 


rated ; but the perforation is reduced in size in Geococcyx, 
Tinamus, and Hesperornis. 

Hind Limb. The hind limb of birds consists of femur, 
tibia and fibula, tarsus, metatarsus, and phalanges. 

In all birds the femur is shorter than the tibia, the pro- 
portions varying much. It seerns impossible to place those 
birds in which the difference is least at the base of the series, 
on account, of course, of a resemblance so far to reptiles, 
since relative importance of the fore and hind limb appears 
to have something to do with the matter. From FUE- 
BEINGEE'S tables ' it is to be gathered that Fregata is the 
bird in which the difference between those two segments of 
the leg is least. It is most pronounced in the divers, flamingo, 
and Tubinares. A bone of some classificatory importance is 
the patella, 2 a sesamoid on the upper surface of the knee. 
This bone is not ossified at all in Colymbns, but is enormous 
in the grebes and in Hesperornis, in which latter, as in 
Phalacrocorax and Biziura lobata, it is perforated by the 
tendon of the ambiens. In Plotus the patella is grooved only 
for this tendon. 

In no bird (except as an occasional abnormality) is the 
fibula a complete bone. It fails below, and does not reach 
the tarsus. It is usually more or less coalescent with the 
tibia. The latter is a strong bone with a crest in front, which 
is enormous in the divers and Hesperornis. The distal end 
of the tibia is formed by a portion of the tarsus, of which 
the remaining portion is coalesced with the metatarsus. The 
tarsus in the embryo 3 consists of three chondrites, a tibiale, 
a fibulare, and a distale. The latter represents the separate 
distal elements of the tarsus fused. The tibiale sends up- 

1 KESSLER (' Osteologie der Vogelfiisse,' Bull. Soc. Nat. Mosc. 1841) has also 
given tables. The value of the ' long bones ' of the leg for ' defining orders, 
families, and often genera ' is plainly set forth in this paper. See also MILNE- 
EDWABDS, Oiscaux Fossilcs de. la France, where further information is to be found. 

2 J. KACZANDEK, ' Beitrag zur Lehre iiber die Entwicklungsgeschichte der 
Patella,' Mt. Embr. lust. Wien, (2) ii. 1887, p. 12. 

3 G. BAUK, ' Der Tarsus der Vogel u. Dinosaurier,' Morpli. J.B. viii. 1883, 
p. 417 ; E. G. MOKSE, ' On the Carpus and Tarsus of Birds,' Ann. Lye. Neui 
York, x. 1873, p. 141, and ' On the Identity of the Ascending Process of the 
Astragalus,' &c., Anniv. Mem. Bost. Soc. Nat. Hist. 1880. 


wards an ' ascending process,' found also in the dinosaurs, 
which is the equivalent of the intermedium, while the 
centrale is represented by a distinct osseous nodule in the 
adults of the Struthiones (including Dinornis) and tinamous. 
In Apteryx T. J. PARKER found two osseous centralia. 

The number of toes and phalanges has been already 
described above. 

The Skull. 1 While presenting many characteristic features 
of its own, the skull in birds shows certain fundamental like- 
nesses to the skull of the reptilia. As in them, and contrary 
to what we find in the mammalia, the skull of birds 

1. Articulates with the spine by a single occipital 

2. Possesses a quadrate bone for the articulation of the 

3. The mandible itself is composed of at least a dentary 
angular and articular portion. 

4. The columella auris is very similar. 

The bird's skull is, however, distinguishable by a number of 
characters, of which the following are the most important : 

1. The bones of the cranium are very closely united and 
fused, this being less marked in the penguins and ratites. 

2. The brain case is large as compared with that of 

3. The bones of the skull, as are those of the skeleton in 
general, are light and contain air spaces. 

4. The columella and the os transversum of the reptiles 
are absent. 2 

5. There is no distinct postf rental. 

The bones of the bird's skull, as that of other ver- 
tebrates, may be distinguished into four categories 
(1) those of the cranium ossified from its cartilage ; (2) those 
of the sense capsules ; (3) those of the visceral arches, and 
(4) membrane bones connected with the several regions 

1 H. MAGNUS, ' Untersuchungen iiber d. Struktur d. knochernen Vogelkopfes , 
Zcitsclir. wiss. Zuul. xxi. 1871. 

2 See however below, under Passerine skull. 


1 . The bones formed by ossification from the chondro- 
cranium are four occipitals, viz. basioccipital, two exoccipitals, 
and the snpraoccipital, forming a complete ring of bone 
round the foramen magnum ; basisphenoid, with two wings, 
the alisphenoids ; presphenoid, with two wings,, the orbito- 
sphenoids (occasionally atrophied, e.g. Apteryx) ; meseth- 
moid, with two lateral wings, the ectethmoids (sometimes 
termed prefrontals) ; these are occasionally absent as distinct 
ossifications, and may sometimes, on the other hand, be very 
large and even appear on the frontal surface of the skull, 
marking the orbit anteriorly ; in those cases they take the 
place of the orbital part of the lacrymal and have a better 
claim to be called prefrontal. 

'2. The investment of the auditory capsule, termed 
collectively the periotic bone, consists of three separate 
elements, the prootic, opisthotic, and the epiotic, which last 
is absent in the Aptery.r. 

3. The first visceral arch, the mandibular, has but two 
bones l ossified from its cartilage, the quadrate and the 
articulare of the lower jaw. The second and third arches 
form the hyoid apparatus ; the ossifications are, first, the 
columella auris, a bone corresponding physiologically, if not 
also morphologically, to the ear bones of mammals ; secondly, 
a median piece in front, composed of two fused pieces, the 
basihyal, with sometimes lateral processes, the ceratohyals ; 
thirdly, the basibranchial, with two long lateral outgrowths, 
of which the nearest regions are ossified to form the cerato- 
branchials ; thirdly, a single median piece (sometimes absent), 
the urohyal, a remnant of the third arch. 2 

The membrane bones of the bird's skull are numerous, 
and may be referred to the same categories as the cartilage 

1. Associated with the cartilaginous cranium are pos- 
teriorly the parietals ; in front of these the frontals ; with 
the frontals articulate the lacrymals, of varying development, 

1 A mento-meckelian has been recently discovered in hawks. 
- For modifications of hyoids see especially GADOW, in Bronn's Thicrrcic/i, 
and GIEBEL, Zcitsclir. f. d. gcs. Naturw., xi. 1858. 


and in the hawks bearing a second and separate bone 

behind. These bones appear sometimes to have a definite 

relation to the cartilaginous ectethmoids. I do not refer so 

much to the fact that they sometimes entirely fuse with 

them (and with the skull wall) as to the varying size and 

relations of the two. In the kingfishers, for example, where 

the ectethmoids are small, the lacrymals are large, and have 

below an expanded plate which supplies the place of the 

feeble ectethmoid. When the lacrymal does not reach the 

orbital margin, as in Corvus, the 

ectethmoid does, and, as it were, takes 

its place. In many birds belonging 

to quite different orders there is a 

small bone connecting the lower end 

of the lacrymal, or of the ectethmoid, 

with either the palatine or the jugal 

bar ; this bone has been termed ' os 

crochu,' os uncinatum, os lacrymo- 

palatinum, and will be described in 

detail in those birds where it is to 

be found. 1 It may be that the os 

uncinatum should have been de- 

scribed as one of cartilaginous bones 

of the cranium. In some birds (tina- FIG. 7G. HYOH> OF Latha- 

mous, Menura, Psophia, and A rbori- discolor 

Cola) there is a Set Of SUpraOrbital &, basihyal ; M, hypnbrancliial : .*, 
, . . , i -j. i ceratobranchial ; , uroliyal ; c, 

DOIieS margining the OrbltS abOVe. entoglossum ; }>, parahyal ; c, cou- 

The base of the brain case is protected 

by a large basitemporal, which has sometimes (e.g. Apteryx) 
a long rostrum in front. The maxillae are sometimes sepa- 
rate from each other, and at times united across the middle 
line by more or less extensive ossifications, of which a pro- 
minent one, and with the appearance of a separate bone, 
is the maxillo-palatine. The premaxillaries in front of these 
send back a long process extending as far as the nasals. 

'2. To this category, perhaps, belong the squamosal, 

1 See under Cariauta, Tubinares, Steganopodes, Musophagi, where cross 
references will be found. 


referable to the auditory capsule, and the nasals above and 
the vomers below to the nasal capsule. The vomers are 
often paired bones, and near to them are other small and 
independent ossifications, such as the septomaxillaries (see 
under Passeres) and the mediopalatines (see under Cuckoos) . 
Connected with the eye ossified, indeed, in the sclerotic is 
the sclerotic ring. 

3. As membranous ossifications connected with or in 
the neighbourhood of the visceral arches are the pterygoids 
and palatines of the first arch, and perhaps the quadrato- 
jugal and jugal, connecting the quadrate with the maxillae. 
The pterygoids are, as a rule, style-like bones, flat, however, 
in the penguins, and may or may not articulate with the 
basis cranii by means of the basipterygoid processes. The 
palatines may be completely separate from each other, or 
fused for a greater or less extent. They are broader or 
narrower, as the case may be. The lower jaw has a number 
of membranous ossifications ; these are the splenial, dentary, 
angular, supra-angular, and coronary. One or two additional 
elements may be present. 

With these general characters the skull of birds shows a 
considerable number of differences of minor importance in 
different families and genera. The first, perhaps, of these 
in degree of importance certainly in general estimation, if 
not actually so is the series of modifications of the avian 
palate which were worked out in detail by HUXLEY, and 
after him by PARKER, but which had been previous to 
HUXLEY'S ' well-known paper studied with some degree of 
success by CoRNAY. 2 

HUXLEY distinguished among the palates of birds the 
following principal modifications :- 

1. Dromceognathism. This characterises not only the 
ratites but also the tinamous, and has been indeed the 

1 In P. Z. S. 1867. See also paper by NEWTON, Ibis, 1868, p. 85, and reply. 
ibid. p. 357. 

- ' Considerations Generates sur la Classification des Oiseaux,' etc., Rev. 
Zool. Soc. Cuvicricnnc, 1847 ; see also HEEIIWAOKX, Bcitrage z. Kcnntn. >L 
Kiefergaumenapparates d. Vdgel, Diss. Inaug., Niirnberg, 1889. 


principal reason for the close association of these birds by 
subsequent writers. In these birds (see fig. 77) the vomer is 
broad posteriorly, and thrusts itself between the pterygoids 
and palatines on the one side and the basisphenoid rostrum 
on the other, and thus prevents their articulation. This is 
the typical dromseognathous state ; but there are certain 
modifications which will be described in detail later. The 
ostrich, for example, is only dromaeogiiathous in that the 
pterygoids and palatines do not articulate with the basi- 
sphenoidal rostrum ; for the vomer in this bird is short and does 
not reach back far enough to prevent (so to speak) the union. 
'2. Desmognathism. In a variety of birds belonging to 
many orders the vomer has either disappeared or is very 
small ; the two maxillo-palatine plates come into contact 
in the middle line, as, indeed, they do in the dromseognathous 
skull. As with all the types of skull to be enumerated, the 
pterygoids and palatines at the point of their union with 
each other articulate with the basisphenoidal rostrum. (This 
kind of skull is illustrated in fig. 78.) 

3. ScJiizognatliism. This type is almost as prevalent as 
desmognathism. The vomer, well developed, terminates, as 
a rule, in a point anteriorly. The maxillo-palatines, variable 
in size and shape, do not meet across the middle line with 
each other or with the vomer. (See fig. 79.) 

4. JEgitliognatliism. Found typically in ' finches ' and 
in passerines generally ; is very like the last type. The 
distinguishing character (fig. 80) is that the vomer is 
broad and truncated anteriorly, lying between the separate 
maxillo-palatines. The skull is thus ' schizognathous ' 
etymologically. To these four divisions Professor PARKER 
has added 

5. Saurogiiatliism. Exemplified in the woodpeckers. 
The maxillo-palatines are extremely slight, hardly extending 
inwards from the maxillne ; hence the skull is widely 
schizognathous. The vomers are delicate paired rods. 

As stated in the foregoing brief epitome of the characters 
of the several types of skull, the facts seem to differentiate 
the five types fully. ELLIOT COUES remarks of desmo- 


gnathism that ' it does not fadge so well as any other one of 
the palatal types of structure with recognised groups of 
birds based on other considerations.' This might be really 
said of saurognathism also ; for the woodpeckers are not so 
far removed from other picarian birds as the structure of 


t-'ni.i; premaxilhv ; J/r/>, maxillo-palatiue ; R, 
rostrum ; Vo, vomer ; J'l, palatine ; Pf, 
pterygoid ; *, basipterygoid process. 


FIG. 78. SKULL OF Dacclo (AFTER 

Ln, laoiymal. Other letters as iu previous 

their skull would lead us to believe. Neither are any of the 
subdivisions, except that of the dromseognathse, really satis- 
factory from the classificatory point of view. Their in- 
efficiency, however, is rendered harmless by the fact that 



they are in reality not such hard and fast distinctions ;is 
might be gathered from the foregoing abstract and from 
textbooks in general. 

PARKER has distinguished four categories of desmo- 
gnathism (a) perfect direct, the maxillo-palatines uniting 
below in the middle line ; (b) perfect indirect, maxillo- 
palatines separated by a chink in the middle line ; (c) im- 
perfectly direct, maxillo-palatines sutured together in the 





*, prefrontal (eetethmoid) ; <IH, quadrate : J/.r. 
Maxilla. Other letters as in previous figure-.. 

FIG. 80. SKULL OF Corvus (AFTEH 

middle line ; (d) imperfectly indirect, maxillo-palatines closely 
articulated with and separated by the median septomaxillary. 
There is also the exaggerated desmognathism (' double des- 
mognathisna ') of the hornbills, &c., where not only the 
maxillo-palatines but also the palatines are united across the 
middle line ; and finally the spurious desmognathism of 
certain segithognathous birds (Megalcema), in which the 


vomers are asgithognathous, but the maxillo-palatines are 

As of desmognathism so of aegithognathism, PARKER 
distinguishes four kinds 

Incomplete segithognathism (Hemipodes) is distinguished 
by the fact that the broad double vomer has a septomaxil- 
lary at each angle, which is only ' strongly tied ' to alinasal 

Complete Var. 1. The vomers are distinct from the often 
long alinasal walls and turbinals : a small septomaxillary 
appears on the angle of the alinasal cartilage, but does not 
run into it. 

Complete Var. 2. Here the vomers are grafted upon the 

nasal wall. 

Compound, where in an aegithognathous palate desmo- 
gnathism is produced by ankylosis of the inner edge of the 
maxillaries with a highly ossified alinasal wall and nasal 

Among the higher families the septomaxillaries are often 


It is clear, therefore, that a very narrow boundary line 
separates desmognathism in some of its forms from schizo- 
gnathism, and that there is a direct relationship between 
desmognathism and aegithognathism. The only type of 
skull which is really distinct is the dromaeognathous. 

In classificatory importance perhaps next comes the 
condition of the nasal bone. (TARROD distinguished birds 
into those with holorhinal and those with schizorhinal 
nostrils. These terms refer to the posterior edge of the bony 
nostril, which in one set of birds, the holorhinal, ends behind 
with a clear oval outline (fig. 81), or in the schizorhinal 
birds runs back as a gradually narrowing chink ; this latter 
arrangement is shown in fig. 82. In the holorhinal bird a 
straight line, drawn across the face from the posterior 
boundary of one nostril to that of the other, passes in 
front of the termination of the nasal processes of the 
premaxilla. Tt is not always the case that a line drawn 
similarly to that of the holorhinal birds passes behind the 



end of the premaxillary process, but it is generally so. In 
the schizorhinal skull it often appears, as in the typical 
schizorhinal cranes and charadriiform birds, as if the outer 
part of the nasal bone were a distinct bone ; for it joins the 


inner lamina at an angle. In the typical holorhinal skull, 
on the other hand, as, for example, in the Rallida?, the two 
parts of the nasal come smoothly together, leaving the clear- 

FIG. 82. SKULL OF Lants. 



cut, ovally contoured nostril. It is not, however, always easy 
to distinguish so clearly as can be done between Rallus and 
Grus. Thus GTARROD admits the schizorhiny of Funt.rii(t> 


and some other Passerines, in which the bony opening of 
the nostrils, although, as he figures it, rounded off at its 
termination, ends behind, or at least on a level with, the 
ends of the nasal processes of the premaxillse. In the same 
way an intermediate condition is offered by Thinocorus and 
Glareola, in which there is much the same kind of arrange- 
ment. But one of the most striking instances which have 
come to my personal knowledge is that of Chung a. Its 
near ally Cariama is, as correctly stated by GAEROD and 
others, holorhinal, which in view of its relationship to the 
cranes is unfortunate. But in Chunga it is clear that the 
holorhiny is secondary, being produced by a slight modifica- 
tion of schizorhiny. 

A careful examination of Chunga shows that the two 
parts of the nasal bone do not join evenly above the opening 
of the nostril, but that the outer descending lamina of the 
bone is divided for some little distance by a crack from the 
premaxillary portion, the two running up in close contact- 
so close that no actual space is left between them, only a 
line of junction to mark their original separateness. In the 
skulls of Cariama that I have examined there is no trace of 
this ; but, considering the nearness to each other of the two 
birds, it seems probable that it is merely disguised. This 
fact favours FURBRINGER'S idea that schizorhiny is more 
primitive than holorhiny, and is so far adverse to GARROD'S 
view that ' the schizorhinal disposition is most certainly one 
which is a secondary development upon the normal holo- 
rhinal nares.' 

It is clear too that the holorhiny of such a bird as Opis- 
thoconms, where the ossified alinasals produce the rounded 
margin of the bony nostril, cannot be accurately compared 
with the holorhinal nostril of a gallinaceous bird, where it is 
the nasals themselves that bound the orifice. 

The presence or absence of basipterygoid processes is 
another matter upon which some stress is usually laid from 
a systematic point of view. One assumes that the existence 
of these processes is the original condition and that their 
loss is secondary. The presence or absence of basipterygoid 


processes is most capricious. Thus among the limicoline 
series they are absent from the skulls of the gulls and more 
unexpectedly perhaps from the CEdicnemidse and Thino- 
coridae. Among the birds of prey the secretary bird and 
the American vultures have these processes, while the 
Falconidse have them not. The goatsuckers may be 
similarly divided into those without and those with basi- 
pterygoid processes. The TUBINARES, again, show variation 
in this respect, as do the trogons. It is therefore, in the 
first place, impossible to compare directly all birds which are 
without these processes, just as it is impossible to put 
together all birds without an ambiens. We may note, how- 
ever, that it is only among groups of birds which show a 
considerable range of structural variation that there is this 
variation of the basipterygoid processes. It is not, so to 
speak, lightly that they have gone. The reason for the 
assumption that the basipterygoid processes are primitive is 
their existence in reptiles and in such widely separated types 
as chameleons, pterodactyles, and Hatteria, and in addition 
I may point out that there is a significant correspondence 
between a primitive arrangement of gut and the presence 
of these structures. It cannot be said that every bird with 
basipterygoid processes has the most primitive arrangement 
of gut, but we do find both in the ratites, Chauna, the 
gallinaceous birds, the charadriiforrn birds, the owls, and 
the goatsuckers. The falconiformes (Haliaetus) are, it is 
true, an exception ; but it must be remembered that this 
group is one that has basipterygoid processes (Serpentarius, 
Catliartes), though they are absent in the true falcons. 

The principal variations exhibited by the cranium of 
birds, apart from those that have been already considered, 
concern the existence of supra-orbital bones, the existence or 
non-existence of occipital fontanelles, the marks of the supra- 
orbital glands, and the presence or absence of a hinge line 
between the skull proper and the face. 

The existence of supra-orbital bones in the form of a 
longer or shorter chain of ossicles was first pointed out by 
PARKER as a reptilian character of occasional occurrence. 



In the tinamous there are a series of these bones, which in 
Psopliia are reduced to a smaller number, and in the passe- 
rine Menura to three only ; 011 the other hand in Perdix 
they are again more numerous. It is possible that in such 
birds as Rhinochetus and (Edicnemus, in which the edge of 
the bony orbit is very sharp, the thinness of the edge is due 
to the fusion of a set of these bonelets with the true margin 
o the frontal. As will be seen, the existence of these bones 
is so rare as to render them of not great service from a sys- 
tematic point of view ; but it is only 
among birds which may be fairly 
on other grounds regarded as 
archaic that they are to be met 

The occipital fontanelles (fig. 
84) are mostly developed in water 
birds, though any connection be- 
tween them and the habit of the 
birds is at least not obvious. They 
are most general among the 
Limicolse and the duck tribe, but 
are found also in the Plataleidse 

and exist temporarily in the gulls ; they are also found in 
the flamingo, Gruidse, and among the auks. 

Almost the same remarks may be made of the impressions 
for the supra-orbital glands. They are very marked in the 
Limicolse, being more usual than the occipital fontanelles ; 
they are met with in the auks, divers, and penguins ; in the 
cranes and Plataleidse they are present, but not so conspicuous. 
The hinge line between face and skull is seen in its most 
fully developed condition in parrots, where the face is actu- 
ally movable on the head. But it is commonly met with 
elsewhere, particularly among the Anomalogonatae ; it is 
associated with holorhiny and with comparatively short 
nasal processes of the premaxillary. 

Apart from their relations to the vomer the palatines and 
ptenjyoids show some variations in structure. As to the 
pterygoids, the most prominent difference concerns the place 

Magellanica. BACK 



where the articulation with the basipterygoid facets occurs ; 
it is usually towards the middle of the bone ; but in the emu, 
in the Anseres and Galli it is quite at the end of the bone, 
the end nearest to the palatines. The palatines differ greatly 
in shape and in breadth. Among the Anseres and Galli the 
internal lamina which meet in the middle line above the 
basisphenoids are practically absent, being only represented 
by a more or less faintly marked ridge. In other birds these 
regions of the palatines are well marked, and may meet for 
a considerable space in the middle line, and, as in Steatornis f 
an anterior portion of each bone may come into contact in 
the middle line further forward, thus giving rise to what has 
been termed double desmognathism, the maxillo-palatines of 
course forming the other junction in the middle line in front 
of them again. In most birds perhaps the palatines are 
gradually rounded oft' behind, but in the Ardeidae and in 
Bliinoclietus, &c., they are as it were sharply cut across 
behind this edge, being at right angles to the long axis of the 
bone. In many Passeres and in some other birds there is a 
well-developed postero-lateral process of the bone, wilich has 
been termed transpalatine ; this, if it came into contact with 
the jugal arch, as it nearly doesiii some Passeres, would bear 
the strongest likeness to the transverse bone of reptiles. 
The palatines are occasionally, but rarely, fenestrate, e.g. 
A nous, Eurypyga. 

The vomer varies from complete absence (Colius) to very 
striking presence. When present there is every grade be- 
tween a thin splint and a broad flat bone, which in the latter 
case is often obviously formed of two lateral halves. These, 
however, do not appear to remain completely separate except 
among the Pici (q.v.), where they are smallish splints. But 
not only among the ostrich tribe, but in Hesjn'raniis and 
passerines the vorners are double in the young. A series of 
small bones lying between the palatines, and called interpa- 
latiiies and septo-maxillaries by PACKER, may be regarded 
as belonging to the vomerine series. They are unpaired 
ossicles which continue, though with a hiatus, the vomer 

L 2 


The degree of development of the lacnjmals and the 
ectethmoid processes varies greatly among birds, and is at 
times of use for systematic purposes. In some birds, as, for 
example, among the cranes, there is no junction between 
those bones ; in the Charadriidge, on the other hand they 
form a complete ring ; in Pterocles, &c., the two are firmly 
blended into a square plate of bone which bounds the orbit 

The lacrymal is occasionally joined to the palatine or to 
the jugal by a small independent ossicle, which PARKER has 
termed the uncinate bone, and thinks to be the homologue 
of the anterior connection of the palato-quadrate arch with 
the skull in the tadpole, &c. This bone is so variable in its 
presence (e.g. Cariama, Tubinares) that it can hardly be 
regarded as of much systematic importance. The last matter 
to which we may refer as of classificatory importance is the 
form of the quadrate, which Miss WALKER has shown to 
vary much and characteristically in different groups. It had 
been long known that the single-headed articulation with 
the skull was a character of struthious birds, excepting 
Apteryx, and of Ichthyoniis, and to a less extent of gallina- 
ceous birds and ducks. 

The value of the base of the skull in classification has 
been continuously debated since the facts were first so clearly 
set forth by HUXLEY. 

From the complications introduced into the originally 
simple series of modifications of the skull instituted by 
HUXLEY, by PARKER, and from the varied criticisms of fact 
and conclusion of a classificatory kind based upon fact, we 
may disentangle one conclusion that many ornithologists 
will agree with that is, the more lacertilian character of 
the skull in the struthious than in other birds. In them the 
palatines are borne off from the basisphenoidal rostrum by 
the vomer (with the exception of Strutliio, in which, however, 
the palatines are still remote from the rostrum), and the 
general disposition of these parts is, as HUXLEY wrote, 'more 
lacertilian than in other birds.' Furthermore in the stru- 
thious birds the double character of the vomer is more 


universally retained than in other birds. As T. J. PARKED 
has shown in Mesopteryx, Anomalopteryx, and (according to 
OWEN) Dinorni-s torosus, and finally in the young of Emeus 
crassus, there are distinct paired voiners. The strong anterior 
and posterior bifurcation of the voiner in living Struthionidse 
is an indication of the partial fusion of the primitively 
separate halves of the single vomer. A double vomer, 
however, is not a feature of the struthious birds alone; in 
the young Hesperonu's the same occurs ; in the woodpeckers 
(hence Saurognathse in PARKER'S nomenclature) there are two 
distinct vomers, while PARKER has recently committed him- 
self to the general statement that ' in most birds that have 
a large or wide vomer it is double at first.' The inference 
is that the single thread-like vomer of many birds is in a 
degenerate condition. In any case a large vomer, double or 
single, and a lacertilian palate generally mark all the stru- 
thious birds, and justify their generally recognised position 
somewhere not far from the root of the avian series. 

As for desmognathism and schizognathism, Professor 
PARKER has aptly remarked : ' The use of such a taxonomic 
character as desmognathism or schizognathism is very 
extensive in some groups and very limited in others ; and 
there is no sharp line of demarcation between the two. The 
most lacertilian palate for openness is that of the woodpecker ; 
the most modified by intense ossification is that of the 
toucan ; yet these two types, each specialised to the utter- 
most, have a postcephalic skeleton, not indeed identical, but 
extremely similar.' 

That the toucans and woodpeckers are exceedingly near 
akin is also shown by many other features of their organisa- 
tion (see below). HUXLEY claimed for his Schizognathse 
that they were a natural group, but hardly claimed so much 
for the Desmognathfe. He admitted, however, that Cariama, 
a schizognathous bird in the totality of its organisation, had 
a palate approximating to the desmognathous. The Cracidae 
and Rhinochetus are in the same anomalous position. The 
Trogonidse, whose nearest allies are the desmognathous 
birds, have a schizognathous palate (as was pointed out by 


FORBES), while the goatsuckers show both desmognathism 
and schizognathism among the members of the group. On the 
other hand schizognathism is not on the taxonornic side 
sharply marked off from aegithognathi'sm. HUXLEY'S defi- 
nition of the vomer in schizognathons birds was that it is 
' pointed in front ; ' this is not the case with many Schizo- 
gnatha3, e.g. Hcematopus, Numenius (cf. below). And, on the 
other hand, PARKER showed that another peculiarity of the 
eegithognathous skull, the union of the vomer with ossified 

O O 

alinasals, was also found in the ' Turnicomorphae,' a group 
which for other reasons should be placed among HUXLEY'S 

It appears, therefore, undesirable to lay too much stress 
upon the modifications of the palate, as seen in the three 
groups just discussed, as a basis of classification. 

HUXLEY defined desmognathism as follows : ' The 
vomer is often either abortive or so small that it disappears 
from the skeleton. Where it exists it is always slender and 
tapers to one point anteriorly. The maxillo-palatines are 
united across the middle line either directly or by the 
intermediation of the ossifications on to the nasal septum.' 

This definition applies perfectly well to the Anseres and 
Palamedese, but not to all other groups of birds. 

But there is the form of desmognathism found in the 
Steganopodes and in the owls and Accipitres. In the 
former group there is no desmognathism (in the sense of 
HUXLEY) except in Pelecamis. The maxillo-palatines of 
Phaethon and Fregata (see below) are perfectly free ; in 
front of them the palate is complete, but that completeness 
is formed by a union of lateral extensions of the maxilla? 
which are distinct from the rounded maxillo-palatines. In 
the cormorant, which is admittedly an ally of Phaethon, there 
appears at first sight to be a true desmognathism, a fusion, 
that is to say, of the maxillo-palatines. The plates of bone 
in question are beneath the ends of the palatines, but, instead 
of running horizontally in the same plane as the palatines, 
they run obliquely upwards (when the skull is regarded from 
below) . If it were not for Fregata, these bones might be 


looked upon as the homologues of the maxillo-palatines of 
Pliaetlion. But in Fregata both structures are present. 
Coexistence clearly disproves homology ; we must, therefore, 
place the ' desmognathism ' of Phalacrocorax upon an 
entirely different footing from that of the Anseres, where 
the maxillo-palatines unite in the typical fashion across the 
middle line. 

In addition to these forms of desmognathism it appears 
to me that we should distinguish that of the American vul- 
tures, where there is no union at all of the maxillo-palatines, 
but only of the alinasals. 

In fact there appear to be three ways of bridging over the 
palate which may be termed desmognathism, though the 
word becomes merely descriptive of a condition and not 
necessarily indicative of affinities ; these are 

(1) Union of maxillo-palatines (more or less complete). 

(2) Union of alinasals. 

(3) Union of maxillae in front of the maxillo-palatines. 
These three are usually combined in various ways ; e.g. 
In hornbills there are all three methods of union ; in 

some Steganopodes only numbers 2 and 3. 

Brain and Nervous System. 

The brain of recent birds chiefly differs from that of 
reptiles by the large size of the hemispheres, the large size 
of the cerebellum, and by the fact that, owing to the great 
development of these two regions, the optic lobes are pushed 
to one side. T. J. PARKEE found in studying the develop- 
ment of Apteryx that the latter feature is acquired during 
development, and that the optic lobes have originally the 
reptilian position. The extinct cretaceous birds had brains 
with smaller cerebral hemispheres and with larger optic lobes 
than existing birds, and were in these particulars more repti- 
lian. On the other hand, if the cast present in the London 
example of the Arcliceopteryx be really, as it was first surmised 
to be by Sir JOHN EVANS, the brain of the bird, it was less 
reptilian than Hesperornis, a fact which may possibly 


suggest the inference that the brain of the cretaceous bird 
furnishes an example of degeneration rather than of the 
retention of an archaic character. 

The hemispheres, although smooth in the majority of 
birds, show faint indications of what may correspond to 
furrows in some others. In the duck, for example, as 
figured by GADOW (after BUMM), there is a central raised area 
marked off by a furrow from the surrounding parts of the cere- 
brum ; faint traces of the same occur in Buteo vulgaris. The 
weight of the brain as compared wdth that of the whole body 
has been studied by several observers, and according to their 
results the Passeres and parrots take the highest place. 
But GADOW justly remarks that ' the attempts to sort birds 
according to the proportion of brain to body have led to no 
practical results, chiefly because the variable conditions of 
fat and lean subjects have not been considered.' 

The brachial and lumbar plexuses, particularly the 
former, have been studied in a large number of birds ; 
FUEBRINGEE has published a quantity of drawings of the 
former, but no classificatory results of reliability appear to 
follow from the facts collected with so much diligence. The 
brachial plexus varies in position and in complexity. The 
former variations are largely correlated with the varying 
length of the neck ; thus in Columba the first spinal nerve 
entering into the plexus -is the tenth, in Pliocnicopterns the 
seventeenth. The number of nerves which together form 
the plexus varies from only three in Biicorvus to six in 

The Eye. The eye of birds presents many resemblances 
to that of reptiles. The minute structure of the retina 
presents many points of similarity, as also the ring of ossifi- 
cations in the cornea, and the pecten. The latter is a folded 
process of pigmented tissue which projects into the vitreous 
humour through the choroidal fissure, which is in the embryo 
the gap left between the edge of the optic cup and the lens 
on one side. The pecten offers differences in various birds 
upon which perhaps some little stress can be laid. In 
Apterijx it is entirely absent. There are very few folds in 


the owls (4 to 7) and in the goatsuckers ; this might lead, in 
connection with the total absence of the pecten in Apter//.r, 
to the conclusion that the less or greater development of the 
organ had some relation to nocturnal or diurnal habits. 
But the existence of only four in the emu seems to throw 
some doubt upon this suggestion. Among the passerine 
birds the largest number (raven 30) of folds in the pecten 
is found. 

Certain small ossifications in the cornea near to the 
entrance of the optic nerve seem to be peculiar to the 
Passeres and to certain picarian birds. The eyes have both 
lacrymal and Harderian gland ; the eyeball is moved by four 
recti and two oblique muscles ; the membrana by two 
muscles, the quadratic and fhepyramidctl-is, both innervated 
by the sixth cranial nerve. 

The Ear. Birds have no external ear (concha), but in 
many a flap forming a valve projects into the meatus from 
the outer margin, a state of affairs which recalls the condi- 
tions found in the crocodile. This outer ear is especially 
well developed in the owls, in which birds also the ear region 
of the skull is often markedly asymmetrical. 

The inner ear has the three semicircular canals of all 
higher vertebrates, but the cochlea is not coiled. The audi- 
tory ossicles consist of a single structure, partly bony and 
partly cartilaginous, called the columella. 

The Affinities of Birds 

In considering the relationship of birds to other verte- 
brates it is probably safe to leave out of consideration the 
mammalia and the amphibia. Points of likeness have, it is 
true, been urged in favour of the latter view of an affinity 
between birds and amphibia by the late Professor PARKER ; 
but apart from warm-bloodedness and the resemblance 
of some of the more simple forms of feathers to hairs 
there is nothing to be said on behalf of a kinship 
between birds and mammals. As to the likeness with 
amphibians, it is possible that the divergence of birds from 


the reptilian stem was at a time when the characters of the 
amphibian had been incompletely thrown off, and at a time 
also when the mammals diverged on their own path from a 
point near to that whence the birds took their origin. The 
general belief is in the origin of birds from some reptile stem, 
but there is not an absolute agreement as to precisely which 
group of reptiles birds are most nearly akin to. The researches 
of MARSH and HUXLEY, besides those of COPE, SEELEY, 
HITLKE, and some others, have led to a general acceptance of 
a nearer kinship with the dinosaurs than with any other- 
group of reptiles. In considering the question, then, which 
forms the subject of the present chapter, we shall commence 
with the dinosaurs. The dinosaurs, ranging in size from 
vast creatures of 70 or 80 feet in length to a diminutive reptile 
half the size of the domestic fowl, are first known from 
the Trias, persisted though the Jurassic and finally came to 
an end in the Cretaceous epoch, later than which no 
remains have been found. So far as we are aware birds 
came into existence in the Jurassic period ; hence there is 
no anachronism in considering them from the dinosaurian 

It was formerly held that birds antedated the Jurassic 
period ; for some of the celebrated tridactyle footprints in 
the sandstone of the Triassic period were put down to birds. 
It seems, however, to be now fairly certain that those foot- 
prints are of dinosaurs. Still with so specialised a form as 
Archceopteryx certainly was, and as Laopteryx probably was 
in the Jura, it would not be surprising to meet with genuine 
avian remains in the Trias. But even then there are 
undoubtedly dinosaurs belonging to that period, so that the 
question of relationship would resolve itself into a common 
origin, not a derivation of birds from dinosaurs. 

The part of the skeleton in which most resemblance is 
shown between birds and dinosaurs is the pelvis. The 
dinosaurian pelvis consists of apparently three elements, like 
that of birds, but the pubis is an L-shaped bone, constructed 
of two pieces, one directed forwards and the other back- 
wards and parallel with the ischium of its side. The latter 


is generally regarded as the homologue of the pubis of birds, 
while the forwardly directed half of the bone is considered 
to be the equivalent of the pectineaj process. The alterna- 
tive is to look upon the prepubisof the dinosaur as the pubis 
of birds, and the postpubis as having disappeared altogether 
in them. In this case the crocodile will be an intermediate 
form ; for in this reptile the prepubis is the 'pubis,' while 
the postpubis is represented by an inconspicuous process 
upon the pubis. The former alternative commends itself to 
us. There are birds in which the pectineal process is 
practically absent ; in others, as Apteryx, Geococcyx, &c., it is 
large ; on the other hand there are dinosaurs, e.g. Laosaurus 
consors, in which the prepubis is generally reduced to a 
length of not more than one-third of the postpubis (pubis 
of birds). The opposite extreme is reached by Triceratops, 
where it is the postpubis which is less than a third of the pre- 
pubis, and Ceratosaurus, where it is even further reduced. 

It is not, however, only in the pubes that the pelvis of 
the dinosaurs is like that of birds. In those reptiles the ilia 
were extended forwards and backwards, as in birds, and in 
Ceratosaurus at any rate the three bones were all firmly 
ankylosed, as in all birds save Arch&opteryx, where the 
separation of the bones conforms to what is found in the 
vast majority of the dinosaurs. Laosaurus, which has just 
been mentioned, is one of the most birdlike of dinosaurs. 
' The two species of the genus first described by the writer,' 
remarks Professor MAESH, ' show these avian features best 
of all, and it would be difficult to tell many of the isolated 
remains from those of birds.' Of the cretaceous dinosaurs 
the same author observes, ' Others were diminutive in size, 
and so birdlike in form and structure that their remains 
can be distinguished with difficulty, if at all, from those of 
birds.' OrnitJiomimus, as its name denotes, is one of those 
especially annectent dinosaurs. In this genus the third 
metatarsal is crowded backwards behind the second and 
fourth, as in many birds. But the dinosaurian metatarsals 
which are most strikingly like those of birds are of Cerato- 
saunts, which MAESH has figured side by side with those 


of a penguin, where the bones show greater traces than 
in any other birds of their distinctness, and are furthermore 
shorter than is the rule. In dinosaurs generally those bones 
are separate, but not in Ceratosaurus, where the degree of 
fusion is almost exactly that of Aptenodytes, 

Furthermore in some dinosaurs, as in birds, the inter- 
medium is prolonged upwards as the ascending process of 
the astragalus. 

The proportions of the long bones of the hind limb are 
distinctly birdlike in some dinosaurs. In Laosaurus, for 
example, the femur, as in birds, is shorter than the tibia, the 
reverse occurring in most forms. In the same animal and 
in others the fibula is commencing to degenerate ; it is 
decidedly smaller than the tibia. 

In their skulls the dinosaurs show no marked approxima- 
tion to birds ; there are nevertheless one or two features 
which may be remarked upon in this connection. The 
earliest known forms from the Trias have perhaps the most 
birdlike forms of skull. MARSH comments upon the light- 
ness and avian appearance of the skull of Anchisaurus ; it 
has moderately developed basipterygoid processes instead of 
those of such great length that are apt to characterise the 
dinosaurs. The great extension backwards of the premaxil- 
laries in some dinosaurs is an avian characteristic ; this is 
seen especially well in Diplodocus and Claosaurus. In the 
former animal, as in some others, the tw r o vomers diverge 
widely posteriorly, as in many birds ; and in the restoration 
of the under surface of the skull in this dinosaur the vomers 
have a very birdlike appearance. 

Finally the height at which the transverse processes of 
the vertebrae are borne seems, as HUXLEY has pointed out, 
to suggest birdlike respiratory organs, while the hollowness 
of many of the bones in many dinosaurs points in the same 
direction. It is, however, undoubtedly in the pelvis and 
hind limb that the most striking likenesses to birds are 
shown by the dinosaurs. It has been attempted to put this 
down merely to bipedal progression. ' It may be said,' 
remarked Professor HUXLEY, ' that all birds stand upon 


their hinder feet, and that, as the Ornithoscelida did the same, 
the resemblance of structure arises from a resemblance of 
function. But I doubt if the majority of the Dinosauria 
stood more habitually upon their hind limbs than kangaroos 
or jerboas do ; and, unless there was some genetic connection 
between the t\vo, I see no reason why the hind limbs of 
Ornithoscelida should resemble those of birds more than they 
resemble those of kangaroos.' In addition to this it may be 
pointed out that Hallopus, which appears to have been very 
probably a leaping dinosaur, has not the specially ornithic 
form of limb ; it has large pubes and no postpubes. 

A recent description by Mr. E. T. NEWTON 1 of the skull, 
brain cast, and cast of the auditory organ in a pterodactyle, 
Scaphogiiathus Purdoni, shows certain most interesting 
resemblances between the pterosaurians and birds. It is 
possible that this pterosaurian, like Pteranodon, possessed a 
homy beak and no teeth. But the presence or absence of a 
beak or of teeth is no more distinctive of birds (cf. Archceo- 
pteryx) than of reptiles. The skull shows more positive 
points of likeness. In the first place, the bones of the 
pterodactyle cranium are early ankylosed and well anky- 
losed, being in this particular avian and not lacertilian. 
The large size and backward extension of the single pre- 
maxillary bone ( = two fused premaxillae) agree with that of 
birds and contrasts with that of lizards. The palate shows 
also certain interesting resemblances, more especially to 
both emu and cassowary. As in the struthious birds and in 
lizards also the palatines are borne off from the middle line 
by the pterygoids ; the latter bones, moreover, as in the emu, 
articulate at their posterior ends with both quadrate and 
basipterygoid processes. The vomer too is birdlike in being 
pushed backwards, owing to the extent of the premaxillse, 
and in being thin, apparently single and bifurcate posteriorly. 
Other general resemblances in the skeleton are the develop- 
ment of air cavities in the bone, the large size of the orbit 
which may, however, in the pterodactyles have had some 

1 ' On the Skull, Brain, and Auditory Organ of a New Species of Pterosaurian,' 
Phil. Trans. 1888, B, p. 503. 


relation to nocturnal habits and the presence of a keel upon 
the sternum. 

It is possible, though far from certain, that Scaphogna- 
thus had not that characteristic reptilian bone, the os tnnis- 

Finally, the general shape of the scapula and the angle 
that it makes with the coracoid are birdlike in the pterodac- 

The brain of the pterodactyles seems also to have pre- 
sented avian characters ; the optic lobes are pushed aside by 
the large cerebellum, which had well-developed floccular 
lobes. In the reptile's brain the optic lobes intervene 
between the cerebrum and cerebellum. 

The pelvis of the pterodactyles has some likenesses to 
that of birds. The ilium has an extension in front of as w r ell 
as behind the acetabulum ; and, if the opinions of SEELEY 
are to be agreed to, there is a rather backwardly directed 
pubis, more or less fused with the ischium, and a long and 
thin forwardly directed piece, the prepubis of dinosaurs and 
the pectineal process of birds. 

The main difficulty, however, in the w r ay of comparing 
pterodactyles and birds is in the fact that both can fly, and 
that each has acquired the power of flight by a different 
method. Having acquired the power of flight it seems clear 
that certain of the points of resemblance between them may 
easily be due to that mode of life and may have been inde- 
pendently arrived at. 


PROFESSOR NEWTON'S article ' Ornithology ' in the ' En- 
cyclopaedia Britannica,' and the preliminary sketch of Dr. 
GADOW in Bronn's ' Thierreich,' contain a digest of, and 
criticisms upon, the main schemes of classification of this 
group which have as yet appeared. I shall, therefore, refer 
the reader to those works for the history of the subject. 
There can be no question, in my opinion, that birds must 
be primarily divided into two great divisions, viz. Saururae 
and Ornithurae, the first to contain Arch&opteryx and possibly 
Ijaopterijr, the latter the rest of birds, both living and extinct. 
As to the Ornithurae, while there is a very general agreement 
with the main subdivisions no one probably will quarrel 
seriously with the divisions adopted in the present work 
no one has (to my mind) satisfactorily arranged the different 
groups with reference to each other. More especially does 
it appear to me that the majority of ornithologists are in 
error concerning the position of the picarian and passerine 

In considering a scheme of classification it is clear that 
we must bear in mind indications of the descent of birds. 
Existing schemes have savoured too much of a mere sorting 

O O 

by combining in various ways characters which are dis- 
tinctively bird characters. However unsuccessful the con- 
struction of phylogenetic trees has been, it is abundantly 
plain that that must be the line to take in arranging a group 
scientifically. It follows, therefore, that in sketching, at 
any rate, the main outlines of our scheme attention must 
be paid only, or chiefly, to those characters which birds 
have inherited from their reptilian ancestors. 


Now this at once lands us in a difficulty, which has been 
too lightly regarded by many systematists. Phylogenetic 
schemes used to be boldly linear, and even so recently as 
the attempt of FUEBBINGER the family tree savours a little 
too much of the linear arrangement. Now the imperfect 
remains of birds that have come down to us from tertiary 
times show that the modern types of birds were fully 
differentiated even then in addition to a few extinct forms, 
such as Odontopteryx toliapicus (if this benot asteganopod). 
But beyond that point there is the most scanty record of 
bird life, limited to the Cretaceous Ichthyornithidae and 
Hesperornithida?,, with a few obscurer forms, and to the 
Jurassic Archceopteryx. So emphatically were all these 
creatures birds that the actual origin of Aves is barely hinted 
at in the structure of these remarkable remains. Moreover, 
at least in the case of Ichthyoniis, they depart fully as 
widely from any bird with the required ' mixed ' characters 
as any living group, while Hesperornis can with safety be 
relegated to the neighbourhood of the existing divers. We 
get, therefore, no help whatever from the Cretaceous birds, 
and, if any, only the scantiest assistance from Archteopteryx, 
in determining what are archaic characters in birds. There 
are no criteria by which we can assert with any degree of 
safety the relative positions of this and that existing group ; 
nor has the study of the comparative embryology of birds 
as yet advanced sufficiently far to give any results, except 
in isolated characters ; such indications are the relatively 
primitive character of the basipterygoid processes, at least 
in certain groups ; for the gulls which are without them 
when adult, have them as young chicks, as have in the 
adult condition most of their near allies, the Limicolse. It 
may, therefore, so far be inferred that the gulls are a 
modification of the limicoline type and not vice versa. 

It would be perhaps held that any type in which a 
number of undoubtedly reptilian characters had survived 
would be on a lower level of organisation than other types 
in which fewer or no such characters could be discovered. 
But the few specially reptilian features in the organisation 


of birds have, so to speak, been distributed with such exceed- 
ing fairness through the class that no type has any great 
advantage over its fellows. PARKER has collected together 
some of the reptilian survivals, and to his series a few others 
may be added. The rudimentary organ of Jacobson found 
by T. J. PARKER in Apteryx is a suggestion of the reptile ; 
but we do not know enough about the development of other 
birds (except the struthious, where PARKER found the same 
rudiments himself) to lay much weight upon the discovery 
as indicative of the low position of Apteryx in the avian 
system. The supra-orbital chain of bones is also on the 
same grounds an archaic character ; but they exist in 
such widely different types as tinamous, Psophia, Menura, 
and quails. The double vomer is reptilian ; this bone is 
double or nearly so in struthious birds, in Hcsperornis, 
woodpeckers, and even in passerines and other types with a 
broad vomer. PARKER has compared the ' os uncinatum ' 
with the anterior suspensory cartilage of the tadpole's jaw 
apparatus ; this is found in so many and such various forms 
as Cariama, Fregata, Tubinares, Musophagida?, Steatornis, 
&c. Basipterygoid processes are distinctly reptilian, found 
as they are in so many forms of reptiles. But among birds 
they occur in nearly every big group, and are therefore most 
undistinctive. The pectineal process, if it is invariably as 
T. J. PARKER says it is in Apteryx the joint product of pubis 
and ilium, is not exactly comparable to the supposed corre- 
sponding process of the dinosaurian pelvis ; but in any case it is 
found in Geococcyx and some other birds far away from the 
Struthiones and the tinamous. A large number of vertebrae 
in the tail is reptilian ; but not only Arcliceopteryx but also 
the swan has a long tail. Opisthocoelous vertebrae are 
found in the Alcse, penguins, and gulls, not to mention the 
darters and parrots. 

As to the viscera, HUXLEY showed the close likeness be- 
tween the various membranes which divide the coelom and the 
corresponding membranes in the crocodile, and I endeavoured 
to show that the ostrich is not in these particulars more 
reptilian than many other birds. The partial persistence 



of the left aortic arch is an approach to the reptile ; but birds 
belonging to the most diverse orders have this arch left in 
varying degrees of perfection ; so that no stress can be laid 
upon this anatomical fact as a mark of low position. It has 
been pointed out that the Struthiones are unlike other birds 
in the absence of a syrinx ; and in the absence of this 
specially bird organ they approach so far to the lower 
forms. This is not, in the first place, true of all the Struthiones, 
for Rhea has, as has been pointed out, a syrinx fully as typical 
as that of most birds, while the American vultures and even 
the storks have nothing in the way of a specialised syrinx. 
In fact, without going into further detail, it seems impossible 
to select any existing group of bird which is distinctly more 
reptilian than any other. 

Since positive characters appear to fail us in discriminat- 
ing between the relative positions of the several groups of 
birds, it seems to be not unreasonable to turn for light to 
negative characters. Birds as birds have many peculiarities 
of organisation, which are impossible in other animals ; for 
example, patagial muscles cannot exist without a patagium 
to contain them. It may therefore be permissible to draw 
with caution some inferences from the absence or simplicity 
of certain peculiarly ornithic structures which, it appears 
obvious, must have originated within the class. The lower 
types will surely possess fewer of these essentially ornithic 
organs or modified organs. 

There is a general belief in the modified character of all 
the birds which GARKOD placed in his subclass Anomalogonatae. 
Nevertheless there is something to be said in favour of their 
primitive nature. Without absolutely urging the acceptance 
of this view, it may be useful to refer briefly to certain 
reasons which might be alleged in support of such a placing 
of the Pico-Passeres. Their small to moderate size is to 
some extent an argument. The most ancient mammalia 
and reptiles are small as compared with some of their later 
and more modified representatives. Universal distribution 
is another argument, as is possibly chiefly arboreal life. In 
anatomical structure we find that many essentially ornithic 


characters have not yet put in an appearance, or have done 
so only to a small extent. There is no member of the 
group, wide though it is, in which there is an ambiens, a 
special bird muscle. It may be that GADOW'S discovery of 
a small independent slip of the rectus femoris, which he 
interprets as a rudimentary ambiens, is really the beginning 
of this characteristic muscle. The remarkable peculiarities 
of this muscle seem to forbid the notion that it is the direct 
descendant of anything reptilian. And we have the un- 
doubted fact that, apart from the possible rudiment already 
referred to, it is present in no pico-passerine bird. If it 
had disappeared in them there would be here and there a 
rudiment left. But nothing of the kind has been hinted by 
GARROD and FORBES, who between them dissected so many 
of these birds, and who would have been especially on the 
look-out for such a point. I should therefore be disposed 
to disagree at once with GARROD 's opinion that those birds 
which have lost the ambiens ' may be set down as having 
possessed the muscle in their ancestral form.' The ambiens 
is so purely a bird muscle, though it may doubtless have its 
homologue among reptiles, that one cannot but think that it 
was acquired within the class ; and the facts discovered by 
MITCHELL (see above) entirely support this way of look- 
ing at the matter, and indeed suggested it. As to the 
muscular system of the wing, a highly characteristic muscle 
is the expansor secundariorum ; this was supposed for 
some time to be absent in the group under consideration, 
but it has been found to occur in some of them. In the 
majority of those in which it does occur its structure is 
decidedly more rudimentary than in some of the Homalogo- 
natse. It is true that FURBRINGER regards this muscle as 
the abortive remnant of a reptilian muscle. But this state- 
ment cannot be made about the patagial muscles, which are 
essentially ornithic. . 

Now it is noteworthy that, with the exception of the 
colies, not a single bird referable to the great group of Ano- 
malogonatae has a biceps slip, while in the majority of them 
the tendon of the tensor brevis is exceedingly simple, being 

11 '2 


without the numerous subdivisions so often observable in 
the same muscle in the (hypothetically) higher birds. Nor 
is there the patagial fan, the junction between the tendons 
of the longus and the brevis, a character, again, so frequent 
among the larger birds. The complications in this case 
seem to be much more likely an effect of specialisation than 
that the simple conditions observable in the picarian and 
other allied birds should be due to a process of degenera- 

The simplicity and relative shortness of the gut is a 
matter, which is perhaps of importance in this connection. 
The average relative length of the gut is much less on the 
whole in the Pico-Passerines than in any other group. There 
are, of course, some startling exceptions, but the general 
statement holds. As to the caeca, it must be confessed that 
they are as a rule small or absent. But the Coraciidse and 
the Todida? are exceptions. The fact that both lobes of the 
liver are frequently in this group of birds enclosed in separate 
compartments, separated from the subomental space, seems 
to me to be a vestige of a condition such as that which is 
found in the crocodile. On the other hand certain of the 
organs of the body show great variety and specialisation ; 
particularly is this the case with the syrinx ; but to find 
most of the organs of the body in a primitive condition, 
while others are greatly specialised, is precisely what is so 
often found in what are believed to be archaic types. 

There can be no doubt that the Archesopteryx, far 
though it may have diverged from the ancestral stock, has 
retained more of the reptile than any other form known to 
us. One or two of the characters are shared by that large 
assemblage of birds which has been termed the Ano- 

In the first place the structure of the foot of fheArchceo- 
ptenjx is that found in passerine birds. That the primitive 
bird was arboreal seems likely, and it is not surprising to 
find that this mode of life has led to various specialisations 
in the foot, such as we find in the hornbills, &c. The 


. I rchceoptenjx has the smallest number of cervical vertebra) 
of any bird (ten), a fact which recalls the nine vertebrae or so 
of the cervical region of lizards and crocodiles. Now, among 
other recent birds there are none which have a smaller 
number than that possessed by certain passerines. Fifteen 
is perhaps the average number of these vertebra?, among 
birds ; but among passerines the low number of thirteen is 
to be met with. Nearly all the Anomalogonatse are holorhinal, 
as was the ArchcBopteryx. It is doubtful, however, whether 
this particular fact advances my argument, as there are 
reasons (see p. 144) for considering the schizorhinal arrange- 
ment to be the older, and for looking upon the holorhinal as 
a derivative. 

There remains for consideration the large assemblage of 
birds which, taken together, correspond to the HomalogonataB 
of GARROD. In a preliminary way we may regard, as I 
have pointed out above, four characters at any rate as 

These are the presence of basipterygoid processes, the 
possession of two carotid arteries and of the fifth cubital 
remex, and the simplicity of the intestinal coils. The only 
birds which have all of these are certain Galli and certain 
Turnices. Allowing for the degeneration of the wing, the 
struthious birds may be referred to a nearly equally low 
place in the system. In many groups, however, we find a 
near approximation to this presumably primitive condition. 
Thus among the anserine birds Palamedea is deficient only 
in the fifth remex. Among ' Gralla? ' it is only the same 
character that is wanting in certain forms to complete the 
four requisite characters. And some of them e.g. Cariama, 
Psophia have this elsewhere missing feather, though those 
particular forms have not some of the other characters. 
Opisthocomus, the cuckoos, and Musophagidas are not far 
off from the base of the series, while the trogons, if they 
had both carotids, would be among the (hypothetically) 
lowest groups. The facts in question may be thus tabu- 
lated : 



( ';mitiil< linsijit. Proc. 

Kit'th Jti-inex 


Colyinbi . 

2 or 1 


Sphenisci . 


4- ? 


Tubinares . 

2 4- or - 

Complex . 

2 or 1 




Simple (in 



2 + 



2 + 


Accipitres . 

2 + or - 




2 4- 



T/irnices . 

2 or 1 4- 




2 or 1 + 

4- or - 




+ (Podica) 



2 _ 

+ (Psoplua, &c.) 



2 or 1 


Limicolce . 

2 + or - 

Simple (some) 


2 + 


Columbse . 

2 + or - 

+ or - 



2 or 1 







2 _ 




2 _ 




2 or 1 + 

4- (?) 



2 or 1 



2 + 



2 4 or - 


Pico-Passerrs . 

2 or 1 - or 4- 

4 or - 

Simple (?) 

I have italicised in the above list those groups which 
come near to the supposed primitive condition ; and it will 
be observed that those groups which are italicised are enough 
to account for the ancestry of the rest. The main difficulty 
is perhaps presented by the Cuculi ; but it may well be that 
that group will ultimately prove to have a simple intestinal 
tract. It will be observed also that the groups in question 
comprise representatives of all the five large divisions of 
existing birds admitted by GADOW, and of all the corre- 
spondingly large divisions of FUEBEINGEE. Incidentally, 
therefore, I find myself in reassuring agreement with those 
authorities. In the systematic part of this work I have to 
some extent discussed the mutual affinities of these different 




Definition. Generally quincubital. Ambiens and accessory femoro- 
caudal always absent ; biceps slip rarely present. Cervical 
vertebrse, 13-15. Atlas generally perforated by odontoid process. 
Skxill holorhinal. 

This group of birds is equivalent to the similarly named 
group of G-ABROD, with the sole addition of the Striges. The 
total absence of the ambiens, even of all traces of that muscle 
(see p. 95), is to my mind a sufficient reason for bracketing 
together all these birds. I am of opinion that the ambiens 
is not degenerate in them, but that it has not yet appeared. 2 
It must be admitted that there are not many other charac- 
ters that run through the whole group. There are, never- 
theless, certain peculiarities of structure that are confined Or 
nearly confined to this group. Thus, with the sole exception 
of the parrots, a forked manubrium sterni is a peculiarity of 
the Anomalogonatse ; so too, with the same exception, is the 
presence of a cucullaris propatagialis. Again, it is only here 
and in the parrots that the syrinx has so complicated a 
musculature. The very prevalent shortness of the intestine 
is a fact (not, it is true, without exceptions) not to be ignored 
in considering the claims of this group to existence. The 
feet are nearly always anisodactyle or zygodactyle, there 
being but few other birds not referable to this group in which 
that structure of foot is to be found ; and those groups will 
be treated of later as possible allies of the present. Powder- 
down patches are exceedingly rare in this group, and but 
few possess the expansor secundariorum. 

I allow nineteen separate families of this group (whose 
main characters are given in the table), of which some may 
be united more closely than others. 

1 SEEROHM, ' An Attempt to diagnose the Pico-Passerine Group of Birds,' &c., 
Ibis, 1890, p. 29. 

'-' A case which appears to contradict this statement is dealt with on p. l(j:-5. 



"3 a 




i en .S 

S ^u en 

S ^ "o 

*^3 x t3 ^ 

S < S S 

o ;. o o 

cc g co ui 


'S . 









America, Ai 




-S i ,-= J 

C a c3 r3 <J 
> cppS '_ P, 

> * 53 33 

2 'S -^ S g 


5 g < g 3 g 

r- 'S =5 p- ' r p- 'E 
S CD -- ~ ^ i 
-2 ^ ^ -^ '^ 
S<; 2 S<j fl<| 
m ^ - --. ' CD 

o < a o 


Is ~ c3 cS 
-/. <1 /- -f. 

pri ^t p- 

a PS J" p 
t CB P t3 











o o o o 

00 jT + 












OS + 




d a? 

, + + + 

















is i 

QJ ^ <3) 

p CC Q 

^.B g g g 
3 a; ^c 

V; J!' QJ Q^ ^J 

^ q ft q 

S r-* 

r^ -5 l-^ ~>- 

^ u- '--I 





r- 1 




i i 




rH i 1 rH 

H 01 01 01 

r rH 01 01 













C^- -j- O 

( 'ucullaris 






+ + , 

+ ** 

P5 05 P5 PH' 










o o o 

o .. + + 

00 .T 









:p -' -~ 

? '? i 

> %, > ^ 









x 7 T 

^ < < 

>H >H >J 


2 QJ 









+ + + 

+ ++ + 












ry. Ffii PTLI 

fa " 1Z " 


-<li 02 







o + 

+ + + 

o o 










Todidae . 




Coliidae . 


... . OS 

83 ^a 

^ m ft 
.PH K P- 
ft T3 P 
p^ .p-l oj 

ft .2 ^ 

P , 83 83 
1 ill 

. ,0 S 
W i 1 fe 

os n oj 

r ^ 03 r^ ^^ 


88 ffl S 1 

P^ -3 a - 
s -g -s '& 

>> p- oS 


In the above table I have used twelve characters. The 
Picidae, Bhamphastidae, and Capitonidae agree in ten of these, 
and are undoubtedly nearly allied birds. The Bucconidae are 
unfortunately not well known, but of the nine characters 
which are set forth in the table they agree in eight with one 
or other of the three families just mentioned. 

The Coraciidse and Meropida? agree pretty well in all the 
characters except the exact arrangement of the deep flexor 
tendons, the carotids being variable. There will be but little 
violence done if these groups are associated. The Cypselidae 
and Trochilidae clearly come near together, agreeing as they 
do in nine of the selected characters. 

The Caprinmlgidse on the one hand, and the owls on the 
other, each form a distinct group with no such near affinities 
to any of the others as those which we have been considering. 

The trogons are the only other group with basipterygoid 
processes ; they do not, however, come very near to the 
Caprimulgidae or to the owls ; out of the selected twelve 
characters they have at least four in which they totally 
differ from the first, and five in which they differ from the 
second. This group may be left as equivalent to the other 
compound groups already considered. 

The Todidaa are placed by GADOW close to the motmots ; 
by FORBES, on the other hand, they are widely separated. 
They agree with them in ten out of the twelve, showing 
thus, it appears to me, a considerable nearness. They agree 
equally closely with the Meropida? on the one hand and the 
Galbulida? on the other. These four groups appear to me to 
be worthy of association into one larger group. 

We have left the Bucerotida?, Upupidae, Alcedinidse, and 
Coliidae. The kingfishers undoubtedly come near to all of 
these, in only at most four of the characters differing from 
any one of them ; but they are as near to the motmots and 
Rhamphastidae. They should form a group apart. 

The same may be said of the colies ; they are very near 
to the hornbills and kingfishers, but equally near to the mot- 
mots and Rhamphastidse ; we may therefore place them in a 
group apart. On the other hand the Bucerotida3 come nearer 


to the Upupidas than to any other groups except the colies 
and Alcedinidae ; we may therefore unite them. 

Finally as to the passerines. It is not possible to place 
them very definitely nearer to one than to another of the 
groups enumerated. They differ at the lowest in four 
characters from any. They are, perhaps, furthest away from 
the Bucerotidse ; the two groups, in fact, are typical members 
each of them of the piciform and passeriform birds of 
GAEROD. Their leanings are perhaps to the Cypselidse and 
to the Pici. 

These conclusions may be tabulated as follows :- 

Group A. Aftershaft present ; fifth remex always present. 
Muscle formula, AXY. Expansor secundariorum present. 
Desmognathous. Vomer present. 

This group contains the families Coraciidae, Meropidae, 
Todidse, Galbulida?, and Momotidae. 

By both G-ADOW and FURBRINGER the Galbulidse are 
placed nearer to the Bucconida?, and by inference to the Pici 
(Picida 3 , Bhamphastidw, and Capitonidse). They differ, how- 
ever, from these by the presence of the expansor secundario- 
rum, which is so rarely present among anomalogonatous 
birds that when present it seems to be of special importance. 
The most salient point of agreement between these latter 
birds is the form of the deep flexor tendons. But the 
Bucconida? are so little known that they may be found to 
differ more than is at present suspected from the Pici. In 
this case it may be desirable to separate both Galbulida^ and 
Bucconidas from the group with which I now associate them 
and place them nearer together. Of the group as at present 
constituted the Todidae have perhaps the most claims to be 
regarded as the most primitive forms. They have a feathered 
oil gland and long caeca, which two characters do not coincide 
in any other of the families now under consideration. 

The next group, that of the Pici, may be thus defined :- 

Group B. Fifth remex always present ; deep flexor 
tendon of type VI. Expansor secundariorum absent. 
Cucullaris propatagialis present. No caeca. Vomer 

AX( )M A LOGON AT^E 1 7 1 

There are other peculiarities that unite these birds, which 
will be found mentioned on p. l^->. 

The third group, containing the hornbills and hoopoos, 
may be thus defined : 

Group C. Oil gland feathered ; fifth remex present. 
Muscle formula, AXY. No expansor secundariorum. Caeca 
absent. Skull desmognathous. 

The colies are the only birds among the Anomalogonatae 
besides the Caprimulgidae which possess the biceps slip ; there 
is a rudiment of this structure in Bncorvux. 

The remaining families are treated of separately and need 
not be defined here. 

This arrangement, nearly coincident with that of GADOW, 
is widely different from that of GARROD and FORBES. The 
former divided the birds (excl. Striges) into the three groups 
of Piciformes, Passeriformes, and Cypseliformes. They were 
thus defined :- 

Piciformes. Oil gland tufted ; ca3ca absent ; external 
branch of pectoral tract given off at commencement of breast. 
Muscle formula, (A)XY. Picida3, Capitonida?, Upupidse, 
Bucerotidae, Coliida?, Alcedinidae, Momotidse. 

Passeriformes. Oil gland nude ; caeca present ; pectoral 
tract simple or with external branch given off beyond middle 
of breast. Muscle formula, AX(Y). Passeres, Bucconidae (?), 
Galbulidae, Coraciidae, Meropidse, Trogonidae. 

Cypseliformes. Oil gland nude ; caeca absent. Muscle 
formula, A. 

To these FORBES added a fourth group Todiformes, on 
account of its having at once caeca and a tufted oil gland. 
This latter group was regarded by him as most nearly repre- 
senting the ancestral anomalogonatous bird. 



Definition. Oil gland nude. Skull segithognathous. Atlas perforated 
by odontoid process. One carotid, left. Caeca present, small.' 
Muscle formula, AXY.- No biceps slip or expansor secunda- 

This is an enormous group of birds, numbering over 
6,000 species, which are spread over the entire globe. As a 
rule they are of small or moderate size ; but some large 
species, such as the raven and other Corvidse, are included 
in the assemblage. In spite of the numerous species and of 
a certain amount of differentiation in external form, the group 
is structurally a uniform one, the difference being in cha- 
racters which have not as a rule been regarded as of primary 

In all Passeres the foot has this structure : the first toe is 
directed backwards, and none of the other toes are ever 
changed in position. Cliolornis is exceptional in that the 
fourth toe is abortive. 

GAEEOD noted, some years since, another peculiar character 
of the group, which may possibly be universal, with the ex- 
ception of Menura and Atrichia. This concerns the arrange- 
ment of the tendon of the patagialis brevis ; and the passerine 
disposition may be understood from a comparison of the two 
descriptions. In the passerine the tendon of the muscle 
does not end upon the tendon of the extensor, as it does 
in the picarian bird, but, though attached to it firmly, 
retains its independence and runs back to be attached near 
it to the extensor condyle of the radius. This difference, 
though small, appears to be constant to the Passeres. Another 
character dealt with on p. 41 of the present work may be 
also an exclusively passerine character. In birds belonging 
to the present family the oblique septa, instead of having a 
separate attachment to the sternum, are either not attached 

1 In a specimen of Gracula intermedia the casca were as long as half an 
inch, an exceptional length. 
' Very rarely AX . 


at all, lying loosely over the liver, or have but one attach- 
ment, which they share with the falciform ligament. 

It is very general but there are a few exceptions, which 
will be dealt with later for the sternum to have a forked 
manubrium in front and a single pair of notches behind. 

The number of rect rices present among passerine birds 
varies. They are, indeed, completely absent in wrens of the 
genus P)ioepijga. Twelve is the usual number, but ten only 
occur in Xeiiicus, Plirenotrix, and Edoliiis, while Menu r a 
snperbah&s sixteen. The aftershaft is ' very weak and downy ' 
when present, and is sometimes (e.g. Paradisea rubra) absent 
altogether. As to the pteryloxis, we may take Ampelis 
cedronon, recently described by SHUFELDT/ as an example 
of passerine pterylosis, mentioning afterwards such varia- 
tions from this type as are met with. In the bird in question 
the dorsal tract is exceedingly narrow from the origin in the 
fairly continuous feathering of the head down to a point in 
the pelvic region, where it is greatly dilated to form a 
diamond-shaped area ; this again contracts to the original 
dimensions, and the tract concludes a little way in front of 
the base of the oil gland. From the lateral angles of the 
diamond-shaped area a tract runs to the feathering of the 
legs. On either side of the oil gland arises a short tract, 
which does not leave the trunk, but appears to be the hinder 
part of the femoral tract of some other birds. It is perhaps 
noteworthy that it is very similar to the corresponding one 
of the Bucconida3 and Capitonidse, also, however, of the 

The ventral tract divides early on the neck, and on the 
breast is increased in breadth, the outer rows of feathers 
being much stronger than the inner set ; the latter (not the 
former) are continued down to the cloacal orifice, which 
they completely surround by a narrowish band of feathers. 
The humeral tracts, which are strong, are connected with 
the head feathering by a special neck band as wide as the 
dorsal tract. NITZSCH does not figure this connection or 
that of the diamond-shaped dorsal area of feathers with the 

1 In a paper dealing with Macrochires in J. Linn. Soc. vol. xx. 


femoral feathering ; but apart from this the same type of 
pterylosis occurs in many types, such as Motacilla, Certliia, 
Oriolus, &c. The principal variation is offered by those 
passerines in which the spinal widened area is not solid, but 
encloses a space. 

We find this in Coracina chplialoptcra and Selencides. 
In others (e.g. EurylcBums) there is the same ephippial space 
within the dorsal tract, but the posterior sides of the diamond- 
shaped space are formed by a single row of feathers, which 
contrast with the mass of feathers which form the antero- 
lateral boundaries of the space. This arrangement culmi- 
nates in, for example, Hiriuiclo and Diphyllodes, where the 
dorsal tract forks, and there is no connection between the 
ends of the fork and the single posterior part of the spinal 

tract. 1 

The skull of Passeres has been mainly investigated by 
PARKER and by SnuFELDT.' 2 Corvus may be taken as a 
type, and the divergences therefrom noted later. As are all 
passerines, it is a?githognathous ; the maxillo-palatines extend 
obliquely outwards and backwards ; they approach each 
other in the middle line, and expand over the vomer. The 
vomer is broad and bifurcate both anteriorly and posteriorly ; 
from the anterior horns a small separate piece of bone goes 

1 For passerine pterylosis, see, in addition to NITZSCH, GIEBEL, ' On Ptery- 
losis of Paradisea,' Zcitschr. f. d. ges. Naturw. xlix. p. 143, and for Philepitta 
ibid. p. 490 ; SHUFELDT has described Chamcea, Journ. Morph. iii. 1889, p. 
475; HELLMANN, ' Beitrag zur Ptilographie u. Anatomie der Hirnndo mstica,' 
,/. /. O. iv. 1856, p. 360 ; GADOW, ' Remarks on the Structure of Certain Hawaian 
Birds,' in WILSON and EVANS'S Avcs Hawaicnses, ii. Sept. 1891. 

- For osteology of Passeres see PARKED, ' On the Structure and Development 
of the Crow's Skull,' Month. Micr. Journ. 1872, p. 217 ; ' On the Development 
of the Skull in the Tit and Sparrow Hawk,' ibid. 1873, pp. 6, 45 ; ' On the 
Development of the Skull in the Genus Turdus,' ibid. 1873, p. 102 ; ' On 
.-Egithognathous Birds,' Trans. Zool. Soc. ix. p. 289, x. p. 251 ; MUKIE, ' On 
the Skeleton and Lineage of Fregilupus,' 1 P. Z. S. 1874, p. 474. LUCAS, 
' Notes on the Osteology of the Thrushes,' &c., P. U. S. Nat. Mus. xi. 1888, p. 


SHUFELDT, ' Osteology of Eremopliila,' 1 Bull. U. S. Gcol. Surv. vi. p. 119 ; 
1 Osteology of Lanius,' ibid. p. 351 ; ' On the Skeleton in the Genus Stur 
in'lltt,' &c., J. Anat. Phys. xxii. p. 309; 'Osteology of Habia Auk,' v. p. 438 ; 
' Osteological Notes on Puffins and Ravens,' ibid. p. 328 ; GIEBEL, ' Zur Osteo- 
logie d. Gattung OcijiderusJ Zcitschr. f. d. ges. Naturw. xxi. p. 140. 


on each side to the maxillo-palatines. These are the septo- 
maxillaries of PARKER, and appear 10 remain perfectly dis- 
tinct in Corvus comix, but not in Corvusfrugilegus. There 
are no basipterygoid processes ; the pterygoids have a long 
foot-like attachment (as long as the free part of the bone) 
not only to the palatines, but to the interorbital septum also. 
The nares are holorhinal and pervious. In Corvus comix 
the lacrymals reach the jugal bar ; there is practically no 
orbital portion, the descending limb being closely attached 
to, but not fused with, the broad and thick ectethmoid. The 
mandibular rami have a large oval perforation near to the 
articular surface. 

Among genera nearly related to the Corvidae are various 
slight modifications of skull structure. In Maiutcodia, for 
instance, the rostrum is broadly ossified and fused with the 
co-ossified palatal plates of the maxillae. The nasal septum 
is complete, and the conjoined ectethnioids and lacrymals 
are enormously swollen. In Pt Honor hynchus violaceus the 
ectethmoids and lacrymals are separate, though in contact ; 
contrary to what is found in Core us, it is the latter and not 
the former which border the orbit above. The palatal con- 
ditions of Mcuincodia are repeated and emphasised in Gym- 
norhina and Strepera. There is a firm union across the 
middle line in front of the vomer, with which, indeed, the 
anterior horns of the vomer are ossified in Strepera. In both 
birds, moreover, the pterygoids are fused with the pala- 
tines, and the nostrils are partly obliterated by bony growth. 

The ' desmognathism ' thus produced in the crows of 
' Notogsea ' is not limited to that family. In Pheiictic/is 
and in Cracticus cassicus there is the same state of affairs. 
Other features in which the passerine skull shows variations 
are the maxillo-palatines, vomer, and pterygoids ; in Gracula 
javanensis, for example, the pterygoid has a very limited area 
of articulation with the palatine ; there is no expanded foot, 
as in crows, &c. ; the maxillo-palatines are very long and 
slender, actually reaching the inner plate of the palatines. 
The vomer is narrow in the body, though the two anterior 
' cornua ' are thick. Trochalopteron is almost desmognathous 


in the sense of HUXLEY ; the two maxillo-palatines, dilated 
at their ends, come absolutely into contact under the vomer, 
the middle part of which they completely cover. In Guis- 
calus versicolor the same arrangement occurs, the maxillo- 
palatines being extraordinarily slender before they dilate to 
overlap each other and to cover the vomer. 

As to the rest of the skeleton, Corvultur albicollis may 
serve as a type. There are fourteen cervical vertebrce ; un- 
impaired hypapophyses extend from the tenth to the first 
dorsal. Five ribs reach the sternum, to the keel of which 
the furcula is joined. 

The most salient variations from this plan are shown by 
many genera in which the furcula does not join the keel of 
the sternum, by Corcorax, in which six ribs articulate with 
the sternum, and above all by GymnorJiina, where there is 
a catapophysial canal, beginning with the seventh and 
ending with the tenth vertebra. In this and other forms 
the hypapophyses of the last cervical and the first dorsal 
vertebrae are reinforced by strong lateral catapophyses. 

In the table on the opposite page are some intestinal 
measurements in inches. The most noteworthy fact in the 
anatomy of the alimentary tract is the absence of the gizzard 
in certain tanagers. 1 

In almost all Passeres - the flexor hallucis is absolutely 
independent of the flexor communis, there being no vinculum 
at all. 

These characters are adding to them those used in the 
diagnosis of the family the only ones that are universal, or 
nearly so, among the Passeres. There are, however, a number 
of anatomical features in which the passerines show differ- 
ences among themselves. The most abnormal Passeres on 

1 Ci'. FORBES, ' On the Structure of the Stomach in certain Genera of 
Tanagers,' P. Z. S. 1880, p. 143, who quotes LUNI>'S earlier (1829) paper on the 
same matter. 

2 For myology of passerines see, in addition to GABISOD, KLEMM, ' Zur 
Muskulatur der Raben,' Zcitschr. f. d. gen. Xnlui'ir. xxiii. p. 107 ; SHUFELDT, 
The Anatomy of the Raven, London, 1890 ; C. L. NITZSCH, ' Ueber die Fainilie 
d. Passerinen,' Zeitschr. f. d. ges. Nat. xix. p. 389 ; C. B. ULRICH, ' Zur 
Characteristik d. Muskulatur d. Passerinen,' ibid. xlv. p. 28. 



S. I. 



Myiophoneus Horsfieldi ... 29 



Geocichla citrina .... 9 



Nesocichla eremita .... 15 



Eimator malacoptilus .... 5-1 



Cracticus cassicus .... 10'2 1 



Pastor roseus ..... 13-75 1 


Seleucides nigra 14-85 1 



Manucodia atra 14-15 1 


Uranornis rubra 16-2 1 

3 -3 

Entomyza cyanotis ...... 12 1 


Garrulax albogularis .... 13 1 

2 -1 

Cyanocorax cyanopogon ... 13 

5 -25 

Cissopis leveriana .... 10 


Struthidea cinerea .... 



Hirundo rustica ..... 6'25 

75 -6 

Anthornis melanura .... 6 - 25 

75 -12 

Strepera graculina .... 20 


Barita destructor 13 


Tanagra sayaca ..... 8 

5 -12 

,, festiva ..... 6-5 

Ptilonorhynchus holosericeus . . 10-25 

75 -25 

. . 10-5 1 

25 -5 (F.) 

Gracula javanensis .... 23 1 

5 -25 

Gymnorhina leuconota ... 22 1 

5 -4 

Corvus corone 29'75 2 

5 -5 

Ampelis garrulus ..... 7 

5 -2 

the whole are the broad-bills Eurylsemidse and the Australian 
Menura and Atrichia, which form a sub-family, Menuridse. 
There are differences of opinion as to which of these is most 
independent of the normal Passeres. FURBR.INGER separates 
the Menuridre, GAEROD and FORBES l the Eurylaemidae. We 
should explain that in first of all discussing this particular 
point we are not proceeding in historical sequence. It was 
the syrinx that first of all attracted the attention of JOHANNES 
MULLER, whose divisions of the Passeres were the earliest to 
be based upon anatomical structure ; and in the sequel we 
shall show that his divisions are in the main correct, even 
allowing for our greatly extended knowledge. It is, however, 
in our opinion, beyond cavil that the major subdivisions of 

1 GABEOD'S contributions to our knowledge of passerine birds are as 
follows : ' On some Anatomical Peculiarities which bear upon the Major 
Divisions of the Passerine Birds,' i. P. Z. S. 1876, p. 506; 'Notes on the 
Anatomy of Passerine Birds,' ii. P. Z. S. 1877, p. 447; iii. ibid. p. 523 ; iv. 
ibid. 1878, p. 143. The following papers are due to FORBES : ' Contributions 
to the Anatomy of Passerine Birds,' i.-vi. P. Z. S. 1880-2. 



the group first concern one of the two views that we have 
referred to above. The reasons which lead us to agree with 
GARBOD and FORBES' s separation of a group Desmodactyli, 
as opposed to the remaining Passeres, which are to be so- 
called Eleutherodactyli, are as follows : The Menuridae 
(Pseudoscines of SCLATER and FURBRINGER) are clearly in 
some respects degenerate forms. The clavicle has become 
rudimentary, and the muscles of the syrinx, while approach- 
ing the typical oscinine form, where these muscles are 
numerous and strong, have become to some degree weakened 
by loss. 

On the other hand the Eurylsemidae, while they have 
retained the typical mesomyodian syrinx typical, because it 

FIG. 85. SYKINX OF Eurylcemus. 

FIG. 8fi. SYRINX OF Cymbi- 
rhynchus. SIDE VIEW. (AFTER 

is distinctive of the vast majority of birds have retained the 
plantar vinculum, 1 which in other passerines has been lost ; 
they have also a simple maiiubrium sterni, this appendage 
being forked in other passerines. In the feet too the third 
and fourth toes are largely bound together, giving to the 
group the name of desmodactyli. 

The family Eurylaernidse 2 contains the genera Eurylamus, 
Calyptomena, Serilopha, Psarisomus, Cory don, and Cymbi- 
rhynchus, all Old- World. They have no aftershaft, and the 
oil gland is, of course, nude. There are twelve rectrices in 

1 Occasionally absent in Calyptomena viridis. 

- FORBES, ' On the Syrinx and other Points in the Anatomy of the Eurylse- 
midse,' P. Z. S. 1880, p. 380. 



Cymbirhynchus. There is a wide ephippial space, elongated 
and oval in form ; the narrower parts of the tract behind 
are two feathers wide. 

The tongue (of Cymbirhynchus) is bifid at tip and 
elongated and cordate. 

The following are intestinal measurements of two species : 

Small int. 
Large int. 


. 7-75 

. 1-25 




Left liver lobe is the smallest. 

The following scheme gives the classification of the 
Passeres according to GABEOD and FORBES : 

A. Mesomyodi 
a. Heteromeri 

6. Homctomeri 


B. Acromyodi 

New World 






OM World 








( = Oscines) 

Of the remaining Passeres the Mesomyodi (also sometimes 
called Oligomyodi) are divided into two subdivisions, according 
as to whether the chief artery of the leg is the femoral or 
sciatic. In the Heteromeri (with the exception of Rupicola} 
it is the femoral, in the others (Homceomeri) the sciatic. All the 
Mesomyodi have but one pair of muscles upon the syrinx or 
none at all. But the name of Mesomyodi is derived from the 
fact that these intrinsic muscles are attached to the middle 
of the bronchial semi-ring which bears them. The Haploo- 
phonse are those mesomyodians in which the syrinx is quite 


of the normal fashion ; the Trachephonae are those in which 
the last rings of the trachea are much modified, and the 
syrinx may be termed tracheal. 

The Mesomyodi with a tracheobronchial syrinx comprise 
representatives from both the Old and the New Worlds. In 
them the syrinx presents a varied form, coupled with the fun- 
damental resemblance indicated. JOHANNES MULLEE has 
figured and described a number of genera. GABKOD has 
figured and described others. In Lipaugus cineraceus (of 
the family Cotingida?) the intrinsic muscle is of great width, 
which seems to foreshadow its division in the Oscines into a 
complex of muscles ; it is attached to the third bronchial 

The first and second bronchial semi-rings resemble the 
tracheal in their flatness, depth, and close approximation. 
Those which follow are slightly ossified throughout. In 
Heteropelma and CJiiromadiceris, which are Piprida3, the 
syrinx is very similar. In Pipra leucociUa the intrinsic 
muscle has a tendency to split into two, a further approxima- 
tion to the Oscines. In Hadrostomus aglaice, a cotingid, the 
wide and thin intrinsic muscle is attached to the first bron- 
chial semi-ring. This semi-ring is close to the last tracheal 
ring, and is like it in structure, being deep. The next 
bronchial semi-ring is separated by a considerable interval, 
and the third by a wider interval still, from the ring in 

In the Madagascar Philepitta, which FOEBES ' was the 
first to refer definitely to the present group of passerines, the 
structure in some respects recalls that of the Euryla?mida3. 
The manubrium sterni is but slightly bifid ; but it has in 
the normal passerines no vinculuin. The syrinx, on the 
other hand, differs from the Old-World Mesomyodi by the 
details of its structure (see figs. 87, 88). The different arching 
of the bronchial semi-rings leaves great membranous spaces 
in the wall of the syrinx. The first two semi-rings are very 
concave upwards ; the two following are not so markedly 
concave ; the next is concave in the reverse direction. The 

1 ' On some Points in the Structure of Philepitta,' &c., P. Z S. 1880, p. 387, 



intrinsic muscles are wide and thin, nearly in contact with 
each other before, and behind they are attached to the first 
semi-rings everywhere bat at their tips. Pitta has a much 
simpler syrinx, approximating so far to that of the Eurylae- 
midse ; the muscles are thin and accurately median in inser- 
tion. Pitta is unique among passerine birds by reason of 
the deep temporal fossae of the skull, which nearly meet 

FIG. 87. SYRINX OF Philepitta. 

FIG. 88. SYRINX OF Pliilepittct. 

behind, in a way that is seen in some other birds not 

In the New Zealand Xenicus and Acantliisitta ' there are 
A B 




only ten rectrices, twelve being the number characteristic of 
the majority of the Passeres. The syrinx of Xenicus as seen 
in the annexed figure is quite typically mesomyodian. The 
last few tracheal rings are consolidated into a large box, to 
the top of which the intrinsic muscles (small and median in 
insertion) are attached. 

1 FORBES, ' On Xenlcm and Acanlltixitt,' <fec., /'. '/.. S. 1882, p. 569. 


We now come to the Tracheophonae, which are exclu- 
sively American in range. Their distinguishing mark is, of 
course, the tracheal syrinx, whose general structure has 
been already explained (see p. 67). Besides these general 
agreements the Tracheophonae show variations in structure. 
Hetcrocnemis, for instance, is unique among them and 
the Mesomyodi in general by the existence of a bilaminate 
tarsus, as in the Oscines. Conopophaga and a few others ' 
have a four-notched sternum, while Furnarius, Synal- 
laxis, and a few others have, as has no other passerine, 
a schizorhmal skull. Again, the maxillo-palatine of the 
Dendrocolaptida3, Furnariidas, and Pteroptochidge are like 
those of oscinine Passeres in being slender and curved back- 
wards, instead of being comparatively wide and blunt, as in 
other Mesomyodi. Pkytotoma is unique among passerine 
birds for the nasal gland groove on the frontal bones, as in 'so 
many water birds. 2 

The remaining group of Passeres, the Acromyodi, are 
sometimes called Oscines and sometimes Polymyodi, the latter 
term having reference to the numerous intrinsic muscles of 
the syrinx. It was discovered by KEYSEELING and BLASIUS 
that, with the exception of the Alaudidse, the Oscines have a 
bilaminate tarsus, the hinder surface being covered by two 
closely apposed scutes. It sometimes happens (in the forms 
which are on that account spoken of as ' booted ') that the 
anterior face of the tarsus is covered by a single scute. The 
syrinx of these birds is complex in the multitude of its 
muscles, of which there are four or five pairs. The only 
exception to the muscle formula AXY exists in this group ; 
in Dicrnrus there is the reduced formula of AX . Ocypterus, 
the only passerine with powder down, is referable here. 
Referable to this group, but separated from the more normal 
members as ' Abnormales ' by GARROD, are the two genera 
Atricliia and Menura, which are also regarded as the types 
of separate families. These two anomalous birds are by 

1 FORBES, ' On some Points in the Anatomy of the Genus Conopophaga,'' &c., 
P. Z. S. 1881, p. 485. 

- PARKEK, ^githognatlious Birds, ii. p. 258. 


some systematists placed in a group Pseudoscines, equivalent 
to the remaining Passeres. 

The anatomy of the two genera has been mainly investi- 
gated by GARROD, who studied principally the syrinx. They 
are purely Australian in range. According to FURBRINGER 
this group of passerines is in many respects intermediate 
between the other passerines and the Pici. With the latter 
group M emir a probably and Atricliia certainly agree in the 
following myological points : the origin of the rhomboideus 
profundus from the pelvis ; in the tendon of insertion of the 
supra-coracoideus upon the shoulder joint ; in the origin of 
the latissiinus dorso-cutaneus from the ilium and its covering 
by the leg musculature. GARROD also called attention to 
the fact that the patagialis brevis was picine and not passerine. 

The syrinx of Menura 1 has three pairs of intrinsic mus- 
cles, which are inserted respectively into the last tracheal 
ring and on to the second and third bronchial semi-rings. 
There are three modified bronchial semi-rings. So too in 
Atricliia, where, however, there are but two pairs of syringeal 
muscles, as shown in the figure. 

The clavicles are rudimentary in Atrichia,', there is no 
hypocleidium in Menura, another picine character. Menura 
has, furthermore, a chain of three supra-orbital bones. 


Definition. Feet zygodactyle ; aftershaft small or rudimentary ; 
oil gland tufted. Muscle formula of leg, AXY (AX) ; Gall 
bladder elongated. Skull without basipterygoid processes. 

The woodpeckers, which form the first family of th& 
assemblage, Picidae, 2 are a well-marked group of birds, contain- 
ing about three hundred and fifty species, as allowed by the 
late Mr. HARGITT. S They inhabit most parts of the world, 
excluding only Madagascar, Australia, and Polynesia. 

1 For passerine syrinx see JOH. MuLLERin Abh. Berlin. Akad., 1845 ; HERRE, 
Dissertatio de Avium Passcrinan<ni Larymjc Bronchial*, Gryphise, 1859 ; and 
GARROD and FORBES in papers already quoted. 

- W. MARSHALL, Die, Speclite. Leipzig, iss'.i. 

3 Brit. Mus. Catalogue, vol. xviii. 


Twelve rectrices is the rule for this family, but the outer 
ones are sometimes feeble, and I found only ten in Tig a 
Shorei. The aftershaft is present. The pterylosis varies 
somewhat. But NITZSCH has pointed out that this ' peculiarity 
. . . which occurs almost universally among them, is the 
presence of a small inner humeral tract running along upon 
the most elevated points of the shoulder parallel to the 
very broad main tract.' That this second humeral tract also 
occurs among parrots may be a matter worthy of considera- 
tion. It appears, at any rate, to distinguish the woodpeckers 
from other picarian birds. 

In Picus viridis the narrow dorsal tract passes down the 
middle of the neck and ends abruptly at the end of the 
scapulae or a little before. Behind this are two oval wide 
patches, which correspond to the foot of the Y in other 
birds, which have a dorsal median apterion. There is a 
break between these patches and the posterior end of the 
dorsal tract, which runs to the base of the oil gland undivided. 
On either side of this and of the oil gland is a fainter and 
narrower tract. In Tig a Shorei there is no break between 
the several regions of the dorsal tract ; the anterior part 
narrows and forks into two branches, consisting of but one 
row of feathers, which immediately after dilate into the wide 
interscapular tracts ; from the lower angle of each of these 
a single row of feathers joins the median posterior part of 
tract. The diamond-shaped spinal aptsrion is thus com- 
pletely enclosed within the dorsal tract. 

Sphyrapicus nuchalis ! has a solid spinal tract dilating 
between the shoulders into a rhombpidal but still solid area, 
as SHUFELDT says, like a passerine. 

In Centurus striatus the arrangement is more like that 
of Picus viridis, but slight scattered feathering unites the 
middle and posterior portions of the dorsal tract. 

The ventral tracts of Picus divide early in the neck and 
at the commencement of the pectoral region ; each gives off a 
stronger outer branch. In Tig a Shorei the separation between 
these branches is not nearly so marked as is shown in 

1 'Observations on the Pterylosis of Certain PicidaV Auk, ISHS, p. 212. 



NITZSCH'S figure of Picus viridi*. This woodpecker is so 
far much more like a parrot. 

In Jynx the interior part of the dorsal tract forms a 
continuous Y, of which the fork is hardly wider than the 
handle. This is completely discontinuous with the median 
posterior portion of the tract. 

The tensores patagii 1 are very simple. The tendon is 
single. There is a conspicuous cucullaris patagialis, but 
no biceps slip. Each tendon, both longus and brevis, is 
reinforced by a tendon from the pectoralis. In Centurus 
striatus the slip to the longus is muscular in origin ; it 
seems to be more usually tendinous. 

The deltoid is an extensive muscle. 

The latissimns dor si posterior appears to be totally absent, 
as in Indicator. 

As to the leg muscles, the accessory femoro-caudal is 
always absent. The femoro-caudal and the semitendinosus 
are always present ; the accessory to the latter may or may 
not be present, its occurrence in different genera being 
shown in the following list : 

"\Vith an Accessory Semitendiuosus 

Without an Accessory Seinitemliuosus 

Pi cits major 
Picus minor 
Picoides tridactylus 
Spliyrapicus variiis 

Gee in us viridis 
Gecinus vittatus 
Leuconerpes candidus 
Melanerpes formicivorus 
Chloronerpes yucatanensis 
MnUeripicus fulvus 
Hypoxanthiis Rivolii 
Jynx torquilla 
Dnjocopus martin* 
Picolaptes affinis 
Tiga Shorei 
Tiga javensis 
Centurus striatus 
Melanerpes erythrocepha- 

Colaptes mexicanoides 

1 NiTzscH-GrcBEL, ' Zur Anatomie der Spechte,' Zcitschr. /. d. gcs. Naturir. 
xxvii. (1806), p. 477. 


The deep flexor tendons are like those of toucans and 
barbets (cf. p. 101). 

The tongue l is elongated, and so are the ceratohyals in 
relation to it, overlapping and grooving the skull. As a rule 
among the woodpeckers the right lobe of the liver is larger 
than the left ; they are equal in Xypoxantlms Eivolii. The 
gall bladder appears to be absent in Leuconerpes candidus 
and \\\Xypoxanthus ; it is long and intestiniform (like that of 
a toucan) in Gecinus viridix, Dryocopus martins, undJynx. 

The intestines are without caeca ; ~ the following are some 
measurements : 

Chloronerpes yucatanensis . . &k inches 

Picas minor . . . . . . 1'2 ,, 

Xypoxantlms Bivolii . . . . 1'2 ,, 

Dryocopus martins . . . . '20 ,, 

The syrinx has no remarkable characters. It is quite 
typically tracheo-bronchial with a pair of extrinsic and 
a pair of intrinsic muscles. 

As to the skull, the woodpeckers have an unusual palatal 
structure, which led PARKER 3 to invent the term Sauro- 
gnathous in order to express this peculiarity. The palate oi 
the woodpecker has, however, been variously interpreted. 
Professor HUXLEY, in his paper on the bird's skull, directed 
attention to an apparent vacuolation of the palatines, which 
is illustrated in the accompanying figure (fig. 90) ; a slender 
bar of bone passes backwards and comes into near relations 
with an equally slender projection forwards of the palatine 
bone, thus enclosing a space which in the fresh skull is filled 
with membrane. The anterior process in the dried skull 
is sometimes continuous (fig. 90), but sometimes not con- 
tinuous with the ascending lamina of the palatine. Where 
it is not continuous Professor HUXLEY found that its appa- 

1 J. LINDAHL, ' Some New Points in the Construction of the Tongue of Wood- 
peckers,' Am. Nat. 1879, p. 43. 

2 Exceptionally present in Gecinus viridis. 

3 ' On the Morphology of the Skull in the Woodpeckers,' &c., Tr. Linn. Soc. 
(2), i. 1875, p. 1. See also KESSLEK, ' Zur Naturgesohichte der Spechte,' Bull. 
Soc. Nat. Mosc. 1844, p. 28=5. 



rent continuation was a separate bone, which he regarded 
as the vomer the vomers thus being paired. 

A few years later a different complexion was given to the 
subject by a paper written by Professor GARROD. ' He con- 
firmed HUXLEY'S description of the supposed vomers, but 
regarded them merely as the perfectly ossified edge of the 
imperfectly ossified palatines. This opinion was chiefly 
based upon the discovery of a small median bone lying 
between the posterior ends of the palatines (x. fig. 90), not 
observed by HUXLEY. This bone, identified by GARROD 
with the vomer of other birds, occupies, 
as he admits, a somewhat posterior posi- 
tion, which is parallel, however, as he also 
points out, at any rate in Megalcema* 

PARKER'S paper upon these birds is a 
long and elaborate one, but contains no 
reference to that of GARROD. The bone 
discovered by GARROD is figured and de- 
scribed as the medio-palatine, and is in- 
variably figured as distinct from the 
palatines, between which it lies. PARKER 
adopts HUXLEY'S identification of the 
vomers, but finds that they are often 
divisible each into several splints of bone ; 

,, ,, , FIG. 90. SKULL OF 

the connection with the palatines m iront WOODPECKER (Geti- 
and a median series of bonelets (collectively nus viridis). VEN- 
termed by HUXLEY the ossified internasal 
septum) are spoken of as septo-maxil- 
laries. It may be sometimes noticed I 
have observed it in Leuconerpes candidus that the anterior 
ends of the pterygoids, which in the Pici run for a consider- 
able way over the outside of the palatines, come into actual 
contact with the commencement of the vomers. 3 

Another peculiarity of the woodpecker's skull is the fre- 

1 ' Note on some of the Cranial Peculiarities of the Woodpeckers,' Ibis, 
1872, p. 357. 

'-' See under ' Capitonidas,' p. 196. 

3 SHUFELDT also (' On the Question of Saurognathism of the Pici,' &c., 
P. Z. S. 1891, p. 122) argues against HUXLEY'S view. 

nns viridis). 



I'l, palatines ; J't, ptery- 
goids ; .c, vomer. 


quently Y-shaped form of the pterygoids. An outer 
forwardly directed process is given off near to their articula- 
tion with the quadrate. The extraordinary hyoid connected 
with the long and exsertile tongue is, of course, a marked 
feature of the group. The long distal piece of the bran- 
chials curves over and furrows the cranium of those species 
in which it exists. 

In the wryneck (Jynx) the vomers are small, but 
thoroughly ' picine.' The pterygoids have, however, no 
forwardly directed process, and do not extend quite so far 
over the palatines. 

There are fourteen cervical vertebra. Haemapophyses 
do not extend behind the first dorsal. There is a complete 
canal for the carotids formed in the cervical region by the 
haemapophyses. Four vertebme enter into the formation of 
this in Thriponax Feddeni. 1 The sternum has two notches on 
either side, and has the spina exterior only, which is bifid. 
Four ribs reach it. The clavicles form a U-shaped furcula, 
but are expanded above, as in the toucans and barbets. 

The Bucconidae is a South American family of zygodac- 
tyle birds whose anatomy - is at present but little known, and 
whose affinities are therefore doubtful. It is only provi- 
sionally that I place them in the present position. 

Bucco maculatus has a nude oil gland and twelve rectrices. 
The inferior feather tract starts from the symphysis, leaving 
a bare space on either side ; it divides at the angle of the 
jaw, and thence the two halves remain separate. Each of 
them gives off a stout outer pectoral branch. The dorsal 
tract is in two quite separate parts, separated by a good 
space ; the posterior part is forked for half its length, 
tapering anteriorly to a point. 

In Monasa panamensis each part of the posterior fork 

This character is nearly unique among anomalogonatous birds. I 
describe it, however, above (p. 176) in Gymnorhina. The nearest possible ally 
in which it occurs is a parrot. 

- GIHBEL, ' Zur Osteologie der Gattung Monasa,' Zeitsclir. f. d. yes. Naturw. 
xviii. 18G1, p. 1'21. 

PICI 189 

is connected with the anterior part of the dorsal tract by a 
single row of very small downy feathers. 

In Malacoptila fusca ' the oil (/hind has a few fine hairs 
at the apex' (NiTzscn). The aftershaft is absent. 

The tongue of Bucco macula tun is long, flat, and thin ; the 
tip is not lacerated ; the base is spiny. 

The muscle formula of the leg is given as AXY . Two 
carotids are stated to be present. 

The deep flexor tendons are like those of the Capitonidse. 
The flexor hallncis supplies digits I., II., IV. The flexor 
profundus supplies only digit III. The two where they 
cross are connected by a small vinculum. 

Monasa nigrifrons has fourteen cervical vertebra. The 
atlas is perforated for the odontoid process. The eighth 
vertebra has two catapophyses, which come into contact at 
their bases in the middle line ; on the ninth is the first 
median hypapophysis ; these continue to the first dorsal ; 
on the thirteenth cervical are large catapophyses, which are 
present, but much smaller, on the hypapophysis of the next 
vertebra. There are only three free dorsals, all of which 
bear ribs reaching the deeply two-notched sternum, which 
has a long spina externa. 

The clavicles are expanded into a wide, thin, roughly 
triangular plate at their attachment to coracoids ; there is a 
thin but broadish hypocleidium where they meet below. 

The skull is desmognathous, and in front of the conjoined 
maxillo-palatines is a gap of about half an inch in length 
in the bony palate. The process of the squamosal very 
nearly reaches the jugal bar. The lacrymal completely 
reaches the jugal bar in front, and the descending process is 
in contact, but not fused, with the strong and swollen 
ectethmoids. The interorbital septum is quite complete, as 
is the intranarial. 

The Rhamphastidae (toucans), with their large bills, 
serrate slightly at the edge, bear a certain superficial resem- 
blance to the hornbills, with which, however, they have no 
specially intimate connection. They are, on the other hand, 


undoubtedly nearly related to the Capitonidre, and show 
slighter but still recognisable points of affinity with the 
Passeres. The toucans, which are purely tropical American, 
consist, perhaps, of only one well-marked genus, Rliamphas- 
tos, which, however, has been subdivided into Pteroglossus, 
Aulacorhamphus, Selenidera, and some others. The toucans 
have a tufted oil gland, an aftershaft, and ten rectrices. 

The pterylosis (cf . NITZSCH) is characterised by the wide 
lateral neck spaces. 

The dorsal tract does not divide on the neck, nor is it as 
a rule l continuous throughout. There is a break between the 
straight anterior portion and the Y-shaped posterior portion. 
The ventral tract divides in the middle of the neck, and 
there is a strongly marked outer branch to the pectoral 
tracts. The femoral tract does not, according to my own 
experience, arise so early from the spinal tract as NITZSCH 

The intestinal tract of the toucans is short but volumi- 
nous, eighteen inches in Rliampliastos dicolorus, nineteen 
inches in R. carinatus. There are no caca. 

The proventriculus is zonary ; the stomach is a weak 
muscular bag. 

In the liver the right lobe is larger than the left, three 
times larger in R. carinatu*. 

The most characteristic feature in the anatomy of the 
family concerns the gall bladder ; this has been principally 
investigated by FoKBES. 2 It will be seen that the gall bladder 
is tubular, and of very great length (4'15 inches in a specimen 
of Rliampliastos dicolorus}, as also in the CapitonidcB and a 
fewPicidae (qq.v.) I found in Eh. tocard an exceptional state 
of affairs. As in all (?) toucans, the gall bladder is long, 
but from the upper extremity two separate cystic ducts arise, 
which soon fuse to separate again ; :J they open, however, in 

1 NITZSCH mentions as an exception an unidentified species. And in 
Selenidera and Aulacorliamplius I could find no break. 

- ' Note on the Gall Bladder, &c., of the Toucans and Barbets,' P. Z. ,S'. 
1882, p. 94. 

3 This may perhaps be regarded as a faint indication of the rctc found in 
some reptiles. 

PICI 191 

common with each other and with the hepatic duct into the 
intestine. There are two pancreatic ducts in this species, 
and one of them is peculiar in having a distinct, though 
short, diverticulum near to its intestinal orifice. 

The tongue is long, horny, and feathered along the 

The toucans as a rule possess only one carotid, the left. 
But in an example of Eh. dicolorus FORBES found both 
carotids present. The second is much smaller ; it is quite 
pervious, however, and blends with the left at the entrance 
to the neck. 

The syrinx- is in some respects peculiar. As in some 
other picarian birds, the last three or four tracheal rings are 
fused to form a bony box, which is marked by a deep furrow 
rnesially and behind. This region of the syrinx is much 
compressed antero-posteriorly. The intrinsic muscles are 
attached to the upper part of the box. 

The first bronchial ring is ossified in front ; posteriorly 
each half is L-shaped and fibrocartilaginous. The descending 
limbs of each L are in the same straight line as the trachea ; 
they are closely applied to each other, but really separated 
by a membrane of very short extent, the bronchidesmus. 
They are in close contact with the inner end of the ossified 
half of the ring. At the lower end of these cartilaginous 
pieces is a separate rounded bit of cartilage, which is simi- 
larly connected with the second ring. 

The tensores patagii are very simple. The tendon of the 
brevis is quite undivided. Each has a tendinous slip from 
the pectoralis. There is a cucullaris patagialis. 

The deltoid is particularly long in its insertion. 

The anconceus has 110 humeral attachment. It arises 
by fleshy fibres (not a Y-shaped tendon) from the scapula. 
The glut ceiis maximus extends below the acetabulum ; there 
is no glutceus externus. 

The muscle formula of the leg is AXY . The semitendi- 
nosus is attached to the leg bone separately from the serni- 
inembranosus ; it has also a tendinous insertion on to gastro- 
cnemius. The deep plantar tendons are as in Megal<Eina. 


There are fourteen cervical vertebra. The atlas is per- 
forated by the odontoid process ; strongish median hypapo- 
physes exist on vertebrae C11-D2. In Rliampliastos ariel 
ohere is in addition a paired haemapophysis on CIO, not 
present in Pteroglossus Wiedi. Five ribs reach the sternum, 
which is four-notched and has a spina externa. The clavicles 
are short and do not unite in the middle line ; they are ex- 
panded above, as in the barbets. There is a small accessory 
scapular nodule, from which deltoid in part arises. The 
skull is desmognathous, holorhinal, without basipterygoid 
processes. The ethmoids are large and quite fused with 
the descending process of the lacrymal, itself fused with 
the skull wall. The vomer is truncated at its anterior 

Fam. Capitonidse. This family of birds, with which I 
include, following GADOW, Indicator, is of the tropics, both 
Old World and New. 

All the members of the family that have been examined 
have ten rectrices, save Indicator, which has twelve. The 
oil gland is feathered, and there is an aftershaft. 

The pterylosis l shows some differences in different genera ; 
in all, however, the ventral tract is single as far as the 
posterior end of the neck, where it divides into the two pec- 
toral tracts ; each of these, again, gives off, as in the toucans, 
Bucconidse, Picidse, &c., an outer branch. This soon termi- 
nates, but the inner branches are continued as far as the 
cloaca. The apteria being entirely devoid of down feathers, 
the tracts are easily denned, but in spite of this NITZSCH 
has, in my opinion, fallen into some errors. 

The typical arrangement of the dorsal pterylosis 2 
(shown in tig. 91) may be considered to be that of Megalama, 
and was found to characterise the species M. asiatica, M. 
rireiis, M. Hodgsoni, M. javensis, M. Franklini. The spinal 
tract is single and narrow upon the neck, thus leaving con- 

1 GIEBEL, ' Pterylose von Tetragonops,' Zeitsclir. f. d.ges. Naturw. li. 1878, 

p. 377. 

- ' On the Pterylosis of Certain Barbets and Toucans,' P. Z. 8., 1890, p. 555. 



spicuous lateral neck spaces ; behind the scapula it forks, 
and the posterior part of the spinal tract is not in continuity 
with the interior. The former is at first a single tract, but 
it divides some way in front of the oil gland, which it 
surrounds, as in the Picidas and RhamphastidaB ; but in the 


former family there is also a median continuation of the 
spinal tract, which stops at the base of the oil gland. On 
each side of the spinal tract is a very narrow lateral tract, 
which is figured by NITZSCH as existing also in the Kham- 
phastidae and the Picidae. Xcuithol(cma rosea shows differ- 
ences from Megalcema ; the spinal tract divides in the usual 
way, but the tracts rejoin, so as to enclose a diamond-shaped 
space ; they then again diverge immediately, and end at the 
sides of the oil gland in the usual way. 

Xantlwlcema has a faint lateral tract on either side. In 
spite of NITZSCH'S figures I am disposed to think that this 
tract, so universal in the barbets, is at most feeble in the 
toucans. In the continuity of the anterior and posterior 



parts of the spinal tract Xantholcema is more toucan-like 
than any barbet. 

In Capita and Trachyplwnus NITZSCH has figured an 
anterior bifurcation of the posterior part of the spinal tract, 
much more marked in the former genus, where, indeed, I am 
inclined to suspect a more detailed resemblance to XantJw- 
la-ma than is suggested by NITZSCH'S figure. If this is not 
the case, the New-World genus will be distinguished from 
the Old- World genera by a double posterior spinal tract 
divided by a break from the forked anterior tract. 

The tensor espatagii consist of a slighter tensor longus and 
a wider brevis ; the tendons of both are simple, and each is 
reinforced by a tendinous slip from the pectoralis. The 
brevis tendon is simple, and terminates, as in the last two 
families, upon the tendon of the extensor metacarpi radialis. 

In Capita, at any rate, there is a cucullaris propatagialis. 

FURBEINGEE had indicated a point of difference among the 
barbets which concerns the rhomboideus profundiis. This 
muscle in Meiglyptes (Picidse) and Ehamphastos is a large 
fan-shaped muscle, arising from the tip of the ilium, as 
well as from the spinous processes of certain of the dorsal 
vertebrae, and is inserted from the tip to about halfway down 
the scapula. In Capita this muscle is plainly divided into 
t w o an anterior and a posterior of which the latter arises 
from a few vertebrae and from tip of ilium. In Megalcema 
the anterior section of this muscle is itself again divided 
into two quite distinct parts. This family, like the wood- 
peckers, toucans, and Passeres, has a latissimus dorso- 
cutaneus, but no metapatagialis. 

The deltoid extends some way down the arm, as in other 
allied birds and in Passeres ; it has no special scapular slip. 
The barbets, like the woodpeckers and toucans, possess a 
sesamoid ossicle, called the ' scapula accessoria,' which is 
developed in the scapulo-humeral ligament ; from it arise 
some of the fibres of the deltoid. This bone, again, is also 
found in Passeres and owls, and in a host of other birds. 1 The 

1 See FURBRIXGEK, p. 229. FORBES laid too much stress upon its classifica- 
tovy value in the present case. 


maxim us is both pre- and post-acetabular in origin. 
There is no gl. externu*. In the leg the formula is the 
typical picarian one of AXY . Both peroneals are present, 
as in the Rhamphastidse, and with the normal attachments. 
The deep plantar tendons have been already described and 
displayed (fig. 58, p. 102). 

The tongue (in M. virens) is sagittate, with horny 
margins, and slightly lacerated apically. In a specimen 
of Megaltzma asiatica the horny apex was bifid, quite 
regularly so. 

The right lobe of the liver in M. virens is a little larger 
than the left. The greatest peculiarity of the liver, however, 
concerns the gall bladder. As in the toucans and some 
Picidte it is of great length and intestiniform, i.e. of narrow 
and regular calibre ; in a specimen of M. virens it was two 
and a half inches long. The presence of a similar gall 
bladder has been noted in M. Franklini and in XantJw- 
l<nna rosea. 

The intestines (devoid of cseca) are voluminous but short 
seventeen inches in M. virens, twelve in M. asiatica. 

The syrinx of I\Iegal&ma (asiatica) is of a very simple 
tracheo-bronchial form ; the last rings do not fuse at all, but 
remain perfectly distinct ; there are no intrinsic muscles. 
The extrinsic muscles are attached to the tips of the costal 

The skull is ' segithognathous, with a desmognathous 
tendency,' holorhinal, and without basipterygoid processes. 
The desmognathous tendency is shown by the fact that (in 
Mcgahcma asiatica) the maxillo-palatines may or may not 
blend with the nasal septum. In others, e.g. Pogonorhyncliit* 
bidentatns, the two bones (maxillo-palatines) blend com- 
pletely across the middle line. These forms are, therefore,, 
genuinely desmognathous, except as concerns the vomer. 
This bone is truncated, as in the segithognathous skull, and 
its two forward limbs join the maxillo-palatines, as in Indi- 

It is single (except, of course, for the anterior bifurcation) 
in ; broader and double in Gumnobucco calvus. 

o 2 


As GARROD has pointed out, the vernier of the Capitonidae 
differs from that of the Passeres in being truncated behind 
the posterior line of the palatines ; in the Passeres the 
truncation is in front of the line. The skull of Megaltenia. 
asiatica has been described and figured by PARKER. 1 Outside 
the Y-shaped vomer there is on each side a small ' septo- 
maxillary ' splint, and in front a median unpaired septo- 
maxillary. The former are, perhaps, the equivalents of the 
vomers of Picidse (q.v.) ; the vomer in that case may have 
something to do with the medio-palatine of PICKS. The 
nostrils are impervious and very much reduced by long 
growths. In Megalcema virens and Gymnobucco the lacry- 
mals and pref rentals entirely fuse with each other and with 
the skull walls to form a solid and imperforate plate of bone, 
as in Pterocles, pigeons, toucans, &c-. 

There are fourteen cervical vertebrae. The atlas is per- 
forated by the odontoid. The axis and the three following 
vertebrae have median hypapophyses ; the next vertebra (the 
sixth) has a bifurcate one. All these processes are more 
conspicuous than in the toucans, where, however, they are 
present. Vertebras C11-D6 (last) and LI have strong 
hypapophyses, that of D2 being bifurcate. In Gymnobucco 
only three dorsals have hremapophyses. Five ribs reach the 
sternum, which is four-notched and has a spina externa ; the 
clavicles do not meet below. The head of the clavicle 
expands into a wide flat triangular plate. 

Indicator is considered to be the type of a distinct sub- 
family of the Capitonidae, which also contains the genus 
Protodiscus. Indicator has been chiefly investigated by 
GARROD. 2 NITZSCH, however, previously and FURBRINGER 
subsequently added to our knowledge of the bird. It has 
twelve rectrices ; the pterylosis has a marked gap between 
the anterior and the posterior parts of the spinal tract ; and 
the latter appears to encircle the oil gland, as in the 

1 ' On the Structure and Development of the Bird's Skull,' Trans. Linn. 
Soc. (2), i. p. 122. 

2 ' Notes on the Anatomy of Indicator major.' 1 P. Z. S. 1879, p. 930. 

PICI 197 

The syrinx is shown in the figure on p. 61 (fig. 40). The 
trachea at its lower end is formed into a bony box by a fusion 
of several rings ; the first bronchial semi-ring, to which 
the single pair of intrinsic muscles are attached, is larger 
than those which follow. The muscular anatomy, excepting 
for minute details, is the same ; one of these small details 
is that the pectoral slip to the tensor patagii is muscular 
and not tendinous. 

The skull is eegithognathous ; the vomer is forked in 
front and joins the separated maxillo-palatines, as in 
Megalama asiatica ; but GABROD has remarked upon the 
fact that in this character (the widely separate maxillo- 
palatines) the palate of Indicator is more like that of the 
Picidte than are the Capitonidse. 


Definition. Aftershaft absent. Muscle formula of leg, AX. Cseca 
absent. Both, carotids present. Skull desmognathous. 

The kingfishers form a natural assemblage of birds 
Alcedinidae which, however, show some variations in their 

Thus, while they generally agree with the bulk of their 
relations in having a characteristically tufted oil gland, 
tltr genus Tanysiptera (three species at any rate) has a nude 
oil gland. In this genus too the rectrices are ten, twelve 
being the more usual number. 

As to pterylosis, the kingfishers are characterised by the 
possession of down feathers not only on the pterylae but 
also on the apteria. There is a difference between Alcedo 
and Dacelo in that the latter has a weak dorsal tract in the 
middle region of the back, it being stronger in front and 
again near the oil gland. In Alcedo the tract is of uniform 
strength throughout. The kingfishers, like the swifts, 
sometimes possess and sometimes do not possess the fifth 
cubital reniex. 

In Dacelo gig ante a there are fifteen cervical vertebrae 
and only three ribs reach the sternum. In Dacelo Gaudi- 


chaudi there are fourteen cervical vertebrae, and four ribs 
reach the sternum a difference which appears to me to 
justify the separation of the latter into a distinct genus, 
Sauromarptis. Halcyon smyrnensis agrees in these points 
with D. gig antea. 

The last cervicals (from 10 J ) and the dorsal vertebrae 
(including the first lumbar) have well- developed, often 
bifurcate, sometimes trifurcate haemapophyses. The four- 
notched sternum has a well-developed spina externa. 
As Dr. CUNNINGHAM has pointed out, the lacrymals of 
Dacelo are larger than those of Ceryle? and, I may add, of 
Halcyon and Todirhamphus (= Sauropatis) . The skulls of 
the genera above mentioned are remarkable for the fact that 
the temporal fossae as nearly as possible meet behind : there 
is only a slight bridge dividing them. The lacrymal, 
moreover, dilates into a wide plate, with a notch, on the inner 
side of which is received the small flat-pointed prefrontal 
process of the ethmoid. I have observed these skull cha- 
racters in the Coraciidae. The clavicle reaches the scapula 
and gives off a longish acromial process. The nares are 

In Sauropatis albicilla, Dacelo gigantea, Cittura cyanotis, 
and Pelargopsis (FURBBINGER) the tensores patagii have 
the somewhat complicated arrangement shown in the figure 
(fig. 94). There are two tensor brevis tendons, of which 
the anterior, after giving off a wristward slip, is continued 
over the arm to the lower side, fanning out as it goes. 
There is in all a passeriniform tendinous slip. Dacelo has 
a muscular pectoralis propatagialis. Sauromarptis Gaudi- 
chaudi and Sauropatis sanctus are quite similar. There is, 
in Pelargopsis at least, a tendinous slip from pectoralis I. 
to both longus and brevis tendons. 

In Callalcyon rufa (fig. 93) there is a simplification, 
only the anterior of the two tendons being present ; the 
passeriniform slip is barely marked. In Alcedo there is a 
still further ' degeneration ; ' not only is the passeriniform 

1 On CIO there is a paired hasmapophysis in Dacelo. 

- SHUFELDT, 'On the Osteology of Ceryle,' J. Anat. Phys. xviii. p. 27'.). 



slip absent, but the wristward slip is hardly shown. There 
is a fleshy pectoralis propatagialis joining the longus tendon. 

In Ccnjlc alcyon the two tensor brevis tendons form a 
broad diffused tendon, to which is joined, before it gives off 
the wristward slip, a peculiar long (hornbill-like) pectoralis 
propatagialis tendon. 

In Alcyon Lessoni we have the most simple form of 
tensor brevis without any 1.) ranch. 

OF Ccrylc alcijon. PATAGII OF Callal- 

cyon rufa. 

OF Sauropatis albicilla. 

In Syma the tensor brevis consists of two tendons, but 
the anterior has no wristward slip at all. 

There is never a biceps slip. The deltoid has a scapular 

The expansor secunda riorum appears sometimes to be 
absent. But it is present in Dacelo, Tanysiptera, Syma, and 
Cittura. Dacelo, at any rate, has no humeral slip to the 

The leg muscle formula is, without exception, AX - . 
In Dacelo there is but one peroneal, which is the brevis. 
The deep flexor tendons vary somewhat. In Dacelo 
gigantea the arrangement of these tendons is, as Professor 


GAREOU pointed out, precisely that of Momotn-s. But in 
Halcyon vagans the two tendons blend completely before 
the branches to the toes, all arising approximately at the 
same level, are given off. The structure of the conjoined 
tendon, however, seems to suggest that the flexor hallucis is 
concerned with the supply of digits III., IV. 

In some kingfishers there is a myological peculiarity not 
found in any other group of birds. Dr. R. 0. CUNNINGHAM 
discovered in Ceryle stellata a strong transverse tendon 
uniting the two biventri cervicis close to the upper belly of 
the muscles. He failed to find this junction in Alcedo ispida 
and Dacelo gigantea. T any sip t era and Cittura l have this 
link, but not Syma, Halcyon, or Sauropatis.' 2 

The following table gives the intestinal lengths of two 
species : 

Ceryle amazona . . 24 inches 

Halcyon sp. 

The right lobe of the liver seems to be always larger 
than the left. 

I have examined the syrinx in Dacelo cervina. The last 
tracheal rings are completely fused in front to form a bony 
box, which shows no traces of the number of rings of which 
it is composed. These rings appear to be five or six in 
number, and, with the exception of the last, are fused 
together in the middle line behind. In front of this box the 
tracheal rings interlock in the usual fashion. The first 
bronchial semi-ring, which is ossified, is firmly united to, but 
not fused wilh, the tracheal box ; the succeeding rings are 
cartilaginous. The syrinx has two pairs of intrinsic muscles ; 
the most anterior is the more slender ; the wider muscle 
arises from the trachea, just where the extrinsic muscles are 
given off, and is attached to the first and apparently also to 
the second bronchial semi-ring. 

In Ceryle alcyon there are no great differences, but the 

1 In one of two specimens of S. vayans it was present. 

- ' Notes 011 some Points in the Anatomy of the Kingfisher,' P. Z. S. 1870, 
p. 280. See also BEDDARD, P. Z.. S. 189G, p. (508. 


rings forming the box are not so completely fused, and the 
larger intrinsic muscle arises n 
the commencement of the box. 

larger intrinsic muscle arises much lower down in fact, at 


Definition. Aftersliaft present ; oil gland tufted. Muscle formula, 
AXY ; biceps slip present. Caeca absent. Skull desmognathous. 

Of the family Coriidse there is only a single genus, African 
in range and including something like nine species. 

The toes aTe remarkable for the fact that the zygodacty- 
lous condition can be assumed ; the first toe can be directed 
forwards, 1 the fourth backwards. There are ten rectrices ; a 
tufted oil gland and an aftcrsJiaft are present. 

The ptcrylosis described by NITZSCH is remarkable for 
the width of the pterylae. The ventral tract almost com- 
pletely covers the ventral side of the body ; towards the 
outside the feathers are stronger, but there is no outer branch. 
The spinal tract is narrow and strongly feathered upon the 
neck ; on the occiput is a bare space, reminding one of that 
in a similar position in the Trochilidse. There is no median 
spinal apterion. 

In their myoluyy the colies are remarkable for possessing 
a fleshy biceps slip. The tensor patagii brevis muscle is very 
extensive, and reaches nearly AS far as the fore arm. Its 
very short and single tendon sends back a ' passerine ' slip, 
oblique in direction, and is also continued over the arm. 

The pectoralis slips to both longus and brevis are present. 

The deltoid, as in so many allied birds, is very extensive. 

The leg muscular formulaic AXY . There is only one 

The seinimembranosus is inserted below and indepen- 
dently of the semitendinosus. The latter gives off a tendinous 
slip to the gastrocnemius. 

The deep flexor tendons blend before giving off branches 
to the toes. 

1 Hence the term ' .pamprodactylous,' sometimes used for this family. 


Besides the hornbills and Macrochires the colies are the 
only flying birds in which the latixxli/uts dorsi metapatagialis 
is absent. 

The colies have only the left carotid. 

The stomach is not very muscular. The liver is small 
and has a gall bladder. There are no cteca. The intestines 
are short, but capacious, measuring nine inches. 

The syrinx has been figured by JOHANNES MULLEE.' It 
is quite typically tracheo-bronchial. 

The skeleton and the affinities of Colius have been elabo- 
rately treated of by MuRiE. 2 

There are thirteen cervical vertebrae. Four ribs reach the 

The skull is holorhinal, without basipterygoid processes, 
and desmognathous. After a careful maceration GAREOD 3 was 
unable to find a vomer, the presence of which had been 
previously 4 asserted (see fig. 95, p. 203). 


Definition. Feet zygodactyle by reversion of second toe. Skull 
schizognathous "with, basipterygoid processes. Oil gland nude. 
Left carotid alone present. Caeca short. Ambieiis absent. Of 
deep plantar tendons Fl. hall, supplies I. and II., Fl. dig. III. 
and IV. Vinculum joins them before bifurcation of each. 

This family is chiefly American, but also African and 
Asiatic. 5 

The feathers of the trogons have very well developed 
aftershafts. The pterylosis is remarkable for the non- 
bifurcation of the spinal tract, which is continuous as a single 
tract to the base of the naked oil gland. It is dilated to form 
a rhomboidal area behind the scapula. 

There are twelve rectrices. 

1 ' Ueber die bisher unbekannten typischen Verschiedenheiten der Stinim- 
organe der Passerinen,' Abli . A\ Abaci. Wiss. 1845. 

'-' ' On the Genus Colius, its Structure aud Systematic Place,' Ibis, 1872, 
p. 263. 

: < ' Notes on the Anatomy of the Colies (Colius),' P. Z. S. 1876, p. 416. 

1 By MURIE. 

' Trogon gallicus is an extinct species from the Miocene of France. 


Of the muscles of the thigh which Professor GARROD 
regarded as of importance there are present the femorocaudal 
and the semitendinosus, the accessories of both being absent. 
The femorocaudal is proportionately larger than in almost 
any bird. There is no glut ecus primus. The obturator 
interims is small and oval. The singular arrangement 
of the deep plantar tendons 
is used in the definition of 
the family. The two tendons 
concerned each supply two 
digits, this arrangement being 
unique. In the fore limb 
there is no biceps slip to the 
patagium. The patagial muscles 
and tendons are much compli- 
cated ; they have been figured 
by GARROD for Trogon puella. 
The very powerful tensor 
brevis muscle runs as a muscle 
nearly to the extensors of the 
fore arm ; it has a short broad 
tendinous insertion on to the 
fascia of the outer surface of 
the arm, and this is specially 
developed, a line running back 
to the humerus, as in the 
Passeres (see p. 172). Deeper 
than this are tw r o parallel 
tendons : of these the one nearer the humerus terminates 
exactly like the single one of the Passeres ; the other tendon 
ends as in the Pici, elsewhere described. There is no expansor 

The tongue of the trogons is short and three-sided. It 
is pointed in front. The left lobe of the liver is a little the 
smaller. Among GARROD' s notes are the following measure- 
ments of the intestines and the caeca in three species of 
trogons, which we reproduce : 

FIG. 95. SKULL OF Colitis caxtano- 



Trogon niexicaims 

Tr. puella I'liaromacrus mociuim 

Intestine . 

1O5 inches 
1-25 and 1 inch 

8-5 inches 
1-1 inch 

16 inches 
1'75 inch 

There is no crop ; the gizzard is thin-walled and large ; the 
proventriculus is zonary. 

The most remarkable matter concerning the osteology of 
the trogons is the curious mistake which was originally made 
as to the nature of the palate. HUXLEY, in his paper upon the 
classification of birds, came to the conclusion, from a single 
incomplete skull of Trogvn Heinwardti, that the skull was, 
like its presumed allies, desmognathous. Later FORBES l was 
able to show in five species that the maxillo-palatines were 
not united across the middle line, but that they terminated 
in a spongy expansion some way from each other. The end 
of the vomer, is thin and filiform. The lacryrnal is somewhat 
styliform, and reaches the jugal bar ; there appear to be no 
ossified ectethmoids. The palatines of the trogons are 
peculiar. Instead of being flat plates, as in Cor&cias, for ex- 
ample, the outer portions of the bones are bent upwards, and 
cling closely to the basis cranii. The tw r o palatines are, 
moreover, fused posteriorly, and the pterygoids where they 
articulate with them are expanded. They are holorhinal 
with impervious nares. The trogons have fifteen cervical 
vertebra. The atlas is perforated by the odontoid process ; 
four or five ribs reach the sternum. The sternum has two 
incisions behind, and the bifid spina externa. 


Definition. After-shaft present Muscle formula, AXY ; expansor 
secundariorum present. Cseca generally present. Desmogna- 

The Coraciidse are entirely Old-World birds, chiefly 
massed in the Ethiopian region, but extending as far as the 

1 ' Note on the Structure of the Palate in the Trogons,' P. Z. tf. 1881, 
p. 83G. 


Australian. The genera allowed by DBESSER in his recent 
monograph of the family are Cort/dax, Eurystomus, Brachy- 
pteracias, Atelornis, and Leptosomus. They are distributed 
by him in three subfamilies ; the first two genera constitute 
the first, the next two the second, while Leptosomus is placed 
in a subfamily by itself. The rollers have an anisodactyle 
foot ; the feathers have an after shaft ; but the oil gland is 
nude. The pterylosis has been studied by NITZSCH, FORBES, 
and by myself. 1 In Eurystomus orient alis the ventral tracts 
commence as two from the very first ; at the angle of the 
mandible they are double. Though NITZSCH has figured 
the pterylosis of the throat of Coracias garrulus as if it were 
continuous, I do not find any difference from Eurystomus in 
this particular. On the breast the two divisions of the 
ventral tract remain single ; there is hardly a trace of the 
outer branch. The tracts are here rather wide. The dorsal 
pterylosis narrows gradually until between the shoulders, 
where the feathering is very strong, and where it divides 
into two branches ; these unite again just at the articulation 
of the femora, and finally terminate a little way in front of 
the oil gland. 

Leptosomus has a slightly different pterylosis ; the 
ventral tract is single to about an inch behind the junction 
of the mandibular rami ; for a considerable distance the 
ventral band is continuous with the dorsal, so that the 
lateral neck spaces do not commence until about three- 
quarters of an inch above the shoulder. About the middle 
of the sternum the pectoral tract of either side gives off an 
outer branch some four feathers wide and slightly stronger 
than the main tract. The two forks of the dorsal tract 
run in between each other, the narrower posterior portion 
between the limbs of the wider anterior portion, as is the 
case with so many birds. FORBES has noted that Atelornis 
has a pterylosis which agrees with that of Eurystomus, 
already described. 

Leptosomus differs, however, from the remaining Coraciidae 
in the possession of powder-down patches, which were first 

1 See anatomical preface to DRESSER'S monograph of the group. 



described by ScLATEE. 1 There are two patches in the 
lumbar region, lying between the dorsal and the femoral 

The tongue of the rollers wants the spiny fringe at the 
base which is so common a feature of this organ in other birds ; 
it is horny in front and entire at the tip. The liver has a 
larger right lobe and a gall bladder. 

The following are intestinal measurements:- 

Small Intestine 

Large Intestine 

2| inches 
1 inch 


1 ,<<!>! osomus . 

9f inches 

1 2J, 2f inches 
H inch 
2-2 inches 

The tensores patagii of Leptosomus are figured in the 
accompanying cut. 2 The brevis tendon gives off a wrist- 

ward slip just before its attach- 
ment to the tendon of extensor 
metacarpi radialis longus, over 
which it passes to be inserted 
below. There is no biceps slip. 
Coracias gar ml us is the 
same save for the fact that 
there are two separate brevis 
tendons, from the first of which 
the anterior gives off the 
wristward slip ; the inner 
thinner tendon of these does not 
cross the wrist. Eurystomus 

does not differ. In these birds 
FIG. 9(5.- -TENSORES PATAGII OF Lev- ,, "11 T i 

tosomus (AFTER FORBES). there is a well-developed ex- 

of the 

/', lutiifdum ;; tensor patagii brevis: pailSOT, extensor metacarpi ;?, its origin. . 

fully developed type, which 
Professor GAEEOD has called ' ciconiiform.' 

The muscle formula of the leg is AXY . The glut ecus 
maximum does not reach below the acetabulum ; the gl. 

1 ' On the Structure of Lcptosoma discolor,' P. '/.. N. LSI Jo, p. (382. 
- From FORBES, 'On the Anatomy of Leptasoma discolor,' P. Z. S. 1880. 
p. 4(i4. 


externu* is present in Eurystomus, represented by a ligament 
in Coracias. The deep flexor tendons are of type V., where 
the two tendons blend before the four branches are given off 
to the TSoth peroneals are present. The carotids of the 
CoraciidcTB are two. In Leptosomus FORBES found that the 
two arteries run up close together, but are not fused in the 
hypapophysial canal. He thinks that they may be, like 
those of Bu-corvns, 1 no longer functional as blood vessels. 

The syrinx of the Coraciidae is quite typically tracheo- 
bronchial. In C. garrulus the intrinsic muscles are attached 
to the first bronchial semi-ring. These semi-rings are ossified ; 
the rest of the bronchial semi-rings are more slender and not 


ossified. In Eurystomus the only difference is in the fact that 
the three semi-rings following the first are closely attached to 
it and to each other, and appear to be ossified ; after these 
are the broader soft cartilaginous semi-rings. The syrinx of 
Leptosomus (fig. 97) is rather different ; it appears to be ' 
an extreme development of the type found in Eurystomus. 
The first three bronchial semi-rings, like the last tracheal 
rings, are ossified ; the first of them appears to be nearly, if 
not quite, a complete ring. The fourth and the succeeding 
semi-rings are cartilaginous ; to the former are attached the 
intrinsic muscles. In the commencing formation of a 
' bronchial syrinx ' Leptosomus evidently gives a hint of 
cuckoo affinities, to which group, however, its structure in 
general does not incline. 

There are fourteen cervical vertebrae in Leptosomus, 

1 Cf. p. 215. 


thirteen or fourteen in other Coraciid&e. The atlas (Coracias) 
is notched for the odontoid process ; C2-4, C10-D2 have 
hsemapophyses. On C13 and 14 there are also a pair of 
downward processes (catapophyses), one on each side of the 
hsemapophysis, which, in the case of C14, arise from a 
common base with it and on Dl from its tip. 

In Eurystomus the atlas is perforated. Four (Lepto- 
sonius ! ) or five (some other forms) ribs reach the sternum, 
which is singly or doubly notched on either side, and has a 
spina externa but no spina interna. The skull is desmogna- 
thous, holorhinal, without basipterygoid processes. The 
rollers have the same peculiar lacrymal that has been referred 
to above in the kingfishers. The bone expands enormously 
below and comes into near relations, but does not fuse, with a 
slight ectethmoid ; the lacrymal reaches the jugal. Another 
peculiarity of the coraciid skull is the very large postfrontal 
process, which descends in a straight line and actually 
reaches the jugal. These remarks apply not only to Coracias, 
but to Eurystomus and Atelornis, in which latter, however, 
the postfrontal process is not quite so long. 

The family Meropidae consists of the genus Merops, and 
of a few others which have been carefully monographed by 
DRESSER. Like the rollers the bee-eaters are an exclusively 
Old-AVorld family, ranging through the Patearctic, Ethiopian, 
Oriental, and Australian regions, but again, like the rollers 
predominating in the Ethiopian. 

As to external characters, the oil aland is nude ; the 
rectrices are twelve ; the feathers have an aftersliaft. The 
pterylosis (described by NITZSCH and by myself' 2 ) is as 
follows : 

The spinal tract is wide, and is at first connected round 
the neck with the ventral tract. About halfway down the 

1 The osteology (and some of the viscera) of Leptosomus is described and 
figured by MILNE-EDWARDS in the Histoirc Naturdl' tic Madagascar. See also 
for the family NITZSCH and GIEBEL, ' Znr Anatomic der Blauracke,' Zcitschr. 
f. d, gcs. Naturw. x. p. 318. 

- In anatomical preface to DKESSEH'S monograph. 


neck it becomes separate, and terminates in a truncate or 
sometimes bifurcate extremity between the shoulder blades ; 
at this point there is a break and the rest of the spinal 
tract is double, enclosing a space bounded by two distinctly 
conical tracts, which gradually get narrower to their point 
of fusion, a little way in front of the oil gland. The ventral 
tract is double from close to its point of origin ; the two 
tracts get wide upon the pectoral region, whence they 
gradually dwindle to a single feather wide close to their 
termination ; the pectoral tracts have no outer branch. 

Nyctiornis has two carotids ; most of the others have 
only one, the left. The bee-eaters have long cceca, like the 
Coraciidae. In a specimen of M. ornatus with the intestines 
only 5^ inches they measured iinch; in a larger species 
1 inch. 

The proventriculus is zonary ; the right lobe of the liver 
the larger, and with a gall bladder. 

The tensor patagii b rev is tendon gives off a wristward 
slip, and a ' passeriform ' slip to the humeral at its insertion 
to the fore arm, which it does not cross. There is no biceps 
slip, but there is a fleshy slip to longus from pectoralis ; that 
to brevis is entirely tendinous, there being in both an 
agreement with Coracias. 

The deltoid extends a long way down the humerus ; it 
receives a tendinous slip from the scapula, which passes under 
the latissimus dorsi and over the anconaus longus. The 
latter muscle has a humeral head, but not in Nyctiornis. 

The expansor secundariorum is present and ' ciconiine.' 

The leg formula is AXY . The deep flexor tendons are 
as in fig. 55, p. 100 ; the flexor hallucis gives off a slip to the 
hallux before it fuses with the flexor communis. 

The Meropidse have fourteen cervical vertebrce. The 
atlas is perforated by the odontoid process. There are 
hsemapophyses on C2-4, C10-D1 ; that of C14 is trifurcate. 
Four ribs reach the sternum, which has two lateral notches, 
of which the outer is the deeper, and has both external and 
internal spina, the latter being bifid, as in Passeres and 
some other birds. The clavicles have an acromial process,. 


as in the kingfishers. The skull is desmognathous,holorhinal, 
without basipterygoid processes. The descending limb of 
the lacrymal nearly unites is connected by cartilage with 
the slender ectethmoid, thus forming a ring. The nostrils 
in the dried skull are pervious. 

The desmognathism of Merops is different from that of its 
allies. The maxillo-palatines are long, slender, recurved 
plates, like those of passerines. They are fused in the middle 
line to a broad plate of bone, but the free ends of the maxillo- 
palatines extend backwards for some distance independently 
of this. The palate in front of the maxillo-palatines is to 
some extent vacuolate. The vomer is a single rodlike bone. 

The Momotidae ' are South and Central American, com- 
prising the genera (perhaps subgenera) Momotus, Hylomanes, 
Baryphthengus, and some others. They are placed by 
GADOW in close association with the todies, but there are 
various points in which they differ from that family, upon 
which stress has been laid by FOBBES. It is mainly to the 
last-mentioned observer 2 and to GABBOD 3 that the existing 
knowledge of the family is due. The external characters of 
the family show some variation ; in Momotus the oil gland 
is quite nude ; in Hylomanes and Eumomota the apex is 
furnished with a few small plumes. Momotus has twelve 
rectrices ; Hy, Prionorhynchus, Baryphthengus, ten. 
A remarkable characteristic of the motmots are the two 
central racket-shaped rectrices, which matter was investi- 
gated twenty years ago by SALVIN. 4 It appears that 
the original account given by WATEBTON of the birds nibbling 
off the vanes is perfectly correct, for it was observed by 
BABTLETT at the Zoological Society's Gardens. As a rule 
the bird only nibbles at the two long central rectrices, but 
SALVIN reports a case where a bird had sought fresh fields 
and had attacked others of its feathers. It is a very remark - 

1 J. MUEIE, ' On the Motmots and their Affinities,' Ibis (3), ii. 1872, p. 383. 
- Collected papers, passim. 

3 Collected papers, passim, ' On the Systematic Position of Momotidce,' 
P Z. S. 1878, p. 100. 

1 On the Tail Feathers of Momotus,' P. 7,. S. 1S73, p. 429. 


able fact that when the rectrices in question first appear 
they are narrower at the points where the nibbling occurs, and 
where they will be ultimately denuded, than they are else- 
where. But an inheritance of this particular acquired 
character can hardly be asserted. 

An a/tcrshaft is present, but small. 

The tensores patagii are simple, and there is no biceps 
slip. There is a fleshy pectoralis propata(//<i!/s. The tensor 
brevis consists of two parallel tendons, the anterior of which 
does not give off a wristward slip. The fan to the ulna arises 
in M. brasiliensis and M. cequatorialis as a continuation of 
the hinder of the two tendons, in M. Lessoni between them. 
Hijlomanes gularis agrees with the first. The anconceus has 
a humeral slip. The somewhat rudimentary cxpansor 
secundariorum only reaches the margin of the teres. 

The deltoid is large ; its attachment to humerus is longer 
in Hijlomanes than in Momutus (f-i). There is a separate 
tendinous scapular slip. 

The muscular formula of leg is AXY . Both peroneals 
are present. The deep flexor tendons of the motmots are 
rather peculiar in their structure. It will be observed that 
the slip to the hallux is given off before the flexor hallucis 
joins the flexor longus. 

The first gluteal (gl. maxim its) is only present in front of 
the acetabulum. The glittfcus externns is absent as a distinct 
muscle, but the insertion of glutceus II. extends so far round 
the head of the femur that it may represent also the other- 
wise missing muscle. 

The gizzard is stronger in Hijlomanes than in Momotux, 
and is almost ' ptilopine ' in section. 

The tongue is long, bifid at the apex, and worn into fila- 
ments. In the alimentary canal the caca are absent ; the in- 
trxtincs measure fourteen inches in M. Lessoni, eighteen inches 
in M. brasiliensis. The right lobe of the liver is the larger, 
and there is a gall bladder. 

There are two carotids. The femoral vein is abnormal. 

The syrinx has been described by GARROD and is figured 
by him. It does not apparently differ widely from genus to 


genus. The last few tracheal rings are fused, but there is 
not a complete pessulus. 

The motmots have fifteen cervical vertebra. The atlas is 
perforated by the odontoid process. Cl-5, C11-D3 have 
median hypapophyses ; on C14-D2 are a pair of inferolateral 
processes, which gradually approach the median line until 
they spring from a common base in Dl , and are just visible as 
rudiments towards the tip of the haemapophysis of D2. In 
Hylomanes there is also a double hypapophysis on CIO. Three 
or four ribs reach the sternum, which has (Momotus) four 
foramina, or (Hylomanes) two notches and two foramina, 
and a bifid spina externa. The skull is desmognathous and 
holorhinal. The lacrymals are rudimentary ; the ectethmoids 
are very small and do not nearly reach them. Nares 
impervious, pervious in Hylomanes. 

The West Indian todies (Todidae) form a very distinct 
family ; their structure has been chiefly investigated by 
MURIE i and FOEBES.* They are small birds with feet in 
which the syndactylism is more marked than in motmots and 
some others. The annexed cut shows that the digits IV. 
and V. are united together as far as the end of the third 
phalanx of the one and the second of the other. The 
oil gland is tufted. 

The skull is very imperfectly desmognathous. The two 
maxillo-palatines are not united together ; they are com- 
pletely free from each other and from a median ossified nasal 
septum. There seems to be no vomer. The descending 
portion of the lacrymals is large and broad ; the ectethmoids, 
on the other hand, are small. The interorbital septum is 
widely fenestrate. There are fifteen cervical vertebrae. The 
intestines are, according to Mr. FOEBES, remarkably short, 
not measuring altogether more than 3^ inches ; on the other 
hand the cceca are as remarkably long (considering the sys- 

1 ' On the Skeleton of Todus,' &c., P. Z. S. 1872, p. 664. 

- ' On some Points in the Anatomy of the Todies (Tocliclse), and on the 
Affinities of that Group,' ibid. 1882, p. 442. See also BEICHEXOW, ' Ueber daa 
Genus Todus,' &c., Journ. f. Ornith. xxxi. 1883, p. 430. 


teniatic position of the bird) ; they measure about one-third 
of an inch. The cseca are narrowed at their origin from the 
gut, and, as in the owls, &c., dilated apically. The deep plantar 
tendons vary from the arrangement common to the group in 
that the slip to the hallux is given off before the blending of 
the two. The arrangement, in fact, is as in the motmots. 
There is an expansor secundariorum ceasing at the axillary 
margin of the teres, in the gallinaceous fashion. The syrinx 
has at the middle a bony box, which is formed of three or four 
bronchial rings united with about two tracheal rings. It is 

FIG. 98. FOOT OF Todus (AFTEK 

FIG. 99. FOOT OF Momotiis 
AS IN FIG. 98. 

only ventrally that the fusion is complete. The intrinsic 
muscles cease at the last tracheal ring but one. 

The Galbulidse are a family of neotropical birds, com- 
prising the genera Galbula, Urogalba, Jacamerops, &c., and 
known as puff birds. 

They are zijgodactyle, with a nude oil gland, twelve rec- 
trices, and a small after shaft. 

The pterylosis of Galbula rufoviridis is as follows : 

The inferior tract is double from the angle of the jaw ; 
just before leaving the neck each tract gives off a short 
branch, about six rows of two feathers, which runs on to the 


margin of pectoralis. The main tract itself is also only two 
feathers broad. It sends off, about halfway down sternal keel, 
a short curved outer branch, which runs outwards and then 
forwards towards axilla, nearly meeting a second outer branch 
which is given off by the tract on opposite side to the inner 
branch, already spoken of. The dorsal tract has a slight 
break ; the very short interscapular fork is of strong feathers 
and connected with posterior part only by a very few 
feathers which are weak and arranged uniserially. 

The tongue is long and thin, tapering to a filament ante- 
riorly ; a gall bladder seems to be absent. In a specimen of 
G. rufoviridis the intestinal measurements were as follows : 
s. i., 4 g 15 inches ; 1. i., -75 inch; caeca, -7 inch. 

The Galbulidse have an expansor secundarioniin, but no 
biceps slip. The tensor patagii brevis tendon of Galbi/lah'&s 
no wristward slip. It is merely a single tendon ; in Urogalba 
there is a wristward slip. 

In the leg the formula of Galbula is AXY, of Urogalba 
AX, both birds, of course, lacking the ambiens. ThegliitfCiiH 
I. and V. are absent, at any rate in Galbula. The plantar 
tendons are picine. 

Both carotids are present. 

The skull of the Galbulidte is very like that of the Bucco- 
nida? ; but there are nevertheless points of difference. 

In Urogalba paradisea there is a long gap in the bony 
palate in front of the conjoined maxillo-palatines, as in Bucco- 
nida? ; but the palatines are more sloped off posteriorly, and 
their posterior halves are more closely in contact. The de- 
scending process of the lacrymal is broader, and it is perforated 
by a large foramen. In Jacamerops and Galbula, on the 
other hand, the descending process of the lacrymal is very 

The ectethmoids are large and the interorbital and intra- 
narial septa complete. 

There are fourteen cervical vertebra ; the sternum has 
two pairs of incisions. 



Definition. Oil gland tufted. Muscle formula, AXY. Caeca absent. 

Skull desmognathous. 

The well-marked family Bucerotidae contains at least two 
distinct genera, Buceros and Bucorvus. The latter (the 
ground hornbill) is entirely African ; the former, which has 
been much subdivided, is both African and Asiatic. 

The great casque? not always equally developed, and the 
long bill, frequently serrated along its margins, and the 
largely black and white or black plumage distinguish these 
birds. But the small Toccus is a less typical form. The 
syndactyle foot, in which the second and fourth toes are 
united to the third the latter for several joints, the former 
for only one is highly distinctive, and is repeated in the 
ground-living Bucorvus. 

The oil gland is tufted. The feathers have no a/tersJiaft. 
There are ten rectrices. 

The pterylosis of Bucorvus abyssinicus has been described 
and figured by NITZSCH. 

The neck is completely feathered, except at its lower end, 
both dorsally and ventrally. The former is the commence- 
ment of the very narrow dorsal space of a long oval form, 
but not extensive. The pectoral tracks diverge at end of 
neck, but are subsequently undivided. 

The carotids are double in Bucorvus ; the left only is 
present in others. The remarkable obliteration of the 
carotids in the former genus, and their replacement by a pair 
superficial in position, have been described by GARROD 2 and 

The tensores patagii are in some ways characteristic of 
the Bucerotidae. In Buceros convexus (cf. FURBRINGER) 

1 OWEN, ' On the Anatomy of the Concave Hornbill, 1 P. Z. S. 1833, p. 102. 

- ' On a Peculiarity in the Carotid Arteries ... of the Ground Hornbill, 
P. Z. S. 1876, p. GO. 

" ' A Description of the Vessels of the Head and Neck in the Ground Horn- 
bill,' ibid. 1879, p. 461. 


there is no patagialis longus. The b re vis receives rather 
low down a very strong slip from the pectoral ; near to its 
insertion it gives off a wristward slip, which is attached to a 
special tendon arising from the lower end of the humerus. 
The main tendon passes over this, not attached to it, to the 
ulnar side. The absent longus is represented only by a 
thinnish tendon arising from the pectoralis. 

The same structures are found in B. malabaricus, B. 

coronatus, B. bicornis, and in 
Toccus. In B. atratus there 
is in addition an excessively 
small patagialis longus muscle, 
arising with brevis and con- 
sisting indeed of but very few 
fibres. In Bucorvus, on the 
other hand (fig. 100), the tensor 
patagii longus is well developed. 
Each tendon has a slip from 
pectoralis (a and a' in fig. 100), 
but that which joins brevis 
receives a tendinous slip from 
the biceps. This, however, as 
it is not figured by FUEBRINGER, 
is possibly individual. 

Quite exceptionally among 
anomalogonatous birds many 
hornbills have a broad humeral 
attachment of the anconceus. 
The muscle itself arises from 
the scapula by a Y-shaped 

(Bucorvus, Buceros) or flat, non-divided (Aceros) head. The 
humeral ' ankerung ' is found in B. subcylindricus, B. bicor- 
nis, not in B. tlatus, B. malabaricus, B. atratus, Bucorvus 
abyssinicus, Aceros, or Toccus. The deltoid has no scapular 

The leg formula of all hornbills is AXY . 
The glut ecus maximus is quite absent; the gluticus ex- 
ternus is only present as a ligament. 

Bucorvus (AFTER BEDDAKD). 

//, humerus ; Jii, biceps ; Bs, biceps slips (?): 
li.r, t.p, tensor patagii brevis tendon ; , a', 
slips from pectoralis. 


The biceps is occasionally (e.g. B. el at us) double at its 
origin, the tendons being separated by quite a quarter of an 

The arrangement of the semitendinosus and adductor in 
Aceros nipalensis, which is somewhat complex, will be 
understood from the accompanying drawing (fig. 101). 

The semitendinosus (St) is inserted on to the tibia by a 
long, thin, flat tendon ; another tendon, joining the first just 
where it passes into the muscle, is attached to the gastro- 

The accessory semitendinosus is in two parts : the larger 
half (Ast) is attached to the semitendinosus just behind the 
origin of the tendon of insertion of the latter ; the second 
half appears to arise from the tendon which connects the 
semitendinosus with the gastrocnemius, it passes up towards 
the thigh, and just in front of its (tendinous) insertion on to 
the femur it receives a tendon from the adductor. This 
latter muscle (the adductor longus) is inserted by three 
tendons (1) to the femur; (2) a small tendon which has 
already been described as joining the second half of the 
accessory tendinosus ; and (3) near to the origin of one of 
the internal heads of the gastrocnemius ; to this tendon is 
also attached the inner head of the gastrocnemius. 

The corresponding muscles l of Bucoruiis abyssinicus are 
rather simpler than in Aceros nipalensis. The adductor 
longus is only inserted at two places : first, by a fleshy inser- 
tion along a considerable length of the lower border of the 
femur ; second, by a tendon in common with the inner- 
most head of the gastrocnemius. The semitendinosus is 
attached by a thin tendon to the tibia, as in Aceros, and by a 
short tendon, also as in that species, to the gastrocnemius. 
The accessory semitendinosus arises chiefly from this latter 
tendon, but there is no division between this part of the 
muscle and that which takes its origin from the fleshy part 
of the semitendinosus. 

In Buceros atratus there is, again, some little difference 

1 GADOW figures most of these muscles in Bronn's TJiicrrcich, ' Aves,' J'xl. 
vi. Abth. iv. Taf. xxiii. b, fig. 1. 


from both the types already described, although the resem- 
blances are on the whole closer to Aceros. 

The adductor longus is attached by two tendinous heads ; 
the upper one of these, as inAceros, is attached to the lower 
border of the femur ; this corresponds to the fleshy insertion 
of the muscle in Bucorvus ; the lower tendon is fused 011 its 
way with the inner head of the gastrocnemius, which is 
continued upwards and reaches the femur, and then bifur- 
cates into two tendons of insertion. The relations of the 


add, adductor ; St, semiteiidiuosus ; As/, its accessory ; Sm, semimernbranosus : 

/us/, gastrocnemius. 

semitendinosus and of the accessory semitendinosus are as 
in Aceros nipalensis. 

In Toccus these muscles are much the same as in 

In Ceratogymna elata I find a closer resemblance to 
Aceros than to any other of the genera mentioned in this 
paper, but there is an agreement with Bucurviis in ih& fleshy 
insertion of the adductor longus on to the lower border of 
the femur. The accessory semitendinosus is distinctly double, 
as in Aceros, and is attached by a short tendon to the 


adductor, though the direction of this tendon is somewhat 
different from what is found in Aceros. 

The skull is doubly desmognathous. There are distinct 
basipterygoid processes, large in B. rhinoceros, almost va- 
nished in A.nipalensis, with which, however, the pterygoids 
do not articulate. The interorbital septum is widely fenes- 
trate. The fused lacrymals and ectethmoids together make 
a large plate of bone ; the postfrontal processes are large. 
The atlas is fused with the axis. The haemapophyses are 
slight. There are fourteen or fifteen cervical vertebrcc ; they 
are median in C12-D2. C14 and Dl have a median and 
two lateral processes. 1 The sternum, which is faintly two- 
or four-notched, has both spin a externa and interna. 

The liver lobes present some differences in different 

Commencing with Bucorvus abyssinicus, in which the 
right lobe is larger than the left, the series terminates with 
Buceros coronatus, in which the left lobe is larger than the 
right. The following table shows the relations of the liver 
lobes in such hornbills as have been examined : 

Bucorvus abyssinicus. R>L. 

Aceros nipnlensis. R > L. 

Bnceros bicornis. 
Sphagolobus atratits. 
Bycanistes subcylindricus. 
Buceros plicatus. 
Buceros rhinoceros. 
Buceros coronatus. 

R = L. 
R = L. 

I have noticed a peculiarity in several species of horn- 
bills which is not found in all other birds. In all birds the 
two lobes of the liver are completely separated from each 
other by the umbilical ligament, which bears the umbilical 
vein (this appeared to be particularly large and well deve- 
loped in all hornbills which have been dissected by me) ; and 
in addition one liver lobe the right- -is commonly separated 
from the abdomen by a thin membranous septum. In horn- 
bills both lobes of the liver are thus shut off ; I have figured 

1 In Dichoceros bicornis there is a tendency towards the formation of a 
hypapophysial canal. 



this condition in Bucorvus abyssinicus ; ' it is exactly the 
same in one or two other species which I have subsequently 

The Syrinx, Aceros nipalensis. The last rings of the 
trachea are fused together to form a solid box, at the sides of 
which, however, the individual rings are recognisable. In front 
the last three rings are thus fused, but behind two additional 
rings fuse with the others to form a wide and deep bony 
plate. The tracheal rings lying in front of these five show 
the dovetailing arrangement which is so often found in the 
tracheal rings. The pessulus is well developed and bony, 
but, owing to the complete fusion of the tracheal rings both 

posteriorly and anteriorly, 
it is impossible to say 
from which rings it is de- 

The intrinsic muscles 
of the syrinx are attached 
near to the boundary line 
between the last and the 
penultimate tracheal 


The bronchial semi- 
rings are cartilaginous, and 
there is a considerable in- 
terval between the first of 
these and the last tracheal 

FIG. 102. SYRINX OF Aceros nipalensis. 


Bucorvus abyssinicus. The syrinx of this hornbill (fig. 
103) differs in many particulars from the last. The tracheal 
rings are not ossified, and there is no box formed by their 
fusion. Only posteriorly are the penultimate ring and 
the two in front of this fused just at the origin of the 
pessulus ; anteriorly the pessulus is fused with the ante- 
penultimate tracheal ring, which forms with it a three-way 
piece ; the last two tracheal rings do not meet in front. The 

1 ' Notes on the Visceral Anatomy of Birds. I. On the so-called Omentum,' 
P. Z. S. 1885, p. 842 ; and above, p. 44, fig. 29. 



slender syringeal muscles are attached to the anterior margin 
of the last tracheal ring. 

The peculiar-shaped tracheal rings are hardly recognis- 
able until about the fourteenth from the end. 

Buceros rhinoceros has a syrinx which is not 
different from that of 
Acer os. The same rings 
are fused to form an os- 
sified box ; but the fusion 
between the several rings 
is hardly so extensive as 
in Acer os ; furthermore 
the syringeal muscles are 
attached to the posterior 
border of the last tracheal 


In Sphagolobus atra- 

tus there is very little 
fusion between any of 
the last tracheal rings ; 
the last three rings, which alone show any signs of ossification, 
are fused for a very short space anteriorly ; posteriorly there 
is no fusion at all, and the pessulus can be plainly seen to be 
connected with the antepenultimate ring. Although the last 
tracheal rings are not fused, they are very closely applied 
together, and no membranous interspaces are left. 

Ceratogymna elata, which is, like the last, a compara- 
tively small species, has a very similar syrinx ; indeed, I can 
find no differences sufficiently tangible to be described. 

Buceros lunatus and B. bicornis, which are both large 
species, hardly present any differences from B. rhinoceros. 

Bycanistes subcylindricus has a syrinx which, although 
of about the same size as that of Ceratogymna elata, shows 
certain differences which are worth putting on record. In 
the first place, the syrinx is much compressed from side to 
side at the level of the last tracheal ring; in the second 
place, the last tracheal ring is very much more arched than 
usual ; it forms, indeed, almost a complete semicircle. The 

FIG. 103. SYRINX OF Bitcorvus abyssini- 


intrinsic muscle of the syrinx in this, as in the other smaller 
hornbills, is very much larger relatively than in the larger 

Anthraceros malayanus, again, is a little different from 
all the types hitherto described. The last tracheal rings are 
but little fused posteriorly ; only the penultimate and ante- 
penultimate rings are so fused, so that it is impossible to be 
certain as to the origin of the pessulus. The intrinsic muscles 
are slender. 

Toccus presents certain peculiarities which I have not 
yet observed in any other hornbills ; the trachea has tioo 
pairs of extrinsic muscles, given off about half an inch apart. 
This condition seems to me to be so remarkable that I have 
preserved the specimen which shows it, though unfortunately 
the insertions of the anterior pair of muscles are lost, and I 
have no recollection of w r here the point of insertion was. 
The intrinsic muscles are relatively small. There appears 
to be no fusion between any of the tracheal rings. 

Cryptorms of the upper Eocene of France is held by 
MILNE-EDWARDS to be a hornbill. 

The family Upupidse ' contains only the well-known 
hoopoe (Up up a) and the but little known Irrisor and 
R liinoponias tus . 

There is a large feathered oil gland, but the aftershaft is 
absent or rudimentary. There are ten rectrices. 

The feather tracts are narrow. The ventral tract divides 
very early 011 the neck, and gives off on each side in the 
pectoral region an outer branch. At the base of the neck a 
triserial tract is given off to the humeral tract, and just 
below it a uniserial tract to the patagium. Between the 
outer and inner branches of the ventral tract is a single row 
of feathers. 

The dorsal tract encloses a spindle-shaped space, the 
pterylse enclosing which are somewhat dilated in the middle. 

1 STRICKLAND, 'On the Structure and Affinities of Upupa and Irrisor,' Ann. 
Mag. Nat. Hist. xii. (1843), p. 238 ; MUKIE, ' On the Upupiclae and their Bela- 
tionships,' Ibis, (3) iii. 1873, p. 181. 


The tendon of the tensor patagii longus ' gives off a 
wristward slip ; the main tendon crosses the fore arm. 
There is a cucuUaris patagialis, besides slips from the 
pectoralis, but no biceps slip. 

The anconfcus has an attachment to the humerus. 

In the hind limb the formula of the muscles is the 
typical picarian AXY . The passerine character shown by 
the existence of a well-marked cucullaris prppatagialis is 
paralleled in the hind limb by the absence of any vinculum 
between the deep flexor tendons. 

The tongue is short and the intestines are without c'c-a. 
The left carotid alone is present. 

There are fourteen cervical vertebra 1 . The sternum has 
a single pair of notches or fenestrae and both spinae. The 
skull is pseudo-holorhinal, desmognathous, without vomer 
and basipterygoid processes. 

The conjoined maxillo-palatines are rather delicate fenes- 
trated bones, and the bony palate for a little way in front 
is somewhat vacuolate. The palatines have long postero- 
external angles, which reach back to a point corresponding 
to rather beyond the middle of the pterygoids. 

The lacrymals are small and ankylosed to the skull. 
The ectethmoids are very large plates, and the distal end is 
segmented off, and is apparently the equivalent of the 
os uncinatum of many other birds ; it reaches the jugal. I 
describe the nostrils as pseudo-holorhinal, because, though 
rounded at their extremities, they are unusually long, and 
reach, or very nearly reach, the ends of the nasal processes. 
They are obliterated in the middle by bony alinasals. There 
is a largish median ioramen 2 a little way above the foramen 
magnum, and a minute one just above the latter. 

1 NITZSCH and GIEBEL, ' Zur Anatomic des Wiedehopfs,' Zeitschr. f. d. IJCN. 
Natnrw. x. p. 236. 

2 This was present in only one of three specimens, in which also alone the 
os uncinatum was present. It has a shorter bill and may be a different 



Definition. Beatrices, ten ; oil gland nude ; aftershaft present. Muscle 
formula of leg, A . Expansor secundarioruni, sterno-coracoideus, 
and biceps slip absent. Caeca absent. Manus very long. Sternum 

This group of birds contains two well-marked types 
the humming birds and the swifts, 1 the former confined to 
America, the latter world-wide in distribution. 

In external characters the generally minute size, the 
frequently brilliant metallic plumage, and the long slender 
bill distinguish the Colibris from the swifts. But Dr. 
SHUFELDT has found - in a nestling humming bird a bill 
hardly longer than that of a swift. 

The rectrices are ten, and in all these birds the oil 
gland is nude. There is an aftershaft. In the swifts there 
are down feathers upon the apteria ; in the humming birds 
there are not. 

The pterylosis of the group has been chiefly studied by 
NITZSCH, to whose account Dr. SHUFELDT has added details 
of value. 

The throat is completely feathered in the swifts, the two 
ventral tracts, however, becoming distinct at the beginning 
of the neck. The ventral tracts widen out in the pectoral 
region, but there is no outer branch or trace of one. The 
narrow dorsal tract bifurcates between the shoulders and 
reunites again to enclose a narrowish spinal space. There 
are well-marked femoral tracts. 

In the humming birds the ventral tract is double up to 
the symphysis of the mandibles, or nearly so ; the dorsal 
tracts are very much wider and form a diamond-shaped patch, 
within which is a very slight dorsal apterion ; there appear 
to be no femoral tracts, and there is a naked space in the 
nape of the neck, dividing the dorsal tract. 

1 GIEBEL, ' Ueber einige Eigenthiimlichkeiten in tier Organisation der Koli- 
bris,' Zcitscltr.f. d. ges. Naturw. 1. 1877, p. 322; W. K. PARKER, ' On the Sys- 
tematic Position of the Swifts,' Zool. (3), xiii. 1SS9, p. 91. 

- 'Studies of the Macrochires,' etc.. J. Linn. Soc. 1888. 


The swifts are among the very few birds which are 
partly quintocubital and partly aquincubital. 

Among the more obvious external characters are the 
magnificent metallic colours which are so usual a feature of 
humming birds. It is on account of the latter character 
mainly that they have been placed in the neighbourhood of 
or more properly confused with the Nectariniida?. It is,. 
however, practically the universal opinion that these two 
families have no near relationship, and the feathers of both 
have been lately submitted to a careful examination by 
Miss NEWBIG-G-IN. 1 The metallic colours of humming birds 
occur in both sexes, though more brilliant in the male ; they 
are mainly to be found on the throat and on the head as a 
crest. It has been pointed out that the rapidly vibrating 
wings would destroy all advantage (in sexual selection) of 
the development of these tints upon the wings. The colours 
are of every shade, and gold and red are often present, two 
colours which are not found among the Nectariniidae. The 
striking difference between the two families, however, consists 
in the fact that while the Nectariniidas have the ends of the 
barbs affected by the metallic colour it is the basal part of 
the barbs which is so coloured in the humming birds. 
Hence the barbs have cilia in the latter case and not in the 
former ; for this reason the rectrices in the humming birds 
can show metallic colours and yet not have their efficiency 
as flight feathers destroyed. The interlocking apparatus 
is there in the form of the cilia. The barbs which are 
thus metallically coloured are in both groups of birds 
(and in other birds which show the same kind of colora- 
tion) composed of a series of roof-shaped laminae, in the 
cavities of which the dark brown pigment essential for 
the due production of the metallic colour is located. 
Further details may be found in the interesting memoir 



The tensor es patagii show a striking resemblance among 

1 ' Observations on the Metallic Colours of the Trochilichi- and the Necta- 
riniida?,' P. Z. S. 1896, p. '2H3. 



the swifts and humming birds. In Chcetura, Cypselus, 1 and 
Phaethornis and other humming birds the tensor brevis is 
fleshy for almost its whole extent. In Dendroclielidon 
the tensor brevis has still a larger muscular portion than is 
usual, but the tendon is more evident and has a passerine 
slip to the humerus. In the Trochili, however, the muscle 
is inserted on to a special tendon upon the fore arm, 2 and 
not on to the extensor metacarpi. In the leg muscles 
the birds of this group agree in only possessing the femoro- 
caudal of those used by GAEEOD in his classification ; the 
formula, therefore, is A . GARROD, however, has left a 
note to the effect that in Ch&tura caudacuta the femoro- 
caudal passes through a muscle arising from both pubis 
and ischium, which is thus possibly a combined semitendi- 
nosus and semimembranosus. 

Though the semitendinosus is as a rule absent, there 
seem to be traces of its accessory in a few swifts. Thus 
in Cypselus alpinus and Chcetura Vauxi the gastrocnemius 
has an origin between the biceps loop and the main body 
of the sciatic nerve from the femur. 

Another peculiarity in the leg of certain swifts (cf. also 
Phaethou) is the absence of a biceps loop ; but the value of 
this character may be gauged from the following table :- 

Without Biceps Sling 

Chcetura caudacuta 
Panyptila melanoleuca 
Dendrochelidon coronata 
Macropteryx mystacea 

AVith Biceps Sliny 

Chcetura spinicauda 
Cluetura zonaris 
Cijpseloides fumic/a tit* 
Cypselus alpinus 

The biceps femoris of humming birds at any rate of 
Patagona gigas is peculiar in the fact of its being two- 

The deep flexor tendons in the swifts vary. In the 
majority of forms the two tendons completely blend ; in 

1 For Cypselus muscles see NITZSCH-GIEBEL, ' Zur Anatomie d. Mauer- 
schwalbe,' Zcitschr. f. d. ges. Nat. x. 1857, p. 327. 

2 This tendon looks like a degenerate representative of the abductor 


Mticropteryx, however, LUCAS ! has described the flexor 
hallucis as going to the first digit only after giving off a vin- 
culum to the tendon of digit IV. As to the humming birds, 
there has been some confusion. It appears, however, that 
the flexor hallucis before going to digit I. gives off a slip to 
flexor comrnunis of digit II., and (according to GADOW 2 ) 
III. and IV. also. 

Both swifts and humming birds have, as a rule, only one 
carotid, the left. The following swifts have two carotids : 
Cluf'tura nitila (right carotid larger) , Cypseloides fumigatus. 
In Micropus (7 Panyptila} melanoleuca SHUFELDT has 
described the left carotid (the only one present) as crossing 
over to the right and being until the middle of the neck free 
of the hypapophysial canal. 

The chief peculiarity of the vascular system concerns the 
femoral vein. In Panyptila melanoleuca and in Cluctura 
zonaris the femoral vein, instead of running deep of the 
fenioro-caudal muscle, comes to join the sciatic artery and 
nerve immediately it has passed the obturator externus 
superficial to the femoro-caudal tendon ; Cypseloides fumi- 
gatus is the only other swift which has been shown to be 
characterised by this structural abnormality. 

The large size of the heart of the humming birds as 
compared with that of the swifts is commented upon by 

The syrinx of the swifts is not in any way remarkable ; 
it is tracheo-bronchial, with the usual pair of intrinsic and 
extrinsic muscles. The former are attached (at any rate 
in Chcetura caudacuta] to the first bronchial semi-ring. In 
Cypseloides fumigatus, however, a swift which is in other 
ways abnormal, there appear to be no intrinsic muscles. 

The humming birds, on the contrary, have an unusual 
form of syrinx, which is remarkable in two ways. 

In the first place the trachea bifurcates very high up in 
the neck, recalling the characteristics of Platalea rosea (see 
below). Each bronchus in TrocJiiliis columns (according 

1 Ibis, is'.lo, p. 298. - Ibi<L p. -J'.i'.i. 



to MAcGiLLiYEAT ! ) has as many as thirty-four rings, which 
are complete and not semi-rings. There seem to be two 
pairs of extrinsic muscles, which form a very prominent 
muscular mass, as in Passeres. Dr. SHUFELDT was unable 
to find any sterno-trachealis. 

The tongue in the swifts is short and sagittate, with 
a spiny base. It is constantly bifid at the tip. 

In the humming birds, as is well known, the long tongue is 
tubular, and for its support the hyoids are bent over the top of 
the, skull, as in the woodpeckers. The tongue itself ' is double 
right down to the unpaired part of the os entoglossum, 
whilst each of the two distal prolongations of the entoglossal 
bone or cartilage is surrounded by a horny sheath, which is 
curled upwards and inwards, in a similar fashion to what 
we have seen in the Nectariniidae. In many species the 
outer and inner edges of these tubes, however, are entire 
and not laciniated. Thus the Trochilidae have developed 
the highest form of tubular tongue ' (GrADOW 2 ) . 

The gizzard of the humming birds is remarkably small ; 
that of the cypselids presents no remarkable characters, 
and SHUFELDT has remarked upon the large size of the 
liver in the humming birds as compared with the swifts ; 
in both the right lobe is larger than the left, and there is a 
'gall bladder in the swifts. 

Caeca are entirely absent in the Macrochires. 3 The 
following are intestinal measurements of the swifts :- 

Cypselus apus ..... 6-25 inches. 
Dendroclielidon coronata . . . 4'30 ,, 
Cluftura caudacuta . . . .10 ,, 
Cypselus alpiniis . . . .10 ,, 

Vait.i-i ..... 3'25 ,, 

A careful account of the trochiline and cypseline skeleton 
will be found in a memoir by SHUFELDT. 4 Though this 

1 In AUDUBON'S Birds of N. America. 

- ' On the Suctorial Apparatus of the Tenuirostres,' P. Z. S. 1883. 
:i CRISP, ' On some Points relating to the Anatomy of the Humming Bird 
(Trochihts colubris),' P. Z. S. 1862, p. 208, observed a ' rudimentary appendix.' 
1 ' Contribution to the Comparative Osteology of the Trochilida?, Caprimul- 




observer is disinclined to allow a very near affinity between 
the birds, it is undeniable that there are resemblances. 

The skull is schizognathous in the humming bird, 
aegithognathous in the swifts. But the aegithognathism in 
the latter is a little abnormal. GARROD has pointed out in 
describing ' the osteology of Indicator that that bird, in 
common with the CapitonidEe, has a truncated vomer, in 
which the truncation occurs behind the line joining the 
maxillo-palatines, while . in the true 
Passeres the truncation is in front of 
this line. The swifts are intermediate, 
the truncation being, as is shown in 
the accompanying figure (fig. 104), 
about on a level with the line joining 
the maxillo-palatines. It is true that 
the lateral processes so characteristic 
of the aagithognathous skull are better 
developed in the swifts than in the 
swallows ; but, on the other hand, it 
must be borne in mind that the un- 
doubtedly a3githognathous Indicator 
is without these processes. In both 

swifts and humming birds the skull is 7 , i( . s ieian<>iciicii*. UNDER 
holorhinal and without basipterygoid VlEW - < AFTER SHUFELDT.) 

/'in.r, preiuaxiila ; M.rp. maxillo- 

DrOCeSSeS. AS tO the VOmer, palatines: Vo, vomer: .V, 

nasal ; /V, palatine ; /Y, ptery- 

HUXLEY described it as truncated ; .goi.i. 

but SHUFELDT finds it to end in an 

excessively fine point. In swifts the vomer is, as already 

stated, truncated. But as to this difference and its value as 

a means of separating the birds cf. the manifold vomer of 


The humming birds have fourteen or fifteen (Trochilus 
Alexandra) cervical vertebra. The Cypselidse have thirteen 
or fourteen. Four ribs 2 join the sternum on each side 

gidas, and Cypselidse,' P. Z. S. 1885, p. 880, and 1886, p. 501. See also ZEHNTNEK, 
' Beitriige z. Entwicklung von Cypsdus mclba,' Arch. f. Naturg. Ivi. 1S90, p. 189 
(transl. in Ibis, 1890, p. 196). 

1 Loc. cit. (on p. 196.) - Fri;i;i;i\<iEH says five or six. 

FIG. 104. SKULL OF Micro- 



in both groups of birds. The sternum in both is unnotched 
and broader behind than in front. 

In the fore limb the length of the hand distinguishes 

HUMEBUS OF Micropiis melanoleu- 


both the families of the Macrochires, whence, of course, the 
name. The nearest approach in length of hand is shown 
in the swallows, petrels, and, oddly enough, in the penguins. 

HUMERUS OF Trochilus Alexandri 

p.f, pneumatic fossa. 


The humerus in both families is extremely short ; the radial 
crest is well developed in both into a long process which 
curves over the shaft in the Trochilidae, but over the head in 
the swifts. 



Definition. Anisodaciyle. Oil gland nude. 1 Beatrices, ten. Aquinto- 
cubital. Aftershafb present. Skull holorhinal. Both carotids 
present. Cseca - large. Ambiens and accessory femorocaudal 
absent. Deep flexor tendons of type V. 

This group of birds shows a considerable amount of 
structural variation, which allows of the separation of the 
genera into at least two families ; they are, however, all 
united by the characters in the above definition. The 

FIG. 109. LEFT FEET OF AntrostonniH FIG. 110. EIGHT FOOT OF Podar. 

i-ociferus (RIGHT-HAND FIG.) AND gas Cuvieri (AFTER SCLATER). 
Nyctidromus albicollis (AFTEE 

external aspect too of these birds, with the widely gaping 
mouth and their generally softly tinted grey and brown 
plumage, enables them to be readily distinguished from other 

1 Sometimes said to be absent in Podargicla;, but FURRRINGER found it in 
- Absent in 


In the typical Caprimulgidse (fig. 109) the claw of the 
middle toe is serrated and the fourth toe has but four pha- 
langes. There is no serration and five phalanges in others. 1 
The aftershaft is present ; in the aberrant Steatornis it is 


not absent (as GAEEOD asserted). Podargns is remarkable 
for the possession of powder-down patches, of which there are 
two, one on either side of the rump (see fig. Ill), first dis- 
covered by Mr. SCLATER. 2 The patches of this bird are well 
defined and very compact, and have not the diffuse character 
that is seen in, for example, RhinocJietns. Pow T der downs 
.are also found in Batrachostomus and Nyctibius. The 

1 SCLATER, 'Notes upon the American Caprimulgidse,' P. Z. S. 1806, p. 1'Jd. 

2 ' Additional Notes on the Caprimulgidae,' ibid. p. 581. 



pterylosis has been elaborately described by GAEEOD for 
Steatornis ; l so we shall select that bird, though it is, as 

vtfi&fft 4 ^ 

" '''V:.::. 

















1 For notes on Steatornis pee, in addition to papers quoted, N. FUNCK, 
'Notice sur le Stcatnrnis rarijwisis,' Bull. Ac. Belg. 1844, p. 371 ; F. STOLZ- 
MANN, 'Observations snr le Steatornis Peruvien,' Bull. Hoc. Zool. France, 
v. 1880, p. 198 ; HUMHOLDT, ' Sur le Sfentornis,' Bull. Soc. Pliilom. 1817, p. 51 ; 


already hinted, in many ways an aberrant form as a type. 
The dorsal tract (see fig. 112) gradually narrows as it 
passes down the neck, but the feathers get stronger ; it 
bifurcates between the scapulae to form a well-defined 
fork, which ends ultimately, having become weaker. Between 
this fork, and not connected with the rest of the dorsal tract, 
appears a spear-headed patch of feathers. The ' shaft ' 
of the ' spear ' becomes stronger as it descends to end 
abruptly at the base of the oil gland. The ventral tract 
is narrow between the mandibles ; it is undivided upon the 
neck. At the beginning of the breast it divides into a wide 
outer and a narrow inner portion, the latter being more 
strongly feathered. The two converge on each side towards 
the cloacal aperture, but do not reunite. 

In Capriwmlgus, on the other hand, the ventral tract bifur- 
cates in the neck, and the two ventral tracts are single, and 
there is no such abrupt break between the two parts of the 
dorsal tract as has been described above in Steatornis. Nor is 
there in Antrostomus, where there is no narrowing of the pos- 
terior part of the dorsal tract. Nyctidromus is much the same. 1 

The tongue in the goatsuckers is more or less abortive ; in 
Podargus it is a curious tough but transparent membranous 

As mentioned in the definition of the group the cceca are 
large ; but as a unique exception ^Egotheles seems to be 
entirely without them. In all the genera the left lobe of the 
liver is rather the smaller ; and all, save Chordeiles, have a 
gall bladder. The intestinal measurements in inches are 
given on the following page. 

The intestine (according to MITCHELL) is primitive and 
owl-like, while the ctcca are dilated apically, as in owls. 

J. MURIE, ' Fragmentary Notes on the Guacharo, or Oil Bird,' Ibis (3), iii. p. 81 ; 
L'HEKMINIEB, 'Memoire sur Je Guacharo,' Nouv. Ann. Mns. iii. 1834, p. 321, 
and note ' Sur la Classification Methoclique clu Guacharo,' &c., Rev. Mag. Zool. 
(2), i. p. 321 ; J. MULLEK, ' Ueber die Anatoniie des Steatornis caripensis,' M.B. 
k. Ahad. Wiss. Berlin, 1841, p. 172, and ' Anatomische Bemerkungen iiber den 
Guacharo,' Arch. f. Auat. it. Phys. 1842, p. 1. 

1 For further details of feathering see CLARK, ' The Pterylography of certain 
American Goatsuckers and Owls,' Proc. U. S. Nat. Mns. xvii. 551. 


Nyctidromus albicollis , 

Caprimnlgiis curopcpus 

) n 

Chordcilcs tcxcnsis 
Podargns Ciivicri . 
Steatornis caripensis 







i iseca 

I-:?, 1-4 


1-1, <) 




The syrinx is highly characteristic in the Caprimulgi. 
Like the nearly related (?) cuckoos, we have both the 
tracheo-bronchial and the purely bronchial syrinx. In- 
deed, the stages are almost identical in the two groups. 
Cuculus and Caprimulgus correspond with a tracheo-bronchial 
syrinx ; then we have Ceutropus and Podargns, and finally 
the culmination in Crotophaga and 
Steatornis of a syrinx furnished w T ith 
a membrana tympaniformis, which 
does not commence until many rings 
below the bifurcation of the tube, 
the intrinsic muscles being attached 
to the first ring which borders upon 
it. It will be necessary to describe 
the various syringes in some detail ; 
they have been studied and figured 
by myself. l In N/jctidromus albicol- 
lis, which will serve as a type of the 
tracheo-bronchial syrinx which ex- FIG. 11:3. SYBINX OF Nycti- 
ists in the Caprimulgidffi (.,), the gSSi )) . flZWcoBM (A 
last four tracheal rings are closely 

applied in contradistinction to the preceding, which are 
separated by copious membranous intervals. The last two 
tracheal and the first five bronchial semi-rings are ossified. 
To the first of the latter are attached the intrinsic muscles. 
In Batrachostomus we have the intermediate type of syrinx, 
which may, however, be called bronchial. The first six 
bronchial semi-rings and the last three tracheal are ossified, 

1 ' On the Syrinx and other Points in the Anatomy of the Caprimulgidffi,' 
P. Z. S. 1886, p. 147. 


and bear much resemblance to each other, which will be 
apparent from the illustration (fig. 115). 

The intrinsic muscles are attached to the seventh bron- 
chial ring, which is soft and cartilaginous ; where the bronchial 
rings change their character is a constriction of the mem- 
brana tympaniformis ; it is, however, of equal breadth before 
and after the change. In Podargus Cuvieri there is a further 
approach to the purely bronchial syrinx of Steatoniis. The 

FIG. 114. SYRINX OF JEgotlic!<'x FIG. 115. SYRINX OF Batrachostomus 


first two bronchial rings are complete. These and the four- 
teen following are closely applied to each other and ossified. 
The intrinsic muscles are attached to the last of this series. 
Mgotheles really belongs to this section of the Caprimulgi, 
though the intrinsic muscles are attached very high up upon 
the bronchi ; but the two rings immediately preceding the 
attachment are complete rings. The final development of 
the bronchial syrinx is seen in Steatoniis (see fig. 48, p. 69), 
where all the rings in front of the attachment of the intrinsic 
muscles low down upon the bronchi are closed and complete 
rings, as in Crotophaga. 

The tensor patagii shows certain differences among the 
goatsuckers. In the genera Caprimulgus, Nyctidromns, and 
Chordeiles there is a biceps slip, absent in the rest. Of 
these three genera the arrangement of the tendon is show r n 



in the annexed cut (fig. 110). Steatornis, as will be seen 
(fig. 117), hardly differs, and Podargus agrees with it. In 
Mgotlieles there is a slight difference in that there is hardly 
any trace of the wristward branch of the tendon. Steatornis 
has an expansor secundariorum, apparently absent among 
the other genera. The muscle is attached to the teres by its 
long tendon. The insertion of the deltoid is extensive, and 
it receives a tendon from the scapula. In many Caprimul- 
gidae the biceps is split for some distance before its insertion,, 
the bifidity even invading the muscle itself and not being 

Caprimulcjus (AFTEK GAEKOD). 

rf, de'.toid ; b, biceps ; h, Immerus ; tpb, ten- 
sor patagii brevis ; ecr, extensor carpi radialis. 

OF Steatornis, BUT OF LEFT WING 

/, triceps. Other letters as in fig. 11G. 

limited to its tendon. In Podargus FURBRINGER describes 
a special slip of rhomboideus profundus, arising separately 
from ilium. The anconceus has a tendinous humeral head. 
In the thigh the muscle formula is either AXY (most Capri - 
mulgi) or XY (Steatornis). The tibialis antic its tendon of 
Podargus is double. The glutaeus I. extends over biceps in 
Nyctidromus, &c., not in JEgothelcs. No glutaeus V. There 
are sometimes both and sometimes only one of the two 
peroneals present. In Steatornis and JEgothdes only the 
brevis is to be found, in Nyctidromus only the longus ; in 
Podargus both. 




have by no means a uniform skull 


In Caprimulgus l the skull may be termed schizognathous. 
The palatines are enormously expanded, and between their 
posterior extremities (not indicated in the figure) are a small 
anterior and posterior medio-palatine, a state of affairs recall- 
ing the Picidse (q.v.) The vomer is a long bone, distinctly 
paired in the young bird, rounded in front. It articulates 

FIG. 118. SKULL OF Caprimuhjus (AFTER HUXLEY). 
I'm.r, premaxilla ; Mxp, maxillo-palatiue ; Vo, vomer ; PI, palatine ; Pf, pterygoid. 

with the hook-like maxillo-palatines. Each of the latter is 
connected by a ligament with the internal forward process of 
the palatine of its own side, the hinder part of which is 
largely ossified. The basipterygoid processes are well de- 
veloped. The lacrymal is large and ' binds upon the zygoma.' 
The ectethmoids are attached to the broad outer flange of 
the palatines by a cartilaginous prolongation. 

1 PARKER, ' On the Structure and Development of the Bird's Skull,' Linn. 
Tr. (2), i. 



In Clwrdeiles ' the skull is much upon the same plan, 
but the maxillo-palatines meet in the middle line, and may 
even become ankylosed. 

The skull is thus desmognathous, in fact. In C. virgi- 
nianns, however, the bones do not meet. 

The skull of Nyctibius .jamaicensis (see fig. 119), de- 
scribed by HUXLEY, is not widely different from that of Capri- 
mulgus. The ligaments which unite the inner angle of the 

Nyctibius jamaicensis (AFTER 

*, preirontal process. Other letters as in fig. 1 18. 

FIG. 120. SKULL OF Steatoniis 

palatine to the maxillo-palatines are, however, completely 

In Podargus ' 2 the skull is completely doubly desmo- 
gnathous. The basipterygoid processes are quite rudimentary. 
There are two small azygous vomers. The palatines have 
coalesced in the middle line. The lacrymal is small, if not 

The skull of Steatofnis has been described, with figures, 

1 SHUFELDT, ' On the Osteology of the Trochilidae,' Ac., P. Z. S. 1885, 
p. 891. 

- PARKER, loc. cit. p. 124 (with figs, on PL xxiii.) 



by GARROD,' and more recently and more fully (also with 
illustration) by PARKER.* The skull is quite owl-like in 
general aspect with its decurved beak, and in the equality 
between greatest length and greatest breadth. The lacryinal 
is very small and is ankylosed to the orbital wall. The 

AS IN FK;. 118. 

maxillo-palatines are completely fused across the middle 
line, the skull being desmognathous. The palatines also in 
their middle part meet across the middle line. The vomer 
is divided into two parts, one lying behind the other. The 
anterior part is small, like the posterior part ; they both 
measure 5.5 mm. The posterior vomer is probably the 
equivalent of the medio-palatine of Caprimulgus and other 
birds. The basipterygoid processes are well developed. The 

1 ' On some Points in the Anatomy of Stcatornis,' P. Z. S. 1873, p. 526. 
- ' On the Osteology of Steatornis caripensis? ibid. 1889, p. 161. 



ectethmoids are continued on to the quadrato-jugal bar by a 
distinct ossified ox nin-inatum, as in Todus, Scythrops, 
Musophagida?, Piaya cay ana, Cariama, and Procellariidae. 
The relations between the procoracoid and the clavicle vary 
somewhat. In Podargus the process is large and reaches 

OF Caprimiilr/us 

Fra. 123. STERNUM OF 
Podargus (AFTER 

OF Nyctibius (AFTER 

the clavicle ; it is small and does not in Steatornis and 

The number of cervical vertebra varies from thirteen 
(Cliordeiles) to fifteen (Steatornis). PARKER has commented 
upon the fact that in Steatornis the atlas, instead of being, 
as is the rule among birds presumably allied to it, perforated 
below for the reception of the odontoid process, is merely 
deeply notched for the same. In Steatornis the dorsal 
vertebra3 are opisthoccelous, as among the parrots alone 
among probable allies. It is the rule among the Caprimulgi 
for four ribs to reach the sternum. The sternum itself is 
one-notched on each side in Steatornis, &c. It has four 
foramina in zEgotlieles, and is doubly notched on each side 



in Podargus. The spina externa is developed and slightly 
bifurcate in Steatornis. There is no spina interna. 

In view of the considerable variation in structure ex- 
hibited by the group, the following tabular statement may 
be of use. 










t 2 * 






'S c/5 




5 ". 

& 1 '| 











Podargus . 



Xnt over 




2 notches 

+ Broueh. 


. Ki/nthelts 










1 notch 
2 fur 

.Vi/ctibius . 





2 notches 

+ V 







Over biceps 



) T~ 

1 notch 















1' 51 










V + 


There can be little doubt, from a consideration of the 
above table, of the naturalness of a family Caprimulgidse to 
include the last four genera. In these forms, in all of them, 
the toes are aberrant in that the last has only four phalanges, 
and further that the claw of the middle one is serrated. 

It will be noticed from the table that the amount of 
structural variation among the CaprirnulgidEe (s.s.) is exceed- 
ingly small, the only character, indeed, of those selected 
which varies being the gall bladder, which is absent in 
Chordeiles, and shows signs of commencing disappearance 
by its small size in some of the others. We might, perhaps, 
add the desmognathism of Chordeiles ; but this is obviously 
but a slight exaggeration of the aegithognathous palate of 
the others. The enormous length of the second primary of 
Maorodipteryx and Cosmetornis is a variation which does not 
appear to be of great importance from a classificatory point 
of view. 

The remaining genera are by no means so uniform a 
group as that which those that have been already considered 
form. In all of them, however (so far as is known), the 
biceps slip is absent, the glutaeus primus is of limited extent, 

1 See BLYTH in Ibis, 186(3, p. 357. 


the skull is desmognathons, the syrinx is bronchial (least 
marked in JEgotlieles}, and the outer toe has five phalanges, 
while the middle toe has no serration. This is a fair 
assemblage of identical characters. It is usual for Steatornis 
to be separated as a distinct family from the Podargidee, as 
has been done by GADOW. It differs from Podarf/us in 
seven out of the thirteen characters made use of in the 
above table. It is often supposed that the Guacharo is 
peculiar among goatsuckers by reason of its vegetable diet ; 
but NEWTON states in his ' Dictionary of Birds ' that the 
Podargidae also partly nourish themselves on fruit. More- 
over JEgotlielcs, which is by universal consent placed in the 
immediate neighbourhood of Podargus, differs from that 
genus in four out of the thirteen characters, and from Stea- 
tornis in exactly the same number. It appears possible to 
place all these genera in one family, which, on account of its 
greater antiquity, has had time to vary more than the Capri - 
mulgida?. It is also among members of this family that the 
greatest number of points of affinity to the owls is met with, 
a further argument in favour of their basal position. 

The relationship of the Caprimulgi to other groups is a 
puzzle hard of solution. This is partly, perhaps, due to the 
fact that the goatsuckers are probably a somew T hat ancient 
group. That they are an ancient group seems to be shown 
by the considerable amount of specialisation of structure 
which they exhibit, by the primitive character of the 
intestinal folds the ca3ca being at the same time well 
developed by the double carotids, and by their wide distri- 
bution, with a restriction in range of some peculiar types, 
such as Steatornis. 

The remarkable series of modifications of the syrinx is 
one of the most striking facts in the anatomy of the group. 
They share this with the Cuculi, and, though to a less extent, 
with the Striges. It is, indeed, with this latter group that 
the goatsuckers seem to be most nearly allied. This con- 
clusion, which is in harmony with much recent opinion, is 
curious in view of the external likenesses ' which bind 

1 For instance, the 'ears' of certain Podargidffi and of Lyiicornis. It is a 

it 2 


together the two groups of birds, likenesses which might 
fairly be put down to similarity of habit. These superficial 
resemblances are, however, enforced by more deep-lying 
structural similarities. MITCHELL has found that of the 
various groups which may be supposed reasonably to be 
allied to the Caprimulgi the owls come nearest to them in 
the primitive character of the gut, while the caeca, swollen 
at the ends, are alike in both. The owls too are nearly the 
only other Coraciiform birds besides the Caprimulgi which 
have well-developed basipterygoid processes. The trogons 
it is true, possess them, but then they differ in many other 
important particulars. 


Definition. Oil gland nude. 1 Aquincubital. Both, carotids present. 
Caeca "well developed, ending in a dilatation. Skull desmognn- 
thous and holorhinal, with basipterygoid processes. No ambient, 
semitendinosus, accessory fenioro-caudal, biceps slip, or expansor 

The owls, formerly associated with the Accipitres ant" 1 
termed 'Accipitres nocturnse,' or ' Nyctharp ages,' are now 
generally placed by themselves away from the haw r ks and 
in the neighbourhood of some of the birds comprised under 
the term ' picarian.' The group itself is characterised by 
a great uniformity of structure, and by the possession, so to 
speak, of so many negative characters. The resemblances 
to the hawks are really only in habits and in beak and claw. 
These, however, will be dealt with more fully later. 

The owls comprise a considerable number of genera, of 
which Strix stands rather apart from the rest, having a^ 
near neighbours the Eastern Pliotodilus 2 and the Madagas- 
car Helivdilu** 

curious coincidence that the term ' morepork ' is applied to Podargus in Australia, 
and in New Zealand to an owl, Spilocjlaux Novte-Zclandice (fide NEWTON, Diet, 
of Birds, sub voce ' Morepork '). 

1 Except Strix and Asio otus. 

- BEDDARD, ' On Photodilm badiits,' Ibis, 1890, p. 293. 

3 MiLNE-Ei>WAiu>s, ' Observations sur les Affinites Zoologiques,' etc., Nonv. 
Arch. MUK. (2), i. 1878, and in Hist. Nat. de Madagascar. See also B. B. SHARPE, 
' A Note on Heliodilus,' P. Z. S. 1879, p. 175. 


It is often given as a character of the owls (and as a, bond 
of union with Pa ml ion) that there is no aftn-xluift. There 
is, however, a small one in Strix. As a rule the oil gland 
is nude, but NITZSCH gives as a constant character of Hybris 
( = Strix) the presence of two minute feathers upon the apex 
of that gland ; in Asia otus too there are two or three small 
down feathers upon the apex of the oil gland all of which 
facts seem to indicate a comparatively recent loss of the 
apical tuft so often found in birds. 

The rectrices are invariably twelve in number, except in 
Micropallas, where there are but ten. 1 A singular external 
character of the owls, carefully gone into by KAUP,~ concerns 
the asymmetry of the ears ; with this is sometimes corre- 
lated an asymmetry of the skull in the region of the ear. :! 

The tensores patagii of the owls 4 are on the whole 
simple. In none of them is there a biceps slip. Very rarely 
is there a recurrent tendon uniting the insertion of the 
brevis with the longus (patagial fan). This occurs, how- 
ever, in Strix Novte Hollandice (? always), S. flammea 
(occasionally), and S. prat in cola (? always). Something of 
the same kind is found in S/jrni/tin alttco, where, however, 
the connection is between the tendon of the longus, 
just at its origin, and the inner of the two branches of the 
lircris. In most owls (Bubo macnlosus, Scops leucotis, 
Pulsatrix torquata, Athene noctua, Strix, Sijrnium nebu- 
losum) the tensor brevis sends off a wristward slip not far 
from insertion of main tendon ; the latter alone crosses the 
fore arm to be inserted on to ulna. 

In Scops Lempiji, S. Asio, Asio otus, Kctupa javanensis, 
Otus vulgaris, Bubo ascalaplms, and Nyctea nivca, there is 
an additional posterior brevis tendon, arising separately from 
the muscle. This latter tendon sometimes in Otus vulgar is 
is connected with the middle one. 

' Fide CLARK, ' On the Pterylography of certain American Goatsuckers and 
Owls,' Proc. U. S. Nat. Mus. xvii. p. 551. 

- A monograph of the Strigidse, Zool. Trans, iv. 

:! Cf. COLLETT, ' On the Asymmetry of the Skull in Strix Tenyinalmi,' 
P. Z. S. 1870, p. 739. 

4 D'ALTON, DC Striyuui Muscidis Comvicntalio, Halis, 1837 ; HEUSINGER ? 
Arch.f. 1'hys. vii. 122. 



In Asia of/is there is a fleshy slip from pectoral to 
longus tendon, and a fibrous slip from humerns to patagial 
muscle just at origin of innermost brevis tendon. In Scops 
Lempiji there are muscular slips to both longus and brevis 
from the pectoral, and a fibrous slip from the humerus to 
the longus tendon just at its origin. 

The humeral origin of the anconceus is always present. 

There is never an expansor secundariorum, so far as 
is known. 

The pectoralis primus is not at all double. 

The only one of the leg muscles used by GARBOD in 
classification that is present is the fcmoro-caudal, the 
formula thus being A . Glutaeus I. is absent, glutseus V. 
small. In Ketupa javanensis, however, V. is absent and I. not 
quite so. Only one peroneal muscle is present, which is 
attached to head of metacarpal. The tendon of insertion of 
the tibialis antictis is divided ; in Puhatrix I found it to 
be trifid at insertion, arid even the muscle itself was divided 
into two for a short distance. 

The deep flexor tendons are of type I. In Ketupa 
ceylonensis the two tendons blend a quarter of the way down 
the tarso-metatarsus ; though blended the fibres can be 
recognised, and it may be seen that those of flexor hallucis 
mainly supply digits I., II., a small part only to rest of com- 
mon tendon. 

Asio otus 
Syrnium aluco 

,, nebulosum 
Bubo ignavus 

virginianus . 

,, capensis 
Ketupa javanensis 
,, ceylonensis 
Athene noctua 

,, passerina . 
Pulsatrix torquata 
Speotyto cunicularia 
Nyctea nivea 
Surnia funerea 
Gyrnnoglaux nudipes 

Small Int 

Large Int. 



, , Ins. 
























3-5, 3-_M 






















The colic caeca are always (fig. 125) dilated at the blind 
end. The liver lobes are subequal, and the gall bladder 
appears to be always present. 1 

The skull of the owls 2 shows some differences in Strix 
from the characters which distinguish the majority of the 

In Strix the skull is elongate, the proportions being in 
Strix flammea 56 length : 36 
breadth. In another species 
(Strix sp. inc.) 62 : 37-5. 

On the other hand in other 
owls the differences between 
breadth and length show gra- 
dually progressive series, cul- 
minating in Speotyto cunicu- 
laria, in which the proportions 
are nearly equal, viz. 38 : 37. 
The skull of Strix further 
differs from that of other owls 
in the swollen character of 
the prefrontal processes, 
which are thin, almost paper- 
like sheets of bone in other 
owls. The interorbital septum 
of Strix is thick, while in the 
remaining genera it is reduced 
to a thin dividing lamina, as is 
the case with most birds. The 

sknll characters of the o-pmis FlG ' 125 '~ CoLIC C ' ECA OF Plt todiltl * 


Photodilus are to some extent 

intermediate between Strix and the remaining genera of the 
Striges. The interorbital septum of Photodilus is not 
so thick as in Strix, but, 011 the other hand, not so thin as in 
other owls, as, for example, Bubo. The prefrontals, although 
not so swollen as in Strix, are not nearly so flattened as they 

1 Absent in Spcotyto; cf. SHUFELDT, 'Notes on the Anatomy of Sweaty toC 
Jonrn. Morpli. iii. 18S'.), p. 122. 

- See for certain details of skull structure PARKER, Linn. Trans. (2), i. p. 138. 



are in Bubo, where, as already explained, they are thin 
plates, hardly thicker than a piece of paper. So far Photo- 
dilus agrees with Strix ; but there are points in which the 
skull of this aberrant owl is nearer to the bubonine section 
of the order. It has not the occipital convexities which are 
so striking a feature of the skull of Strix. Finally Pliotodiliis 
is strigine in the non-extension over the occcipital region of 
the temporal fossse, which do so extend in many of the 
Bubonidae. In Strix there is but one notch on either side 


E, prefrontal process ; W, maxillo-palatines. 

of the sternum ; in other owls, including Photodilus, there 
are two. The vomer of the owls is not large, 1 and behind it 
there is a medio-palatine, at least occasionally present. 

The lacrymals, like the maxillo-palatines, with which 
they come into contact, are swollen and spongy. The 
nostrils are often partly covered by ossified alinasals, and 
there is a largely bony internasal septum. 

1 Said by SHUFELDT to be absent in Speotyto. 


iM'. > 

In the foot of Strix (see fig. 127) the first phalanx of digit 
III. is much less than the second in length ; in most other 
owls (fig. 128) these two phalanges are small and subequal ; 
Plwtodilus is intermediate. The latter genus has the pecu- 
liarity that the last digit has only four phalanges instead 
of five, the two basal ones being fused. 

The number of cervical vertebrae does not appear to vary. 


I have found fourteen in Strix, Photodilus, Ketupa, and 
other genera which I have examined. In Bubo bcngalensis 
the ring of the atlas is incomplete above ; the significance of 
the occurrence of this same deficiency in Pandiou- is largely 
lost, owing to the fact that Cariama is similarly characterised. 
The haemapophyses in Ketupa javanensis commence as 
single processes on Cll, and extend to D3 ; on CIO is a 


bifid hypapophysis, and on C9 the catapophyses nearly form 
a canal. Posterior catapophyses begin on C12, whence they 
gradually climb the hypapophyses. Bubo bengalensis is 
much the same, save that there is not a bifid haemapophysis 


on 10. In Strix the catapophyses on C9 are distant. Photo- 
dihix is like Ketupa. 

To the sternum, which is one-notched in Strix and two- 
notched in' other owls, five ribs are attached. Dr. COUES ' 

Key to N. American -Bm/.v. 



has mentioned the existence among the Striges of a ' scapula 
accessorial The coracoids are not in contact at their articu- 
lation with sternum ; the procoracoids are moderately large, 
and the clavicle reaches both them and the scapula. 

If it were not for Photudilns, it might be possible to 
divide the Striges into two families, Strigidse and Bubonidse. 
As it is, it may perhaps be permissible to regard the order as 
containing but one family, but two sub-families, viz. Striginse 

FIG. 129. SYRINX OF Scot's leucoti* 

FIG. 130. SYRINX OF Bubo 

and Bubonina?, to which possibly a third, Photodilinae, might 
be added. 

The syrinx of the Striges has been chiefly described by 
WUNDEELICH ! and by myself. 2 This group is one of the 
few that present the remarkable variety of the voice organ 
which has been termed the bronchial syrinx. All the owls, 
so far as they have been examined, possess one pair of 
intrinsic muscles and the usual one pair of extrinsic muscles. 
Scops leucotis has the most modified syrinx. In this bird 
(see fig. 129) the intrinsic muscles are attached so far down 
the bronchus as to the tenth bronchial ring, and, as will be 

1 ' Beitriige zur vergleichenden Anatomic und Entwicklungsgeschichte des 
unteren Kehlkopfes der Vogel,' Nov. Act. Lcop. Akad. xlviii. 1884, p. 1. 

2 ' On the Classification of the Stviges,' Ibis (o), vi. p. 


seen from the figure, the bronchial rings in front of this 
attachment are complete rings, with no membranous inter- 
space left. In >S7rw, on the other hand, and in IJiibo and 
Syniiiu/i, the intrinsic muscles are inserted on to the first 
bronchial semi-ring. In Asio seven complete rings intervene 
between the bifurcation of the trachea and the first incom- 
plete bronchial semi-ring, to which the muscles are attached. 
In Photodilus the intrinsic muscles are inserted on to the 
second bronchial semi-ring. 

Until lately the owls have been almost invariably placed 
in the immediate neighbourhood of the diurnal birds of prey. 
Latterly, however, the opinion has been gaining ground that 
it is to the picarian birds (in a wide sense) that they are most 
nearly allied. This opinion, more than hinted at by GARROD 
and NEWTON, has been given a practical shape in the classi- 
fications of FURBRINGER and GADOW. The latter has in- 
geniously pointed out that it is impossible to imagine that 
the Striges have been derived from the Accipitres, since, 
although without an arnbiens, they have much the same 
structure of foot as the Accipitres with an ambiens. Hence 
it is difficult to believe that they would have lost it ; he con- 
cludes that they are derived from some bird without an 
ambiens, and the failure of MITCHELL to find the last trace 
of the missing ambiens obvious in some birds which are 
clearly the descendants of birds with an ambiens still 
further supports that way of looking at the matter. Even 
in the skull, where the principal likenesses between the 
Accipitres diurrife and nocturnae (as the two groups in 
question have been called) have been seen, there are really 
many differences. It is only, for example, in the skulls of 
those Accipitres to which the owls have been supposed to 
have the least resemblance, i.e. the Cathartidas and Serpen- 
tariidse, that there are basipterygoid processes. The owls 
are decidedly not desmognathous (in the sense of a maxillo- 
palatiiie union), and their lacrymal is quite different from 
that of the hawks and eagles. The palate, too, is incom- 
plete in front of the maxillo-palatines, not solid, as in the 
Accipitres. As to other anatomical features, it is harder to 


find likenesses than differences. The Accipitres have rudi- 
mentary caeca, a biceps slip, the expansor secundariorum, a 
tufted oil gland, an aftershaft (except Pandiun) ; the deep 
flexor tendons are different, and, in short, the differences are 
as great as those which separate any two groups of carinate 
birds. 1 


Definition. Twelve rectrices ; - aftershaft present; aquintocubi- 
tal ; zygodactyle. Skull desmognathous, holorhinal, "with- 
out basipterygoid processes. Biceps slip and expansor se- 
cundariorum absent/' Muscle formula, AXY + or . No 
caeca ; a crop present. 

The parrots are an almost cosmopolitan group, being 
most abundant, however, in the tropics. Count SALVADORI, 
in his British Museum catalogue of the group, allows five 
hundred species, distributed among seventy-nine genera. 
The parrots are a very sharply defined group, there being 
no dubious outlying forms. They are usually brilliantly 
coloured, and lay white eggs in hollows of trees. "With the 
exception of the owl parrakeet (Stringops) of New Zealand 
the parrots are arboreal birds, as, indeed, the zygodactyle feet 
denote. As to external characters, the exaggeratedly hawk- 
like bill is well known. The almost universal twelve 
rectrices distinguish the group, but in other external and 
internal characters the parrots show considerable diversity 
of structure, as is sometimes the case with large and widely 
distributed groups ; compare, for instance, the pigeons, which 
present many other analogies to the parrots. 

The oil gland 4 is a structure which may be wanting or 
developed. The table on p. 268 indicates some of the genera 

1 See also under ' Caprimulgi,' p. 243. 

- With the sole exception (cf. GADOW) of Oreopsittacus Arfttld. 

:! See below, p. 2(il. 

4 External characters and many other points in the anatomy of parrots are 
dealt with by GAKUOD, ' On some Points in the Anatomy of the Parrots,' Ac., 
P. Z. iS'. 1874, p. 247, and 'Notes on the Anatomy of certain Parrots,' ib'nL 
187(>, p. ()'.)! ; see also FORBES, ' On the Systematic Position of the Genus 
Lathaimts,' ibid. 1879, p. 160. 


111 which it is present or absent. When present it is in- 
variably tufted, and generally of fair size. In Cacatua sul- 
l>1tiirca, however, the oil gland, though present, is small, and 
has but a single small down feather upon it. 

The parrots are a group of birds which agree with the 
Accipitres in that some genera have powder-down patches 
while others have not. The table already referred to indi- 
cates the facts so far as they have been ascertained. The 
degree of development of the powder-downs, however, differs 
considerably, though in no parrot is there more than a single 
pair of definite powder-downs which are lumbar in position. 
In Cacatua sulplmrea, for instance, there are a pair of such 
patches, one on either side of the dorsal tract. These send 
up a few scattered powder-down feathers as far forward as 
the neck, and a few to carinal spaces and between the 
branches of the ventral tract. 

In Calopsitta Novte Hollan,dicc there are the same lumbar 
patches of a reniform outline ; but the powder-downs are 
entirely confined to this region of the body. 

Calyptorhynchus stellatiis is more like Cacatua, but the 
lumbar patch is not so well developed. 

In Psittacula passerina there are lumbar patches more 
elongated but narrower than those of the parrots already 
referred to ; there are also scattered powder-downs not 
aggregated into patches. 

Brotogerys tirica has no defined patches, but simply a 
few scattered powder-downs, which, however, are more nume- 
rous in the lumbar regions. In Coracopsis, Chrysotis, and 
Pionits there are the same generally diffused powder-down 
feathers not aggregated into definite tracts. The same may 
be said generally of Psittacus, though such powder-downs as 
there are are limited to the lumbar region. 

The general pterylosis of the parrots is as follows : From 
the general covering of the head arises a narrow dorsal tract, 
which bifurcates in the interscapular region. Between the 
arms of this fork are the arms of another fork, which unite 
near the oil gland to form the single straight, short posterior 
part of the dorsal tract. On the ventral surface the tract is 


single, or double on the neck, and where it widens out 011 
either side of the carina sterni a strongly feathered lateral 
branch is given off. 

The variations shown in the pterylosis are not great. 
They concern the more or less definite separation of the 
lateral ventral tract and the slighter or more pronounced fea- 
thering of the anterior end of the posterior dorsal Y. Thus 
in Lathanuts and Platycercus the lateral ventral tracts are 
well marked, and the posterior fork of the dorsal tract does 
not diminish much in width where it comes into contact 
with the anterior fork. In Trickoglossus the exact reverse 
of these conditions obtains, the lateral ventral tracts being but 
obscurely delimited from the main tract, and the dorsal tracts 
of the hinder part of the body almost ceasing before their 
junction with the anterior half. 

PsepJiotus, Cyanorhamphus, Pyrrhulopsis,Agapornis, &c., 
agree with Platycercus. In Ara the outer branch of the 
pectoral tract is not definitely separable, but the dorsal tracts 
are more like those of Platycercus. Conurus is much the 

The syrinx of the Psittaci ! seems to show two main 

a. In the following species there is a syrinx of the type 
which will be described immediately : 

Cacatua cristata. 

,, triton. 

Microglossa aterrima. 
Calyptorliynclms Banks i. 
String ops habroptilus. 

The syrinx is in these species remarkable for the fact that 
the first semi-rings of the bronchi are weak and cartilaginous, 
and are usually separated from each other by considerable 
tracts of membrane. Cacatua itself represents the most ex- 

1 The syrinx has been chiefly studied by GIEBEL, ' Zur Anatomic der 
Papageien,' Zcitschr. f. d. ges. Wiss. xix. p. 133, and by PARSONS and myself, 
' On certain Points in the Anatomy of Parrots,' Ac., P. Z. S. 1893, p. 507. 


treme type ; in Cacatua cristata, for example, when the syrinx 
is seen on a lateral view the membrane occupies a great deal 
of the outer lateral region of the commencement of the 

The semi-rings of the bronchus are at first very small, and 
do not extend across the side of the bronchus ; they gradu- 
ally increase in length, until at the sixth or seventh they 
come to extend right across the syrinx. In Microglossa 
aterrima the syrinx is in certain respects less abnormal ; 
the rings are still feeble, but on a lateral view of the organ 
they extend completely across, and there is on such a 
view no bare tract of membrane, such as we have figured 
in Cacatua. 

Calyptorhynchus Banksi is intermediate between the 
two extremes ; the first semi-ring only is incomplete, inas- 
much as it does not reach from one side of the syrinx to 
the other or rather we should say from the anterior to the 
posterior side. 

String ops habroptilns has the same weak cartilaginous 
bronchial semi-rings ; but on a lateral view of the syrinx they 
are seen to extend right across. 

b. The second group contains the following genera :- 

Chrysotis. Tanygnathus. 

Pyrrhulopsis. Eos. 

Triehoglossus. Poly teles. 

Lorius. Platycercus. 

Pionus. Pceocephalus. 

These genera' are differentiated from those of the first 
division by the fact that the bronchial semi-rings are as a rule 
ossified, and are frequently more or less fused together ; at 
the same time the first ring is commonly concave upwards, 
whereas in the parrots of the first-mentioned group the 
bronchial semi-rings are straight. 

The most extreme type is perhaps offered by Chrysotis; 
of this genus I have seen the following species : 


Chrysotis versicolor. Chrysotis Bodin. 

,, erythrura. ,, viridigenalis. 

,, leucocephala. ,, Levaillanti. 

In all these species the first two rings of the bronchus 
are closely fused together, and form a bowed piece of bone 
forming with the last tracheal ring a semicircular outline ; the 
space between the two is, of course, occupied by membrane. 
In Chrysotis Levaillanti, for instance and there is no great 
difference in the other species the double character of the 
apparently single first bronchial semi-ring is only to be seen 
at the two ends. In a number of other parrots the first 
bronchial semi-ring is larger than that which follows, though 
not fused with it ; this is the case with Trichoglossus, 
Pyrrhulopsis, and Chalcopsitta ; the genera Eos, Poly teles,. 
Platycercus, and Tanygnathus have syringes which are con- 
structed on the same plan. In Conurus there is a little 
difference ; here the first two rings of the bronchus are equi- 
sized ; this at any rate applies to the two species Conurus 
aureus and Conurus cruentatus, which are the only two that 
we have examined from this point of view. The genus Ara 
(species Ara Leari, Ara militaris) agrees with Conurus. 
Psittacus is like these genera ; but Pionus agrees more closely 
with Chrysotis. 

It will be obvious that no hard and fast line can really 
be drawn between the two groups of parrots ; if it were 
thought desirable to draw such a line, it would be between 
the genus Cacatua on the one hand and all the remaining 
parrots on the other. Cacatua alone has a syrinx in which 
the first bronchial semi-rings are incomplete, leaving a bare 
tract laterally which is easily visible when the syrinx is 
viewed from the side : but in this genus there is another 
peculiarity the intrinsic muscle of the syrinx ends in a very 
narrow point, which passes into a fine tendon of attachment ; 
in Chrysotis, Eos, &c., the muscle is comparatively broad 
down to its actual attachment. In this particular Micro- 
glossa and String ops agree with Cacatua, although they do 
not show the incomplete rings that have been mentioned as 



characteristic of the latter genus. These genera, in fact, are 
to this extent intermediate between Cacatua and the more 
normal (at any rate more usual) form of syrinx in the 
parrots ; the rings are still, however, soft and cartilaginous, 
thus different from Conurus, which is a further step in the 
direction of Chrysotis ; Clirysotis seems to represent the 
opposite extreme to Cacatua. Ara is a genus which is 
also intermediate in the characters of its syrinx ; it has 
weakish and straight rings, as in Stringops, for instance ; 
but the muscles are as in the second group of parrots, and 
the general aspect of the syrinx is more in accord with this 
placing of it. 

Finally it should be added that occasionally (e.g. Polyteles 
melanurus) the extrinsic muscles are attached not to the 
sternum, but to the membrane covering the lungs, being con- 
tinued there by thin tendons. In Platycercus Barnardi 
there would seem to be no extrinsic muscles at all. 

Parrots are very much alike in their myology ; there are, 
however, a few points in which they show differences, and 
which may be useful for the purposes of classification. 

The tensores patagii of the parrots are like those of 
many homalogonatous birds in the broad aponeurotic 
character of the tendon of the tensor brevis, which, however, 
has two or three thickened bands in it corresponding to the 
discrete tendons of most other birds (e.g. Charadriidae) . Of 
these thickened bands the anterior commonly gives off a 
wristward slip ; but there appears to be never any patagial 
fan. The aponeurosis is inserted, as usual, on to the tendon 
of the extensor metacarpi radialis, and is continued over it 
by two tendinous slips, of which the posterior runs obliquely 
to the elbow joint, like the ' passerine slip ' of many birds. 
The common tensor patagii muscle is usually very large, and 
often completely covers the posterior deltoid (d. major}. 
GAREOD dissected away the anterior thickened tendon of the 
brevis in Deroptyus accipitrinus, and found it to arise from 
a distinctly separate slip of the patagial muscle attached to 
clavicle. The tendon in question is inserted on to the lower 
external humeral process, and may represent with its muscle 

I'SITTACI :>/i!) 

a primitive passerine tensor patuyii brevis, to which has 
been subsequently added an extension of the deltoid. 

With the general structure that has been described the 
parrots show much difference in the details of the patagial 

The arrangement of the tendons of the tensor patagii is 
very much the same in Nestor, Sir ing ops, and Calijpto- 
rlnjnclms ; in all three the tendons are relatively very long- 
when compared with the fleshy part of the muscle, and they 
are all close to one another, so as to give the appearance in 
Stringops of one tendon. In Calyptorhijnchus the anterior 
tendon leaves the others in the lower part of the patagium 
and runs forwards after its usual fashion, so that the main 
distinctive point of these three genera, as far as the tensor 
patagii goes, is that the middle and posterior tendons are 
close together. In Cora cops is these tendons are separated 
by a slight interval, but closely correspond to the arrange- 
ment in the birds last named. 

Eos, Loriiis, Pwoceplialus, and Gaica have a character- 
istic and almost uniform arrangement of the patagial tendons. 
In them the three tendons are very difficult to distinguish, 
because the fibrous membrane between them, of which they 
are only specialised parts, is as thick as they are. The result 
is that in these birds the patagial muscle seems to be inserted 
by a broad, short, membranous-looking tendon. 

Conurus shows a transitional stage between these last 
genera and the typical arrangement ; the three tendons are 
more distinct, and they are equally short and show the same 
mode of attachment to the fleshy part of the muscle. 

Latliaiiiiis is remarkable for having the anterior tendon 
separate in its whole length from the middle one, instead of 
being fused with it in the upper part of its course. 

Chrysotis and BolborlujncJi//* have a small extra tendon 
between the middle and posterior ones; in C. Gnildiinji this 
was only present on one side, but in C. leucocepliala it was 
found on both. 

Psittacus has three tendons w r hich are completely separate 
in the whole of their course, and in this respect it corre- 

s 2 


spends to Lathamus. The anterior tendon may represent the 
fused anterior and middle tendons of Latlicuuus, and the 
middle tendon may be an extra one, as in Chrysotis. Our 
reasons for this are that there is a considerable interval 
between the two tendons, and that they do not diverge, as 
in all other cases. If this view is correct, the patagial 
tendons of Psittacus closely resemble those of Chnjsotis, 
while they also agree in having the anterior deltoid larger 
than the posterior, in the absence of a lower head to the 
anconaeus, and in having the deltoid completely covered by 
the tensor patagii. 

The two deltoids are but small muscles, and are largely 
covered by the relatively enormous tensor patagii. It is 
better to use the terms ' anterior ' and ' posterior ' for the del- 
toids, since their relative dimensions vary considerably. The 
major is by no means always the larger. Sometimes the 
two deltoids are entirely covered by the tensor patagii, some- 
times the posterior is partly exposed. Thus in Nestor the 
muscle is exposed, in Dero2)tyus and Chrysotis it is covered. 
In Nestor and Stringoj)s the anterior deltoid is the smaller, 
in Caica it is the larger. In Tanygnatkus, Bulborhynchus, 
and E 'elect us the deltoids are narrow and equisized. In Eos 
cardinalis the anterior deltoid (which is the larger) is di- 
visible into two distinct parts. 

As regards the relative sizes of the two deltoid muscles, 
where they differ, such genera as are known may be arranged 
as follows : 

A. Delt. Larger Post. Belt. larger 

Deroptijus, Psittacula, Aprosmic- Nestor, Stringops,Calyptorhynchus t 

tus, Lorius, Caica, Eos, Pyrrlmlopsis, , Cacatua. 
Latliamns, Palaornis, Loriculus, 
Psepliotiix, Poroccphaliis, Cyano- 
rhamphus, Psittacus, Melopsittacm. 

In some parrots the anconaits longux has an accessory 
head from the humerus, which is especially broad in Stringops 
The table on p. 268 shows the distribution of this accessory 
head among the genera. 

The expansor secitndarionim is stated by G-ARROD to be 


absent from the Psittaci. This is certainly almost univer- 
sally the case. But FURBRINGER speaks of a rudiment a 
short length of tendon in Platycercus palliceps. 

No parrot has a biceps slip. A muscular cuculhiris 
patagialis is generally, if not always, present. 

It is well known that the ambiens muscle is present in 
some parrots, and absent from the leg of others. The actual 
occurrences of this muscle are shown in table (p. 268). Strin- 
fjops is peculiar in that the muscle is sometimes complete 
and quite normally developed, and sometimes ends in a thin 
tendon on the capsule of the knee joint. This recalls 

Of the other muscles of the leg used by GARROD in classi- 
fication A, X, and Y are nearly always present, the only 
exception, so far as I am aware, being Chrysotis Guildingi, in 
a specimen of which I failed to find Y. 

Sometimes (as in Ara chloroptera) the semitendinosus 
gives off a tendinous slip to the gastrocnemius, but in 
Chrysotis there is no such slip. The tibialis anticus is 
usually inserted by a single tendon. This Mr. PARSONS and 
I found to be the case in the majority of parrots which we 
examined. But in Chrysotis the tendon is distinctly double. 
In Deroptyus, Caica, Pceocephalus, Platycercus, and a few 
others, there are more or less evident indications of a double 

The deep flexor tendons of the parrots are gallinaceous, 
with a vinculum such as is illustrated in fig. 54 (p. 100). 
There are some inconsiderable variations of this ground plan ; 
for instance, in Platycercus Barnardi the vinculum is divided 
into two parts, one to digit II., the other to III. and IV. 

Peroneals. The peroneus longus and brevis are, as far as 
we have observed, always present in parrots, but the origin 
of the former differs somewhat in different genera. 

In String ops and Nestor the peroneus longus rises from 
the front of the bony fibula and its membranous continuation 
for about the upper half of the leg. The muscular belly 
overlaps that of the peroneus brevis very much near its 
origin, and the muscle is large and well marked. 


In CJirysotift, on the other hand, the peroneus longus is 
very small and only rises from the membranous continua- 
tion of the fibula in the lower part of the leg ; it is so small 
that it does not overlap the peroneus brevis at all, but lies 
behind it. 

The parrots have a well-developed crop and a zonary 
prove ntriculus. In the liver the right lobe is the larger ; 
rarely are they subequal. The gall Madder is as a rule 
absent ; but it is present in Cacattia and in Calopsitta, though 
' small and easily overlooked ' in the latter. 

The intestinal measurements in a series are as follows :- 


Stringops . . .73 

Eclectus polyclilonts . 93 

Banksi . . .61 

Ara ambigua . . 62 

,, ararauna . . 50 

Cacatna sulf/trra . 33'75 

,, cristata . 37 

,, triton . . 51 

I'ionus saiilis . . 5O5 

,, Maximiliani 49 

Eos rctlculata . . 25'5 

,, iinlica . . 26 

Platycercm Barnardi 29-5 

pallidiceps 20 

Nestor nicridiotialis 38 

Microglossa aterritna 34 

Pyrrhulopsis splendcns 43 

Chrysotis collar la . 45 

,, festira . 36 

Geopsittacus occlden- 

tali* 15 


Tanygnatlms Midler i 57 
Psittacus eritliacus . 48 
Conurus Petzii . .12 
Palceornis Alexandra 30 
Aprosmictus erythro- 

pterus . . .41 

Lorius lori . . 33 

Loricuhis clirysonotus 12 

,, galgulus . 1(5 

Chalcopsitta scintillata 37 

Psittinus malaccensis 12 

Euphema pulclicUa . 12 

,, splendida . 12 

Deroptyus accipitrinus 31 

Pionopsitta pileata . 39 

Latliamus discolor . 18 

Coracopsis Barkleyi . 33 

Dasyptilus Pecqueti . 17-25 

Brotogerys tirica . 31 

tori 21 

The most obvious comment upon the above list is to 
draw attention to the very great length of the gut in E elect u ft, 


whose relations in this particular to other parrots are almost 
those of Didunculus to other pigeons. 1 

The most recent and elaborate essay upon the osteology 
of the parrots is by MiVART, 2 who has described the entire 
skeleton of Lorius and Psittacus ; some of his illustrations 
are reproduced here. Fourteen is the prevalent number of 
cervical vertebra (e.g. Lorius, Psittacus, Platycercus, Caica). 
Strinr/ops, however, has fifteen. The atlas is notched (Ara 
in.ilitaris) or perforated (Pyrrhulopsis) for the odontoid 
process. Five (Ara, Psittacus) or six (Platycercus, Pyrrhu- 
lopsis) ribs articulate with the sternum. The sternum has 
as a rule an entire posterior margin, which in Licmetis is 
entirely unnotched and unfenestrated. Most parrots have a 
pair of fenestrse which in Deroptyus and Microglossa are 
converted into notches. The sternum has a spina externa, 
slightly forked occasionally (e.g. Psittacus erithacus, Callo- 
ceplialon,),\)\\i no spina interna. The carina is deep deeper 
in Platycercus (without furcula) than in Caica (withfurcula), 
the species being approximately of the same size. The 
furcula is sometimes present and sometimes rudimentary. 

It is present, and forms a complete U, in Nestor, Conurus, 
Caica, Licmetis, Microglossa, Ara, Palceornis, CYC. 

An intermediate condition is observable in Eos, where the 
furcula thins much towards its sternal end. A still further 
reduction is seen where the two clavicles are separate below 
and only bound by cartilage. Finally there are those parrots 
with a quite rudimentary pair of clavicles, consisting only of 
a small piece of bone at the coraco-scapular end. This is the 
case, for example, with Pyrrhulopsis and Platycercus. 

The following table shows the number of cervical verte- 
brae and the position of the first and last haemophyses in a 
number of parrots : 

1 For structure of tongue see CIACCIO, ' Nota preventiva sull' interna 
struttura della lingua del Papagalli.' liendic. Sess. Ace. 1st. Bologna, 1877-8, 
p. 157. 

2 ' The skeleton of Lorius flacn^t Hiatus compared with that of Psittacus 
erithacus,' P. Z. S. 1895. There is no account of the bones of the limits 



No. of 
C. V. 

First last 
Hnem. Hsem. 

( 'atap. 

Pyrrhulopsis persona ta 1 4 
Conurns hcemorrhoiis . 14 

09 D3 
C9 D2 


Platycercus Pcnnanti . 14 
Ara militari* . . 14 

C9 Dl 
C8 D3 


Eclectus poiychloros . 14 

Nestor notabilis . . 14 

C8 Dl 

CS * D2 

C11-D1 ' 

Caica mclanocepliala . 13 
Callocephalon galeatiun 14 
Calopsitta Nova Hollandicr. 14 
Conurus cruentatus . . 14 

C8 D3 
C9 Dl 
C9 Dl 

C8 D2 


The liumerus of parrots is peculiar, and, as GAKKOD 3 has 
pointed out, there are features of resemblance to the Columbse 
and to the Alcidae. This peculiarity will be found described 
and figured in the chapter dealing with the Columbse. The 
skull is very uniform in its structure throughout the group. 
It is desmognathous, holorhinal, and without basipterygoid 

The front part of the face (nasals, maxillae, and premaxillae) 
articulates by a transverse joint with the frontals, which is 
movable. The mobility of the anterior part of the face is 
aided by the movable articulation to it of the palatines and 
the jugals. The palatines have a peculiar form ; for the 
most part they are laterally flattened plates of great depth 
and considerable extent. The quadrate of parrots too is 
peculiar in the great length of the neck, which bears the 
squamosal articulation. In many parrots the lacrymal bone 
joins the forward process of the squamosal, thus completely 
encircling the orbit with bone. 

The hijoid has been extensively studied by MIVAET ; 4 in 

1 On Dl the median part of the hsemapophysis has vanished, leaving only 
the lateral. 

2 On this vertebra is a double ruemapophysis, forming a canal. 

3 See also for osteology of parrots BLANCHAKD, ' Des Caracteres Osteologiques 
chez les Oiseaux de la Famille des Psittacides,' Compt. Betid, xliii. p. 1097, 
and xlix. p. 518 ; MILNE-EDWABDS, ' Observations sur les Caracteres Osteolo- 
giques,' &c., Ann. Sci. Nat. (6), vi. p. 91 ; L. VON LOEENZ, ' Uber die Skeletc 
von Stringops liabropiilus u. Nestor notabilis,' S.B. k. Ak. Wien, Ixxxiv. 1882, 
p. 624. 

1 ' On the Hyoid Bone of certain Parrots,' P. Z. ,V. IWto, p. 1C.2. 


this paper references will be found to previous figures and 
descriptions. It has features which absolutely distinguish 
this group of birds. 

The entoglossal has a considerable median foramen, or 





I/, basihya! ; r, eutoglossum : p. jiuraliyal profess : it, uruhyal ; /i'o, liypobranebial ; 

<!>, ccratobranchial. 

more usually is composed of two separate pieces united in front 
by cartilage. The basihyal is broad, and it develops on either 
side a forwardly directed piece (figs. 131-3), for which Dr. 
MIVART has suggested the name of parahyal piece. This 


latter is merely a short process in Ara, Psittacus, and 
String ops (fig. 131) ; in Lorius, Eos, and Trichoglossus the two 
parahyals (figs. 13'2, 133) unite and form a single Y-shaped 

FIG. 132. HYOID or Lorius flavopalliatus (AFTER MIVAKT). 

e, cup-like excavation ; //. syiupln >fs "f ininihyals. Otliur letters as in fig. lol. 

bone inclined obliquely upwards. The only bird which 
seems to present much resemblance to the parrots is the 
eagle, which, according to the figure in BRONX'S ' Thierreich,' ' 

1 Arcs, PL xxxi. tic;. 23. 



has a broad basihyal with the short angular processes which 
suggest the more elaborate parahyals of the parrots. 

The classification of the parrots has been attempted by 
more than one naturalist ; but, as GADOW has justly ob- 
served, ' our knowledge of the anatomical structure of these 
birds is at present too incomplete in relation to their large 

GADOW himself has practically divided them by the 
structure of the tongue into two families, Trichoglossidae and 
Psittacidn?. In all the Trichoglossidae the orbital ring is 

FIG. 133. HYOID OF Loritts domicclla (AFTER MIVAIIT) 

AS IN FIG. 131. 


incomplete, and it is possible that the remarkable structure 
of the hyoid, described above, may serve to distinguish this 
family. The first family contains only Nestor, the lories, 
Cyclopsittacus, and Lathamus ; but the two latter are very 
iinperfectly known. The remaining genera are relegated to 
the second family. 

GARROD'S arrangement of the group, anterior to that of 
GADOW'S in point of time, is based upon the variations of 
the ambiens, oil gland, furcula, and carotids. The facts, 
with a few others added, are displayed in the following- 
table :- 





o s 

a s 




"s g 





a! s 











New Zealand 

Palceornis . 




Africa, India, China, 

East Indies 





Australia, Austro- 







E elect us 




Moluccas, Papua 





Philippines, Celebes, 

New Guinea 





Philippines to Celebes 

Eos .... 




Moluccas, Papua, Solo 

mon Islands 





Australia, Celebes, 

Papua. Timor, New 






Moluccas to Solomon 


Loriculus . 




India to Papua 

Coripliilus . 




Society and Marquesas 











Malay Peninsula to 


Agapornis . 



Ethiopian region 





Australia and Tas- 





South Australia 




South Australia 





i Australia to Philip- 








Calops'Ma . 







2 or 1. 






Papua, North Australia 






Australia, Austro- 

Malay, Philippines 











Tropical Africa 






Tropical Africa 







New Zealand 

Nasiterna ' . 



New Guinea 

Ara .... 






Mexico to South 




' + 



Mexico to South 







Mexico to South 


Caica .... 





Guiana and Amazons 





Costa Eica, South 


Deroptyus . 




Guiana, North-East 


' FORBES, ' On some Points in the Structure of Nasiterna,' &c., P. Z. S. 1880, p. 7G. 



r Wi 




fii .5 




-.2 '-3 

2 c g 

^ <2 ^ 




3-4-1 ^ 


B C 






+ Central and South 

LatJiaiiius . 




+ Australia and Tas- 


Coracopsis . 




+ Madagascar 

Pyrrhulopsis . . + 


2 + Fiji 

Dasyptilus . . . + 
CJiry satis . 


+ New Guinea 
+ Mexico to South 



2 + Mexico to South 


rtrotogeri/fi . 

2 + Central to South 


Platyccrcus . . + 

2 Australia, Tasmania, 
Norfolk Island 

Psepliotns . . . + 2 Australia 

Cyanorlunnplnts . . + 


New Zealand, New 
Caledonia, Society 


Psittacula . 




Mexico to South 


Nympliicus . 



+ New Caledonia to 
Loyalty Islands 

From these facts and a few others may be derived the 
following scheme : 

Sub-Order Psittaci : 

Fain. I. Palseornithidse. Two carotids. Ambiens ab- 
sent. Oil gland present. 

Subf. (1.) Palgeornithinse. Paleeornis, Eclectus, Apros- 
mictus, Eos, Tanygnathus, Prioniturus, Psittinus, 
Loricnlns, Trickoglossus, Lorius. 
Snbf. (2.) Cacatuinae. Orbital ring complete. 

Calopsitta, Calyptorhynchus, Licmetis, Eolo- 
pJms, Cacatua. 
Subf. (3.) Stringopina?. Furcula lost. 

Stringops, Euphema, Geopsittacus, Melopsitta- 
cus, Agapornis. 

Fam. II. Psittacidse. Left carotid superficial. 
Div. a. Ambiens present. 


Subf. (4.) Arinae. Ara, Coiiunis, Bolborhynchus, 

Caica, Psittacus, Pceocephaliis, Nestor. 
Div. b. Ambiens absent. 

Subf. (5.) Pyrrhurinse. PyrrJnira, Lathamas, Coni- 

copsis, Pyrrhulopsis. 
Subf. ((3.) Platycercinse. Furcula lost. 

Platycercus, Psephotus, Cyanorhamphus, Psit- 
t acid a. 

Subf. (7.) Chrysotinae. Oil gland lost. 
Chrysotis, Pionus, Brotogcrys. 

A phylogenetic tree accompanies GAREOD'S scheme, in 
which it is assumed that the ancestral parrot possessed the 
normally running carotids, an ambiens, an oil gland, and a 
complete furcula. From this the main stems are given off, 
in one of which the carotids remained normal, while in the 
other the left became superficial. The loss of the other 
characters leads to further branching of both main branches. 

The Stringopina?, especially Stringops itself, are the 
nearest living representatives of the ancestral stem. 

FURBRINGER ' also argues for the low position of Strin- 
gops, in contradiction to MARSHALL, who holds that it is an 
extremely modified form. Its owl-like plumage, defective 
carinal keel, and associated loss of the power of flight are 
undoubtedly modifications, but it seems more probable that 
they are modifications of an ancient than of a modern type 
of parrot. FURBBINGER'S views are chiefly based upon the 
flexibility of its anatomical characters. I have already 
referred to the variability of the ambiens : the sternum offers 
another fact of the same kind ; sometimes it is entire, without 
notches or foramina, sometimes there are one or two upon 
one or the other side, and occasionally two incisurge, one 
upon each side. This variability must not be associated, as 
variability may often be associated, with a rudimentary 
structure ; the xiphosternum is not rudimentary, though 
the keel is. 

1 ' Einige Bemerkungen Uher die Stellung von Sfrinr/ojm,' Ac., Jouni. f. O. 
1889, p. 230. 


Mr. PARSONS and I have pointed out certain likenesses 
between Stringops, the Cacatuinas of GARBOD, and Nestor, 
which he places in an altogether different family. These 
partly concern the syrinx, to which attention has been 
already directed, partly the muscular system. In those birds 
the posterior deltoid is larger than the anterior. It may be 
noted also that powder-down patches are best developed and 
more universal among the Cacatuinse, while it is in that 
family only that the gall bladder exists. 

Of extinct parrots among the most remarkable is Lopliopsitta- 
cus mauritianus, 1 characterised, as was also Nccropsittacus rodcri- 
canus, by its enormous jaws. 

The principal interest attaching to other remains of parrots is 
the light that they throw upon the former distribution of the 
group ; for Psittacus has been found in the lower Miocene of 

The determination of the affinities of the parrots to other 
groups of birds is one of the hardest problems in ornithology. 
They have been likened to the Accipitres (mainly, perhaps, 
on account of the hooked beak and its cere), and to the 
gallinaceous birds, in the neighbourhood of which they were 
placed by GARROD. It seems to me that the parrots, like 
the cuckoos, are a group of birds which are on the border- 
land between the Aiiomalogonatee and the higher birds. It is 
remarkable what a number of points there are in which they 
show resemblances to the Passeres the complicated muscu- 
lature of the syrinx, the absence of biceps slip and expansor 
secundariorum, the presence of a cucullaris propatagialis, 
found in the Passeres and in the somewhat passeriform 
Upnpa and Pici, the small number of cervical vertebrae, the 
total want of ca3ca, allying them not certainly to the Passeres 
but again to the Pici and many Anomalogonatee, the reduced 
clavicles of some genera. Zygodactyle feet, moreover, are 
not found among the higher birds except in the Cuculi and 

1 Sir E. NEWTON and H. GADOW, ' On Additional Bones of the Dodo and 
other Extinct Birds of Mauritius,' &c., Tr. Zool. Soc. xiii. p. 281. See also for 
a figure and account of this bird NEWTON'S Diet. Birds, sub voce 'Ex- 


the Musophagi, which are, similarly to the parrots, on the 
border line between the Anomalogonatae and higher birds. 
It is noteworthy also that of the Anomalogonatae which 
present a catapophysial canal (found at any rate in one 
parrot) it is the Pici and the passerine alone. But while it 
is not so difficult to point out likenesses to the Anomalo- 
gonatae it is much harder to indicate resemblances to any 
of the higher groups of birds. It must be held, in my opinion, 
that they have emerged from a low anomalogonatous stock . 
at a time not far removed from that at which the Cuculi and 
Musophagl also emerged, but that there is not a common 
starting point of the three groups. 


Definition. Feet zygodactyle by reversion of fourth, toe. Skull 
desmognathous, without basipterygoid processes. 1 Oil gland 
nude. Quintocubital. Two carotids. Caeca longish. Ambiens 

The family which is defined by the above characters is a 
large one, comprising, according to the recent catalogue of 
Captain SHELLFA", 165 species, which are distributed by 
that ornithologist into forty-two genera. 

The family is almost world- wide in range, being most 
abundant, however, in the tropics. Correlated with its 
numerous genera and species and wide range we find a 
certain amount of structural variation in the family, which 
permits of its division into several subfamilies, concerning 
the number and extent of which there is some divergence of 
opinion. It may be convenient, however, to consider the 
general anatomy of the family before dealing with its major 

Apart from the facts used in the definition of the family 
the cuckoos are characterised by the absence of, or the 
presence of only a rudimentary, aftersliaft to the contour 
feathers. The number of rectrices is not, as was stated by 
NITZSCH, constantly ten ; for in Saurothera, Guira, and Croto- 

' Well-marked rudiments in Rhinococcyx and Endynamis. 



pliaga there are only eight. The feather tracts are somewhat 
diverse in their disposition. The pterylosis of the European 
cuckoo (Cuculus canurus) has been described by NITZSCH in 
his ' Pterylography.' The feathering on the throat com- 
pletely occupies the intermandibular space. The ventral 
pteryla is divided upon the neck into its two halves, which 
are not again divided ; each passes backwards, gradually 

Fin. 184. PTERYLOSIS OF Eudynamis orii'iitalis. VENTRAL VIEW. 


diminishing in extent, until it ends in a single row 7 of feathers 
in the neighbourhood of the cloaca. Over the sternum this 
pectoral tract is very wide ; later its three rows of feathers 
become separated by a slight interval, two on one side and 
one on the other, which, however, reunite before ending at 
the cloaca. The spinal tract is narrow in the neck region. 
It bifurcates on the shoulder to enclose a lanceolate space. 




Of other cuckoos whose pterylosis has been studied } Caco- 
mantis, Play a, Saurothera, Diplopterus, Coccuzus, Chryso- 
coccyx, and Coccystes agree in most points with Cue til its. 
But in the American genera Piay a, Diplopterus, Saurotlicrn, 
and CWr//,:^.s-, the ventral tract is double from the very first 
that is to say, in the mandibular region. 

A more complicated pterylosis characterises certain other 



cuckoos. Iii the genus Centropus the feathering upon the 
throat is close and continuous, the tw r o ventral tracts diverg- 
ing at about the junction of the neck and trunk. Each of these, 

1 BEDDARD, ' On the Structural Characters and Classification of the Cuckoos,' 
P. Z. 6'. 1885, p. HIS; SHUFELDT, ' Contributions to the Anatomy of Geococcyx,' 
ibid. 1886, p. 960. 



again, divides into two separate tracts, of which the inner 
is at first two feathers wide, which number is reduced to one 
just before the termination of the row a little way in front of 
the cloacal aperture. The outer branch consists of one row 
only, and terminates some way in front of the end of the inner 
branch, without, however, showing any signs of being fused 
with it. Pyrrhocentor, Gcococcyx, Crotophdga, Endi/)iaiix, 

FIG. ISl'i. PTEKYLOSIS oi' Play a cat/ana. VKNTKAL VIKW. 

Scytlirops, a,n.d. k&v&a pterylosis which is much 
like that of Ccntrupu-s. The two types of pterylosis observ- 
able in the cuckoos are illustrated in the accompanying wood- 
cuts. Unfortunately the pterylosis of some important genera, 
such as Cuua, is not known. 

All cuckoos possess the (unbii'/is. In all cuckoos w r e also 
find the semitendinosus, the accessory semitendinosus, and 

T 12 


thefemorocaudal; in Centropu-s and its allies, both in the 
Old World and in the New, the accessory femorocaudal is 
also present. In many groups of birds the arrangement of 
the tendons ending in the patagium is very complicated. 
This is not the case with the Cuculi, where the disposition of 
these tendons is very uniform. For the most part the sim- 
plicity is suggestive of the perhaps allied picarian birds. In 
Cuculus canorus, for example, which has been figured by 
GARROD, the tensor patagii brevis is inserted on to the fore 
arm without any bifurcation. So too in Piaya, Saurothera. 
The only exception to this which has been noted occurs in 
Geococcyx, where the said tendon bifurcates just before its 
insertion, the anterior branch being inserted on to the exten- 
sor metacarpi radialis a little way in front of the main 
attachment. This is also the case with Guira and Phcenico- 

No cuckoo has any biceps slip. In some genera, e.g. in 
Saurothera, Coccyzus, Pyrrhocentor, there is an attachment 
between the anconaeus and the humerus. In Guira there 
is none. The expansor secundariorum is what GARROD 
(see p. 85) has termed ' ciconiiform. The glutceus primus 
is extensive in most cuckoos ; l its origin in them reaches 
behind the head of the humerus as well as in front. There 
is no glutaeus V. The deep plantar tendons of Pyrrhocentor 
and Centropus are peculiar in that no branch is sent to 
the hallux. 

The syrinx in the cuckoos shows greater variability than 
in any other group of birds excepting the goatsuckers. We 
meet with the typical tracheo-bronchial syrinx in a consider- 
able number of genera, while in others is the much-modified 
bronchial syrinx. The latter was first described many years 
ago in Crotophaga by JOHANNES MULLER. In that genus 
the syrinx closely resembles that of Steatornis, which has 
been already described in the introductory chapter (see p. 69). 
There are, however, as might be imagined, differences of 
detail. In Crotophaga the membrana tympaniformis com- 
mences at the seventh ring of the bronchi, the rings in front 

1 Apparently absent or very slight in CiucuTu*. 



being perfectly complete rings, the trachea dividing, ' as in the 
mammalia.' From the seventh onward all the bronchial rings 
are semi-rings, the intrinsic muscles being attached to the 
tenth. This is one extreme of the series, the other being 
offered by such a type as Piaya. In Piaya (see fig. 137) 
there is a purely tracheo-bronchial syrinx. The third bron- 
chial semi-ring is of compara- 
tively speaking enormous size, 
and to it are attached the in- 
trinsic muscles of the syrinx. In 
Satirothera we have a syrinx 
which is quite similar save for 
the fact that the third bron- 
chial ring is not enlarged. 
Diplopterus is much the same 
as the last. Cuculus has also 
a perfectly typical tracheo-bron- 
chial syrinx. In Eudynamis 
there is a cuculine syrinx, the 
last tracheal arid the first three 
bronchial semi-rings being ossi- 
fied ; the intrinsic muscles are 
attached, as in Piaya, to the Fl(t< 137 ._ SyBINX OF p iayu ca , /ana 

third bronchial semi-ring. (AFTER BEDDARK). 

Plifjcnicopliaes is much the same. 

The remaining genera of cuckoos whose syrinx is known 
are nearer akin to Crotophaga, though in them the bronchial 
syrinx is not quite so typical. In Centropus ateralbus, for 
instance, the first fifteen rings of the bronchi are incomplete 
internally and are closed by membrane, but the membranous 
area is narrow ; this area widens out at the sixteenth ring, 
which with the following is much stronger than the pre- 
ceeding and succeeding rings of the bronchus ; to the sixteenth 
ring are attached the intrinsic muscles of the syrinx. Pyr- 
rhocentor and Geococcyx have a similar syrinx. The syrinx of 
Guira is in many respects very remarkable. On a superficial 
view it is not unlike that of Cuddus. The voice organ in 
this genus is placed further forwards than in the genera just 


considered. But the first two or three rings of the bronchus 
are complete rings with no membrane internally. From the 
fourth onwards the rings are semi-rings. Upon the sixth 
are inserted the syringeal muscles. Coua is somewhat inter- 
mediate. The first seven bronchial semi-rings have their 
inner extremities separated by a narrow area of membrane. 
To the seventh are attached the intrinsic muscles of the 
syrinx. From this point commences the tympaniform mem- 

FIG. 138. SYRINX OF Centropus atcralbus (AFTER BEDDABD). 

brane. The accompanying woodcuts will serve to illustrate 
the varying form of the syrinx among the Cuculi. 

As with some other large groups of birds, such as the 
pigeons and parrots, the gall bladder is present in some 
cuckoos and absent in others. The gall bladder exists in 
the genera Saurothera, Coccyziis, Guira, Pyrrhocentor, 
Scythrops, and Cuculus ; it is absent in Cro'tophaga and in 
some species of Centropus. Coua has a gall bladder ; Eudtj- 
iKtinix appears not to have one. 

In Guira the right lobe of the liver is five or six times as 
large as the left. 

The main artery of the leg is the femoral in Piaya, Cen- 
tropus, the sciatic in Diploptera, Saurothera, Cocci/:://*, 


The following are some intestinal measurements, princi- 
pally of the caeca : 

Cucnliis canorns . 
Piaya cayana 
Diploptcrus ncevius 
S<iu rothcra dominiccnsis 
Coccijzus americanits 
Pyrrhoccntor cclebcnsis 
Centropus ateralbus 

phasianus . 
PJut'tiicophacs sp. 
Geococcyx affinis . 
Crotophaga sidcirostris 
Guira piririgua . 

S. ]. 













1-7, 1-5 






1-75, 1-5 












' N 




As for the skull of the cuckoos, that of Scythrops has been 
described by PARKER l and by myself. 2 SHUFELDT has 
studied the skeleton of Geococcyx. 3 Scythrops is doubly des- 
mognathous, the maxillo-palatines being united for their whole 
lengths, and the palatines also being fused with each other 
posteriorly. In other cuckoos the maxillo-palatines diverge 
posteriorly for a short space, and there is no union between 
the palatines. Scijtlirops has, according to PARKER, two 
small vomers, situated one behind the other. 4 In cuckoos the 
ectethmoid processes are large, and the lacrymal has often 
(e.g. Scijtlirops, Crotophaga) a large descending process 
nearly touching the jugal . Scijtlirops and Eudynamis have an 
os uncinatum lying between the descending process of the 
lacrymal and the ectethmoid. 

The holorhinal nostrils are much obliterated by bony 
growths, the degree varying. As a consequence the nostrils 
are as a rule impervious ; but in the dried skulls of Pyrrho- 
centor and Cuculus there is a considerable foramen. 

No cuckoo has more than fourteen cervical rriii'brce, and 
some have only thirteen. Three or four ribs only articulate 

Trans. Linn. Soc. (2), i. '- P. Z. S. Is'.is. 

7. Anat. Phys. xx. 1886, p. 244. 

I could only find one lying entirely between the palatines. 


with the sternum. The sternum may be fenestrated or 
marked by posterior incisions. 

The atlas of Bliiiwcoccyx 1 is perforated by the odontoid 
process ; C11-D1 have hypapophyses in addition to the axis, 
and two or three following, C12-14, have also catapophyses, 
which on 14 ascend on to the haemapophysis, 

In Squrothera CIO has closely applied paired hsemapo- 

The modifications of the syrinx, the pterylosis, and the 
leg muscles permit the family to be subdivided thus, the 
subdivisions, as will be observed, corresponding to the 
geographical range of the birds : 

Subfamily I. Cuculinae. Syrinx tracheo-bronchial. Ven- 

tral feather tract single. Muscle formula AXY + . 
a. Ventral tract single at commencement. 

Coccystes (?) 
b. Ventral tract double at commencement. 
Saurotliera % 

Diplopterus (?) 

* I New World. 



Subfamily II. PhoenicophainsB. Syrinx tracheo-bronchial. 
Ventral feather tract bifurcate. Muscle formula 
ABXY + . 

Scytlirops \ 

Eudynamis mid World. 

PlioenicopJiaes > 

Subfamily III. Centropodinae. Syrinx bronchial. Ven- 
tral feather tract bifurcate. Muscle formula 
ABXY + . 

1 In Scythrogs there is a notch nearly completely converted into a foramen ; 
in Eudynamis a notch less nearly converted into a foramen, also in Gnira and 
Diplopterus. In Cuculus and Sawothcra there is a foramen. 


a. Ventral tract occupying whole of space between 

Old World. 

Con a (?) 

b. Ventral tract only occupying median region. 

Geococcyx \ 
Crotopha'ga hNew World. 

The question of the affinities of the cuckoos is a difficult 

By GADOW they are placed nearest to the Musophagi and 
next nearest to the Psittaci. FUEBEINGEE'S views do not 
greatly differ. There seems to be no doubt that these birds 
are an archaic group not far from the point where the Ano- 
malogonatae and Homalogonata? of GAEEOD diverge. They 
are, like the Musophagi, quintocubital ; their intestines are 
simple ; and they have the complete muscle formula (B being 
in some forms absent). These characters are found in others 
among the more primitive of the higher birds. 

The likenesses wiiich the cuckoos show to the Pico-Pas- 
seres are mainly in the structure of the foot, in the simple 
character of the tendons of the patagium, and the marked 
resemblance in the syrinx to that of the Caprimulgi, and in a 
less degree to the Striges. As has been pointed out, precisely 
the same series of modifications between the extreme bron- 
chial syrinx of Crotophaga and the purely tracheo-bronchial 
syrinx of Cucidus are to be seen among the Caprimulgi. The 
syrinx is really the only salient point in the anatomy of 
the group that can be laid hold of for purposes of com- 
parison, and, considering the dissimilarities in the voice and 
habits of cuckoos and goatsuckers, it is particular])' note- 

As to fossil cuckoos, the two most interesting facts are, 
perhaps, the occurrence of Centropus and Phoenicopliaes ' in 

1 See MILNE-EDWARDS in Comptes Rendus for 1894. 


Europe, especially the latter, as it has some claims to repre- 
sent the most ancient form of cuckoo, with complete muscle 
formula and tracheo-bronchial syrinx. 1 


Definition. Oil gland tufted. Aftershaffc present. Quintocubital. 
Rectrices, ten. Muscle formula of leg, ABXY + . Expansor 
seoundariorum present. Biceps slip absent. Caeca absent. Both 
carotids present. Skull holorhinal, desmognathous, without 
basipterygoid processes. 

This group of birds, purely African in range,' 2 is divisible 
into three genera, Corythaix, Musophaga, and Scliizorhis. 
These genera do not show a large amount of structural 

As to the pterylosis, the two ventral tracts are double 
upon the neck (in C. albocristata) ; they remain separate 
until just in front of the cloaca, being especially weak and 
narrow in the breast region. Longitudinally arranged rows 
of feathers connect the pectoral tracts above with the 
humeral. The other important external characters are stated 
in the definition. 

In Schizorhis the normal arrangement of the leg arteries 

In Corytliaix and Musophaga the femoral artery is the 
-one developed. The right jugular is the largest, and in 
Corythaix albocristata, seems to have entirely disappeared. 

In the liver the right lobe is the larger, 3 sometimes 
considerably so. The gall bladder is present, and sometimes 
is elongated in form. The tongue is short and triangular ; 
the proventriculus is zonary, the gizzard weak. The in- 
testines are capacious and short, without ca3ca. The following 
are a few measurements : 

1 Cf. also partial persistence of basipterygoid processes in Phoenicophainse. 
; The extinct Necrornis of French Miocene may be a Touraco. 
3 The viscera are described by OWEN for Corythai.i- /><>i-j>Jii/rcoIo}i]iit, I'. Z. ,S' 
1834, p. 3, and by MARTIN for Corythaix Hiitfnnii. ibid. is:-{li, p. 3'2. 



Corythaix erytli nilopli us 
,, albocristat<! 
,, persa 

Musophaga violacca . 

Sell izo rh is africanus 

17 inches 



The windpipe of Corythaix persa is slightly swollen 
along its course, narrowing again at the bifurcation. It is 
much ossified. Counting as 
the last tracheal ring that to 
which the pessulus is attached 
in front, the intrinsic muscles 
are inserted on to the third 
in front of this. The first 
two bronchial semi-rings are 
ossified ; the third, between 
which and the second there is 
a considerable membranous 
interval, is the first of the 
purely cartilaginous series. 
The extrinsic muscles are 
stout, and arise seven or eight 
rings from the end of the 
trachea, and pass at once to 

their insertion ; they run no distance along the trachea, as is 
so common. 

Musophaga has no intrinsic muscles ; Schizorhis has. 

As to muscles, the tensores patagii are very simple, and 
the biceps slip is entirely absent. 

The tensor breves sends off a wristward slip just before 
its insertion in both Musophaga and Conjthaix ; l it is 
reinforced by a pectoral slip and by a fibrous slip from the 
humeral crest. 

The anconceus long us has not, at any rate in Corythaix 
albocristata, a humeral head. Glutceus I. is large, covering 
the biceps ; glutans V. is absent (Corythaix erythrolophiix, 
Musophaga} or present (Schizorhis africanus}. The muscle 


ehlorochlamys (AFTER MITCHELL). 
^shorWircuiting vessel divided. 

1 Absent perhaps as an individual variation in C. albocristata. 



formula of the leg, as stated in the definition, is ABX Y + . 
Both peroneals are present. The deep flexor tendons are 
bound by a vinculum, which is single in Corytlmix and 
double in Schizorhis. 

The skull of Conjthaix is barely desmogiiathous, and 
by no manner of means especially like that of a cuckoo, 
to which group the Musophagi have been often compared. 
The hinge in the middle of the face is nearly complete, but 
there is a bridge on each side, formed b t y an ankylosis 
between the frontal and nasal. The holorhinal nostrils are 
situated very far forwards, and each has, as PAEKEE ' has 
pointed out, an osseous fold upon the ossified internarial 
septum. The maxillo-palatines diverge from each other 
posteriorly for a much longer space than in any cuckoo. 
The ascending laminae of the palatines come into contact 
for a very brief space over the rostrum in front, and are 
continued forward for a short distance as a sharp spike. 
Between them lies a minute interpalatine splint (or vomer). 
The interorbital septum is moderately fenestrate. The 
lacrymal bones are of some size, and the descending process 
is closely applied to, but does not fuse with, the square 
ectethmoid process ; connected with both is a small os 
uncinatum, which reaches the palatine. 2 

Bony nostrils . 


Maxillo-palati ncs 
Jugal bar 


dorsal Not so continuous. 

Close to fronto-nasal Near end of bill. 

Continuous with 

wall of orbit. 
Ascending laminae come 

into contact posti'riurlij 

for a considerable 

Ascending laminae come 
into contact anteriorly 
for a sliort space. 


Move completely fused. 
Expanded where it joins 

maxilla on, or close to, 

ventral surface. 

Less completely fused. 

Not expanded ; junction 
with face higher up, 
very much as in 

In view of the general opinion as to the nearness of the 

1 In his paper upon Opixtliocnunis. See below, p. 286, footnote 7 . 

2 J. T. REINHARDT, Yid. Medd. Kjobenhavn, 1871, p. 3'2(>. 


alliance between Musophagidse and Cuculidte it may be 
useful to tabulate the principal divergences in the skull. 

Corythaixh&s~L4: cervical vertebra' ; ' the <tt1x is notched, 
not perforated by the odontoid process. Four ribs reach 
the sternnm, which is doubly notched and has a strong 
spina externa. 

In the pelvis the prepubic process is very markedly 
developed, as in Geococcyx. 

The clavicle comes into contact both w r ith the scapula 
and with the moderately large procoracoid ; the latter is 
fused with the acrocoracoid, 2 making thus a complete bridge 
over the sulcus supracoracoideus. The coracoids slightly 
overlap at their articulation with the sternum. 

The heemapophyses are characteristic. In Corytlia ix albo- 
cristata Cll has paired processes ; on C12 and C13 the hsema- 
pophyses form a continuous ventral keel to those vertebrae. 
The three following vertebrae have hsemapophyses. The 
extremities of those of Dl and D2 are expanded with a 
median ridge, ow T ing to the catapophyses having descended 
them. 3 


Definition. Aftershaft present. Oil gland feathered. Rectrices, ten. 
Q/uintocubital Muscle forrrmla, ABX~S~ + . Biceps slip and 
expansor secundariorum present. Carotids, two. Caeca present. 
Skull holorhinal, schizognathous, without basipterygoid pro- 
cesses. Sternum peculiar in form, wider behind than in front ; 
the spina externa ankylosed with furcula. 

Tjiere is no doubt that this group, consisting of but 
a single genus and species, Opistliocomus cristtus, the 
hoatzin of British Guiana, forms a well-marked group of 

The external characters of the adult and young have been 
chiefly described in recent years by NITZSCH, PYCBAFT, 4 and 

1 In Corythaix persa the atlas ring is incomplete. 

2 Cf. Galli, in which the same fusion occurs. 

3 E. BLAKCHARD, ' Remarques sur 1'Osteologie des Musophagides,' 
Rend. xlv. p. 50!). 

4 ' On the. Pterylography of the Hoatzin,' Ibis, 1895, p. 345. 


myself, 1 the internal anatomy of the soft parts by PERRIN, 2 
GARBOD, a GADOW, 4 MITCHELL/' and myself. 

The pterylosis of Opisthocomus shows a less sharp differ- 
entiation into pterylae and apteria than in many birds ; many 
of the spaces are covered with a sparse feathering of semi- 
plumes. This condition may be, as GARROD has suggested, 
immediately derived from a continuous feathering. Never- 
theless there is a sharply marked apterion upon the carina 
sterni, which is necessarily, therefore, of limited extent. The 
dorsal apterion is but feebly marked. In the young chick 
the ventral apterion was as clear as in the adult, but I could 
discover no trace of a dorsal bare space. In the young un- 
hatched chick the ventral feathering was closer than the 


The skin lying upon the carina sterni is dense and thick- 
ened, a state of affairs which appears to have some relation 
to its habit of squatting close to the branch upon which it is 

The principal recent papers upon the osteology of this 
bird are those of HUXLEY G and the more recent and more 
elaborate treatise of PARKER/ The cervical vertebra are 
nineteen in the adult ; but PARKER found only eighteen in 
the unhatched young. The atlas is notched for the odontoid 
process. Hsemapophyses are very feeble. The last two or 
three cervicals are ankylosed with each other, and form part 
of the dorsal series of ankylosed vertebrae. There is some 

1 ' Contributions to the Anatomy of the Hoatzin,' &c., Ibis, 1889, p. 283. 
See also C. G. YOUNG, ' On the Habits and Anatomy of Opisthocomus crixtatux ; ' 
Notes Lcj/den Mus. x. p. 169 ; and .T. J. QUELCH, ' On the Habits of the 
Hoatzin,' 'ibis, 1890, p. 327. 

2 ' On the Myology of Opisthocomus cristatus,' Tr. Zonl. Soc. ix. p. 353. 

3 ' Notes on Points in the Anatomy of the Hoatzin,' P. Z. S. 1879, p. 109. 

4 ' Description of the Modification of certain Organs,' &c., Zool. J.B. v. Abth. 
Syst. v. 1891, p. 629, and Proc. R. Irish Ac. (3), ii. p. 147. 

r> See below, p. 288, footnote, for reference. See also a recent paper 
by GOELDI in Ornith. M.B. May 1895, and more fully in Bol. Mus. Para. 


B ' On the Classification and Distribution of the Alectoromorphss,' P. '/,. S. 


On the Morphology of a Reptilian Bird, Opisthocomus cristatus,' Tr. '/.. S. 



variation as to the number of cervical vertebrae, which bear 
long rib stylets. As a rule five complete ribs exist, of which 
all bear uncinate processes. The sternum (see fig. 140) is ex- 
ceedingly remarkable in its form. It is wider behind than in 
Front, with a pair of notches, and outside of these a pair of 
foramina ; the keel is shorn away anteriorly, but well deve- 
loped posteriorly. The furcula, which is shaped like a fork 
with nearly straight lines, is completely ankylosed on the 
one hand with the coracoids, and by its median region with 

. 140. STERNUM OF Opisthocomus. 


tliocomus. FRONT VIEW. 

the spina externa sterni. The region of the furcula, however, 
which comes into contact with the sternum was found by 
PAEKER to be a separate ' needle of bone,' which he regarded 
as the iuterclavicle (see p. lol). The scapula is provided in 
the young with a distinct suprascapula, segmented off from 
the scapula. 

The pelvis is especially compared by HUXLEY with that 
of Gotuni-i.r ; it has no ileo-pectineal processes. 


In the skull the rostrum is articulated with the frontal 
region by a well-marked hinge. The skull is holorhinal, 
schizognathous, and there are no basipterygoid processes. 
The vomer is expanded and bifid in front in a fashion that 
recalls the segithognathous skull and that of certain of the 
Charadriiformes (see below). 

In the young skull PARKER figures basipterygoid pro- 
cesses, not, however, articulating with the pterygoids ; they 
appear to be not unlike those of Aptornis (see below). The 
holorhinal nostrils are partly obliterated by an ossified ali- 
nasal, as in so many picarian and passerine birds. 

The alimentary canal of Opisthocomus is remarkable in 
more than one way. There is, in the first place, the enormous 
crop, which has been most recently and most fully described 
by GADOW. This organ is very large, and rests upon the 
furcula and the fore part of the sternum, for the abortion of 
the anterior part of whose keel GADOW thinks the crop is by 
its pressure responsible. The crop too is exceedingly muscu- 
lar, and has numerous parallel folds in its interior, some of 
which are continued into the oesophagus below. The gizzard 
is much reduced in size. Probably the crop is not a mere 
storehouse, but a compartment where at least trituration of 
the food (chiefly leaves) takes place. 

The most remarkable feature about the intestine ' is 
the long and coiled rectum, a feature which is also found 
among the struthious birds and in the archaic Cliauna. 
The general arrangement of the coils of the small intestine is 
intermediate between those of Pterocles and pigeons. There 
is in the middle loop a faint trace of the spiral found in the 
corresponding loop of the pigeon's gut. There are also like- 
nesses to the form of gut in the cuckoos. The cccca are 
fairly developed. 

The peculiarities of the muscular system mainly concern 
the hind limb, and chiefly characterise the amUens. The 
muscle formula is complete, i.e. ABXY+. The ambiens 
however, is subject to variation. GAREOD found that in all 

1 P. CHALMERS MITCHELL, ' A Contribution to the Anatomy of the Hoatzin 
(Opisllwccnmis cristatus),' P. Z. S. 189(3, p. 618. 



of six knees that lie examined the ambiens was present, 
though small ; but in only one knee did it cross the knee to 
be inserted in the usual fashion in connection with the flexors 
of the leg. MITCHELL dissected this muscle in two specimens ; 
in one the ambiens \vas completely absent above the knee, 
but in each case (see fig. 53, p. 96) a ligament left the fibula, 
and, dividing into three, joined each of the three perforated 
flexors in the way in 
which, as has been 
already described, it oc- 
curs in birds which have 
this ambiens rudiment. 
In the second case there 
was an ambiens above 
the knee, but it became 
lost upon the fascia of 
the knee, and not con- 
nected with the ambiens 
rudiment springing from 
the fibula, which was 
there present. So in 
this bird there are many 
stages in the reduction 
of this characteristic 
muscle, which is clearly 
in them on the wane. 

It is apparently the 
rule among birds for 
there to be a viiiculum 

between the two superficial flexors of digit III. This slip 
is wanting in Opisthocomus, as it is, according to MITCHELL, 
in Asia otus and Rhytidiceros plicatns. The deep flexor 
tendons are connected by a strong viiiculum. 

The syrinx has been described by GAEEOD and myself. 
The accompanying figure is from GAREOD'S paper. The 
last few rings of the trachea are solidified into a tracheal box, 
and the intrinsic muscles do not reach this box, being only 
continued 011 to it by a ligarnentous continuation. There is 


FIG. 142. SYRINX OF Opisthocomus. 



some variation in the number of rings which coalesce to form 
the box, while the fibrous continuation of the intrinsic muscles 
may reach the first bronchial semi-ring. This muscle is 
evidently decaying in Opisthocomus. 


Definition. Quintocubital birds with, an aftershaft. Muscles of leg 
generally ABXY + . Expansor secundariorum present. Entepi- 
condylo-ulnaris present. Caeca large ; a crop present. Skull 
schizognathous, holorhinal, with sessile basipterygoid processes. 
Palatines without internal lamina. 

This very large group of birds, universal in range, shows 
a considerable amount of structural variation. 

The oil gland is generally tufted ; '-' but it is nude in the 
Megapodes and absent altogether in Argus. 

The pterylosis of the Galli is, according to NITZSCH, 
singularly uniform. He figures Gallus bankiva, Pavo 
cristatus, and Meleagris gallo-pavo, describing also a few 
other types. There are lateral neck spaces in all ; the dorsal 
tract is single in Gallus, widening out on the back ; in the 
peacock it widens out in a more pronounced fashion and 
further back than in Gallus. In the turkey there is a 
narrow space in it between the shoulder blades. 

The ventral tract divides early upon the neck, and each 
tract gives off on the breast a wider, denser outer branch ; 
the two median tracts then continue nearly to the cloaca, 
where they unite. 

InPerdix and Tetrao there is a dorsal space, as in Meleagris. 

Among the Cracidse there may or may not be a space in 
the dorsal tract. 

The pectoral muscles of gallinaceous birds, like those of 
the tinamous, meet over the keel of the sternum ; this is at 
least the case with Euplocamus Vieilloti and some others. 

1 H. SEEBOHM, ' An Attempt to Diagnose the Sub-Orders of the Great Galli- 
naceo-Gralline Group of Birds by the Aid of Osteological Characters alone,' Ibis, 
1888, p. 415. 

2 Callipepla calif arnica has a small tuft ; in C. squamata I have observed 
both the complete absence of a tuft and the presence of a very small one. 

GALLI l'91 

The deltoid may or may not possess a special tendinous 
slip from the scapula. This slip is absent in Mitua tomentosa, 
Excalfactoria cliitiensis, and Callipepla, but present in 
Ortalis alhiri'ntris, Crax Sclatcri, C. Daubeiitoni, Crosso- 
ptiloii mantchuricum. It is evidently, therefore, not of great 
use in classifying the group. 

The biceps slip is generally present, but absent in 
Ortalis albiventris, Crax, Mitua ; it is present in. Megapodius 
and Megacephalon', 1 absent in Talegalla, Niiiniifa, and 
Mi'leagris. The same remark may, therefore, be made about 
this muscle. The humeral head of the anconcBus is not 
always present. 

The tensor patagii brevis of gallinaceous birds has a 
thin, wide, diffused tendon, as in the tinamous ; there is no 
patagial fan. 

The entepicondylo-ulnaris is another muscle which they 
share with the last-mentioned group. 

The expansor secundariorum is a muscle which appears 
to be invariably present among the Galli, but to have vary- 
ing relations at its scapular insertion. 

'In the majority of the gallinaceous birds,' wrote Pro- 
fessor GARROD, ' the expansor secundariorum, with the normal 
origin from the secondary quills, has a different method of 
insertion, which has led M. A. MILNE-EDWARDS to describe 
the muscle in the common fowl as a part of the coraco- 
braclnalis (brevis) in his superb w r ork on fossil birds. 

' In the genera Tetrao, Francolinus, Rollulus, Phasianus, 
Enplocaiiius, Gall us, Ceriornis, and Pavo, the muscle, instead 
of being inserted into the scapulo-sternal fibrous band, above 
referred to, after blending to a certain extent with the axillary 
margin of the teres, ceases by becoming fixed to a fibrous 
intersection about one-third down the coraco-brachialis 
brevis muscle. 

'In Francolinus Clappertoni from among the francolins, 
Coturnix, Odontophorus, Ortijx, Eupsychortijx, and Niunida, 
the tendon does not go so far as the short coraco-brachialis, 
but ends either by simply joining the axillary margin of the 

1 Absent, according to Ft'-RBRiNGEB. 

u 2 


teres or by at the same time sending a tendinous slip 
behind it to the scapula. In Arc/us gi.gaiiteus the tendon, 
running from the elbow, turns round the axillary border of 
the teres to end by joining a triangular muscular fasciculus, 
attached by its base to the upper portion of the thoracic 
surface, which appears to be nothing but a differentiatioii-off 
of the upper portion of the last-named muscle. In the 
Crttcidff this insertion into the scapula is also found, but it 
is tendinous, like the upper element of the thoracic band 
above described in the storks and Chauna ; and in them 
there is also a second tendinous slip from the axillary margin 
of the coraco-brachialis longus (not the brevis) . Iri the Mega- 
podidse also the attachment to the coraco-brachialis brevis is 
wanting, the tendon ending either by blending with the teres 
margin or running on to the scapula.' 

The glutceus primus is a large muscle covering the 
biceps. Gliitff'its V. appears to be alwa} T s present, but is 
sometimes (Tliaumalea picta) quite tendinous. 1 

Most gallinaceous birds have the complete muscle formula 
ABXY + . The femoro-caudal, however, varies in size, 
and is quite absent in Pavo and Meleagris. It is very 
slender in Crax and Ortalis.' 2 

The deep flexor tendons belong to type I., and are illus- 
trated in fig. 54 (p. 100). 

There are two carotids in all but the Megapodes, where 
the left only is present. A gall bladder is present. 

Some intestinal measurements are given on p. 293. 

The trachea has in a few gallinaceous birds two pairs of 
extrinsic muscles, thus resembling, it will be observed, the 
Anseres and Palamedeae. 

Thus in Crax Daubentoni, besides the usual sterno-trache- 
ales, which arise in the ordinary way from the costal processes, 
there are a pair of cleido-tracheales, springing from the 

1 For the tail muscles of the peacock see HEMMING. Proc. Linn. Soc. 1844, 
p. 212. 

2 GAEROD has figured (in MS.) an ''abnormal Gallus domesticus with a 
peculiar additional muscle springing by tendinous slips from femur, femoro- 
caudal, accessory ditto, and semi ten dinosus, and running to gastrocnemius. 1 
It was the same on both sides. 



Name of Bird 

Small Iritest. 

Large Intrst. 

i ii'oa 




I'avo cristatus ..... 




nigripennis ..... 



8 and 9 

) ..... 




muticus ... 




,, spicifer .... 

46 ,- 

, 3-5 


Caccabis cliukar ..... 



Argus giganteus ..... 

70 to 84 


5-5 to 7-5 

i >* ..... 




Ithaginis Geoffroyi .... 

31 , 

, 3-5 


Polyplectron chinquis .... 




t? ,, bicalcaratum . 




Rollulus coronatus .... 




Arboricola torqueola .... 




Coturnix communis .... 


Ortyx virginianus .... 




cristatus ..... 




Gambelii ..... 




Odontophorus dentatus 




Perdix cinerea 


Phasianus versicolor .... 




Thaumalea Amhersti'se 




Euplocamus Swinhoii .... 




erythrophthalraus . 




Vieilloti .... 




nycthemerus . 




,, cristatus .... 




,, albo-cristatus . 








nobilis .... 




Ceriornis Temmincki .... 




<? ,, satyra ..... 




Lophophorus impeyanus 




Crossoptilon inantchuricum 




Lobiphasianus Bulweri 




Gallus bankiva ..... 








Numida meleagris .... 



,, ptilorhyncha .... 




,, cristata ..... 

33 _ 

, 3 


,, Edouardi .... 




vulturina .... 



9 and 10 

Tetrao urogallus 








,, phasianellus .... 






Meleagris ocellata 




Francolinus afer 


-- , 2 


,. gularis .... 




Crax globicera 



7 and 8 





Daubentoni ..... 




,, Alberti 




globulosa ..... 


- 5 


Mitua tuberosa ..... 




,, tomentosa ..... 



r> f) 

Penelope cristata .... 





Name of Bird 

Small Iiitest. . Large Intc-r. Cajca 

Penelope cujubi . 

,, cujubi . 

pileata . 

,, jacucaca 
Pipile jacutinga . 
,, cumanensis 
Aburria carunculata 
Ortalis albiventris 
Talegalla Lathami 
Megacephalon maleo 
























> , 









anterior end of the sternum and from fibrous septum between 
it and the pessular process ; these muscles run up the sides 
of the trachea, reaching further than the sterno-tracheales. 

Where the trachea is convoluted it sometimes happens 
that the extrinsic muscles are quite abnormal in their attach- 
ments ; thus in the males of Penelope pileata and Ortalis 
albiventris the muscles in question do not enter the thoracic 

FIG. 143. SYHINX OF Pctro spicifer. 


cavity, but pass close to the carina sterni and are inserted 
at the very end of the sternum. 

The syrinx of gallinaceous birds has been chiefly studied 
by GAEEOD.' From his paper the following account of this 

1 ' On the Conformation of the Thoracic Extremity of the Trachea in the 
Class Aves, Pt. 1, The Galling,' P. Z. S. 1879, p. 354. 



organ has been mainly drawn. The simplest form of the 
syrinx is seen in the peacock (fig. 143), where the modification 
of rings at the bifurcation is of the slightest. The last two 
tracheal rings are partly fused behind. The accompanying 
series of illustrations (figs. 143-147) show some of the princi- 
pal forms of syrinx among the Alectoropodes, in which it will 

FIG. 145. SYHIXX OF Callipepla califor- 


be noticed that intrinsic muscles are but occasionally present. 
When present they do not descend to the bronchi, but cease 
upon the trachea some way in front of the bifurcation. The 
most remarkable modification of the intrinsic muscles is in 
Callipepla calif or nica (figs. 145, 146), where the muscles 
descend the trachea posteriorly, and are inserted on to the 
bronchidesrnus. This state of affairs is not unlike what is 
found, and will be described (see below r ), in the condor. It is 
uncertain whether these muscles may be not more accurately 



referred to the cleido-tracheal extrinsic pair (see below). A 
very singular syrinx is that of the male Tetrao tetrix (fig. 147). 
On each side of the trachea at the lower end is an ' immense 
irregular tumefaction, connected with its fellow by a bridge 
of fatty tissue.' It appears to be mucous in its chemical 
nature ; but it may possibly have some relation to the tracheal 

i.'t v h \ 

FIG. 147. SYEINX OF MALE Tetrao tetrix. FRONT VIEW. 

box of the male ducks, and be thus another of those many 
unexpected resemblances between the two groups. 

The Cracidse (fig. 148) generally possess the intrinsic mus- 
cles, which are, however, short, as in the other Galli. The 
syrinx itself has no salient characters by which it may be 
distinguished from the Alectoropodes. 

GARROD l has also described and figured (fig. 150) the 
syrinx of the megapode Megacephalon maleo. It is rather 

1 ' On the Anatomy of the Maleo,' P. Z. S. 1878, p. 629. 



peculiar in form, but has a pair of intrinsic muscles, which 
reach the first bronchial semi-ring ; in this point the syrinx 
is more primitive than that of other Galli. 

FIG. 148. SYRINX OF Abiirria carnii FIG. 149. SYRINX OF SAME. BACK 

The skull in gallinaceous birds is in many respects 
remarkably .duck-like. 1 The basipterygoid processes a little, 

FIG. 150. SYBINX OF Mcgaceplialon malm. A. FRONT VIEW. 

but not much, more pronounced in the Megapodes are oval 
sessile structures, with which again, as in the ducks, the ptery- 

1 As pointed out by PARKER. 


golds articulate by their anterior ends. The palatines too are 
devoid of an internal lamina, and the angle of the mandible is 
recurved and produced ; it is enormous, extended upwards, in 

In the typical gallinaceous birds the maxillo-palatines 
are generally small or even obsolescent. In Gallus bankiva ' 
they are triangular plates of fair size ; in Tetrao urogalhis 
(fig. 152) they are small, narrow, backwardly projecting 
plates, not quite so long and thin, and not so curved as those 
of TalegaUa. In Ptilopacliys they are somewhat inter- 
mediate ; in Callipepla calif ornica they are still longer. In 
Numida and Meleagris they are much the same. 

The lacrymals are not large, 2 and have a feeble or aborted 
descending process. It is very general for the post-frontal 
and zygomatic processes to fuse and enclose a more or less 
triangular foramen, and sometimes, as in Tetrao* and Crosso- 
ptilon,, the zygomatic bar extends forwards a considerable 
way in front of its junction with the other. 

The interorbital septum is sometimes (Coturnix, Calli- 
pepla, Pcrdicula, Ptilopacliys) considerably fenestrated. 

The vomer, in gallinaceous birds generally, is thin and 

As to the Megapodes, there are some differences in the 
skulls of the' two genera TalegaUa 4 and Megacephalon. The 
latter has the well-known hammer-shaped projection of the 
back part of the skull. In both genera the palatines are 
slender, and there is some ossification of the nasal septum. 
The interorbital septum is not much fenestrate, but it is 
deficient in front. In Talegallaihe maxillo-palatines are thin 

1 SIIUFELDT, ' Observations upon the Morphology of Callus bankii m a,' &c. 
Journ. Comp. JlZt'rf. Surg. ix. p. 343. 

- WOOD-MASOX has described (Ann. Mag. Nat. Hist. xvi. 1875, p. 145) supra- 
orbital bones in certain partridges. Cf. as to this point Tinamous, Psopliia, 
and Meniira. 

3 SHUFELDT, ' Osteology of the N. American Tetraonidse,' Bull. U. S. Geol. 
<S'/MT. vi. p. 309. 

4 The skeleton of this bird is described by PAKKEB, ' On the Osteology of 
Gallinaceous Birds and Tinamous,' Tr. Z. ,S'. v. p. 100. See also W. K. 
PAKKEB, ' On the Structure and Development of the Skull of the Common 
Fowl,' Phil. Trans. 1870, p. 15 ( J. 

GALL I L>!)9 

plates, ending in a curved point very much like those of some 
passerines (e.g. Pteroptochus). Each is vacuolate posteriorly 
in Talegalla. They do not nearly come into contact in the 
middle line. In Megacephalon these bones are spongy plates, 
which do nearly come into contact ; the palatines, slender in 
both birds, are more bowed in Talegalla, and thus enclose a 
wider interpalatine vacuity. The lacrymals of Talegalla are 
small and ankylosed to the skull wall ; the ectethmoids are thin 
plates. A curious difference in the skulls of these two birds 
concerns the nasals and premaxillaries. In MegacepJialon 
there is nothing worthy of special remark except the tumid 
outer part of the nasals ; in Talegalla the premaxillary pro- 
cess of the nasals approach each other in the middle line, 
and cut the nasal process of the premaxillary into two, an 
anterior and a posterior portion. 

In the Cracidas the maxillo-palatines are largish plates, 
concave inferiorly and convex above, which in Crax globicera 
actually come in contact for a short space, and fuse with 
each other and w r ith the median septum. In Pauxi galeata 
this fusion (perhaps owing to the great casque) is even better 
marked. There is nearly a fusion in Aburria caruncidata ; 
in Ortalis albivcutris, on the other hand, the maxillo- 
palatines are well apart. 

In Ortalis and Aburria there is no ossification of the 
nasal septum ; in Crax there is a median piece, which expands 
below to become attached to (C. S elate ri) or fused with 
(C. globicera) the maxillo-palatines. In Pauxi galeata the 
nasal septum, as might be expected, is quite complete and 
very strong. 

There are also in this group of gallinaceous birds, a series 
of stages in the development of the zygoma and the post- 
orbital processes. In Ortalis they are short ; in Abiu-ria 
longer and convergent ; in Crax Sclatcri and Pauxi galeata 
they meet distally, and enclose a triangular foramen ; finally 
in Crax globicera they are completely fused throughout, and 
form a stout triangular process. 

The lacrymals in this family are large, with a large 
descending process. The ectethmoids are but slightly ossified. 



The interorbital septum is more fenestrate than in the Mega- 
podes, but still not markedly so. 

The Galli have sixteen cervical vertebrae. In Gallus 
bankica the catapophyses of vertebra 10 nearly enclose a 
canal ; on the next vertebra the two processes have almost 
fused into a single one, the two processes being closely 


FIG. 151. SKULL OF Cra.v globiccra. SIDE VIEW. (AFTER HUXLEY.) 
/'/.?-, premaxilla ; />', nasal septum. 

soldered together for their whole length, and not, as is so often 
the case, divergent at the end. All the remaining cervical 
vertebrae have strong median haemapophyses, those of the 
sixteenth being fused with the two following at their ex- 
tremities (cf. Musophaga, p. 285). The last cervical vertebra 
and the three anterior dorsals are themselves fused. 

It will be simpler to compare the vertebrae of Gallus 
with those of some other Galli by means of the following 
table : 

First Haam- 

Last Hcem. 

Heem. fused 




Crossoptilon mantcliuricum 

Cll D4 



Numida cristata 

Cll D3 



Talegalla Lathami 

Cll D2 



Meqacephalon malco . . CIO ' Dl 



Callipcpla calif arnica . C13 D2 C16-D2 


Ptilopachys v entrails . Cll D3 D1-D3 


Aburrla carunculata . . Cl'2 D3 D1-D2 


Bifid at end. 



It is evident that not much of classificatory importance 
is deducible from the above facts. HUXLEY has used with 
more success the remaining parts of the skeleton. 

The gallinaceous birds are divided by HUXLEY into two 
main subdivisions, Peristeropodes and Alectoropodes. 

FIG. 152. SKULL OF Tctrao 
urocjallus. VENTRAL VIEW. 

il.rp, maxillo-palatines ; Vo, vomer ; 
MX, maxilla ; I'l ', palatines : 1't ptery- 
Koi'l- ; +, kisiijterygoid facets ;fmf, 




FIG. 153. STERNUM OF Crax 
globicera (AFTER HUXLEY). 


The former contains the Cracidae and Megapodidae, the 
latter the remaining families. 

The Peristeropodes may be thus denned : 

1. Sternum with not very deep inner notches (fig. 153) and 
with short obtuse costal processes, the anterior edge of which 
is at right angles to the long axis of sternum. 

2. Hallux on a level with other toes. 


The Alectoropodes in this way : 

1. Sternum with very deep inner notches (fig. 72, p. 128)', 
and with long costal processes, whose long axis corresponds 
with the long axis of sternum. 

2. Hallux attached above the level of other toes. 

The former group, which undoubtedly, so far as the above 
characters are concerned, is the more primitive, consists of 
the curassows and the Megapodes, limited respectively to 
Central and South America and to Australia and some of the 
Indian islands. 

These two families are distinguished by HUXLEY mainly 
on account of the differing form of the hallux ; but there 
are other points of dissimilarity, not known at the time when 
he wrote. The two families may be thus differentiated :- 

Cracidffi Megapodidse 

1. Hind toe not so long in Hind toe longer in pro- 

proportion to rest. portion. 

2. Oil gland feathered. Oil gland nude. 

3. Biceps slip never Biceps slip sometimes 

present. present. 

4. Two carotids. Left carotid only. 

5. Trachea generally coiled. Trachea always straight. 

The Alectoropodes may be readily divided into three 
groups, which may be thus differentiated :- 


1. Second metacarpal without backward process. 

2. Costal processes outwardly inclined (thus forming 

a transition between Alectoropodes and Peris- 

The remaining families, which will be distinguished from 
each other immediately, agree with each in differing from 
the Numididse in both the points mentioned. 


1. Postacetabulum longer than preacetabulum. 

2. Postacetabulum longer than broad. 

GALLI 303 

3. Furcula weak and straight (viewed laterally), with 
straight rod-like hypocleidium. 

In the remaining gallinaceous birds the preacetabular 
length is greater than the postacetabular (equal in Tetrao 
eiipido) ; the postacetabular area is broader than it is long ; 
the contour of the furcula is curved, with an expanded 
hypocleidium. The series of birds which have these cha- 
racters may be divided into the galline and the tetraonine 

In the former the postacetabular region is only moderately 
broad ; the hypocleidium is oval in contour ; the tarso- 
metatarsus is more than half as long as the tibia, and there 
are a number of smaller osteological marks. 

In the grouse-like birds the postacetabular region is very 
broad ; the hypocleidium has a triangular form, and the 
tarso-metatarsus is not half as long as the tibia. I do not 
give HUXLEY'S characters in detail, since he has pointed out 
that the two series meet among the partridges and quails, and 
cannot thus be sharply marked off. 

The Alectoropodes have a range which is related to their 
anatomical differences, as have the Peristeropodes. The 
Meleagridse are confined to America, the Numidida? to 
Africa, the Phasiaiiida3 to the Oriental region just encroach- 
ing upon the Palaearctic, while the Tetraonidse are Palsearctic 
and Nearctic. 

The Galli seem to be undoubtedly an ancient group of 
birds, a view which is upheld by their points of likeness to 
many diverse groups. 

That they are an ancient group is also shown by their 
quintocubitalism, the primitive character of the gut con- 
volutions, the often complete muscle formula, and the 
existence of basipterygoid processes. The existence of these 
structures has led to their comparison with other groups of 
birds. Less weight, however, appears to be due to these 
more general points of resemblance than to other slighter 
but equally constant similarities. The existence of the 
entepicondylo-ulnaris muscle is an example to the point. 


The Galli share this muscle with the Tinami only ; and no 
one will doubt on other grounds that the gallinaceous hirds 
and tinamous are distinctly related. It is, however, with 
the Anseres that the greatest number of special resemblances 
of this character exist. These have been admitted by 
HUXLEY and PARKEE ; and more recently, in one of his 
alternative schemes, SEEBOHM united the two groups. In 
these two groups the palatines have the peculiar character 
of wanting the internal lamina, which is at most indicated 
by a slight ridge ; in both of them the basipterygoid pro- 
cesses can hardly be described in those words, as they are 
but oval facets for the articulation of the pterygoids. The 
two pairs of extrinsic muscles of the syrinx are one of the 
strong reasons for uniting the Anseres and the Palamedese, and 
we have among the Galli forms in which there are the same 
extra pair of muscles present. The general habit of a 
gallinaceous bird is, it is true, somewhat remote from that 
of an anserine bird ; but Palamedea might with truth be 
described as a goose-like bird with external likeness to a 
curassow, or as a gallinaceous bird which had put on the 
characters of the Anseres. Its likeness to both is considered 
on another page. 

The existing genera Tetrao, Lagopus, and Francolmus are 
known from Pleistocene deposits in countries which they at 
present inhabit, and the species from which these few remains 
have been described are existing species. Similarly Coturnix 
Nova Zcalaiiilin' and Talegalla Lathami have been met with in 
the Pleistocene of Australia and New Zealand, both of them being 
species now living in the localities whence their fossil remains 
have been extracted. Phasianus is represented by a number of 
extinct species from Europe as old as the Miocene. Three totally 
extinct genera described by MILNE-EDWAKDS are PalcEortyx, 
Palceoperdix, and Taoperdix. These are all Eocene or Miocene 
and European. Taoperdix presents affinities to Mclcagris and 



Definition. Aftershaft absent. Oil gland absent or nude. Crop 
present. Caeca nipples. Syrinx with, asymmetrical extrinsic 
muscles. Of the leg muscles ABX always present ; deep plantar 
tendons of type I. Biceps slip present and peculiar in form. 
Skull schizognathous, sshizorhmal. Basipterygoid processes 
present, except in Didus. 

This group of birds contains something like 458 species, 
which are divided into sixty-eight genera. Pigeons are cos- 
mopolitan in range, and show a considerable amount of 
structural variation. 

The absence of an aftcrshaft and the often rudimentary 
condition of the oil gland are among the most important 
variable external characteristics of the pigeons. The latter 
is never feathered, and is sometimes totally absent. But its 
presence or absence cannot be made use of as a fact of great 
systematic importance ; for we find in the same genus 
Ptilopus two species, Pt. coronulatiis and Pt. superbus, with 
a minute oil gland, and two others, Pt. assimilis and Pt. 
puella, from which the gland has entirely disappeared. 

The rcct rices vary in number from 12 (e.g. Cohimbula) 
through 14 (e.g. Phaps), 16 (Goura), to 20 (Otidiphaps) . 
But the variations do not invariably coincide with the limits 
of genera. Phaps chalcoptera has 16 rectrices, P. elegans 
only 14. PJilogcenas Stairi has 12 rectrices, P. cruentata 14. 
There are other examples. Though for the most part 
aquincubital, it is remarkable to find in Columbula a quinto- 
cubital bird (MITCHELL) . 

I have by me a careful manuscript account of ihepterylosis 
of Didunculus by Mr. FOEBES. The dorsal tract is strong 
and broad upon the neck, narrowing a little below ; where 
it narrows it becomes stronger. Below the shoulder it bifur- 
cates and becomes suddenly weak ; below this point the whole 
dorsal region is covered with a weak and diffused feathering, 
which is especially weak over the head of the femur. In the 

1 C. J. TKMMIXCK, Histoirc Naturcllc Gtntrale des Pigeons et ctcs Gallimtcics. 
Amsterdam et Paris, 1813-1815. 




middle line posteriorly the dorsal tract is again strongish. 
The lumbar region is also strongly feathered, and the thigh is 
covered by a strong tract which ends very abruptly below. 

The ventral tract is much weaker, and on the neck 
gradually merges into the lateral space which extends from 
a short way below the head to the shoulder. Below this the 
ventral tract is still weaker, and does not bifurcate until the 
upper end of the carina sterni. The median apterion is 
oblong and narrow, and reaches the cloacal aperture. In 
the abdominal region the two tracts get stronger. The 
pterylosis of Coluiriba lima as figured by NITZSCH hardly 


GIZZARDS OF, , Ptilopus jcu/ibu, latraiifi. l>, ONE OF HORNY TUBER- 

differs, and is typical of the pigeons in general ; there is no 

All pigeons have a well-developed crop, the presence of 
which organ is presumably related to their fruit- and grain- 
eating mode of life. 

In Carpopliaga, a fruit-eating pigeon, the gizzard is very 
weak ; but in other pigeons this organ is very strong, its 
walls being even ossified in Caloen as tiicobarica. 1 In Ptilop//,* 
a cross section of the gizzard has the peculiar form shown 
in the accompanying drawing, where it is compared with a 

1 YERREAUX and DBS Mrns describe an exaggeration like this in Plimnorhina 
goliatli, where also the tubercles are ossified (Rev. Mag. Zool. 1&G'2, p. 138). 
See also FLOWEK, ' On the Structure of the Gizzard of the Nicobar Pigeon,' 
P. Z. S. 1860, p. 330. 



pigeon showing a more normal state of affairs. There are 
four muscular pads in its walls, so that in transverse sections 
the lumen is cruciform. In Ptilopus coronulatus the lumen 
of the gizzard is not regularly cruciform, like Pt. janibit, 
but irregular and asterisk-shaped ; so with Pt. superbus. 
Chryscena viridis has also a ptilopine gizzard. In Carpo- 
phaga paulina the transverse section shows a close approach 
in the structure of its gizzard to that of Ptilopus. In C. 
l(ttrans (fig. 155) the gizzard is lined with extraordinary 
conical horny processes. 1 The right lobe of the liver is 
larger than the left, and 
the gall bladder may be 
absent or present. 

' In the Columba? 
which I have examined 
(Cohuubce of several 
species (fig. 156), PJilu- 
goenas cruen.tata)," re- 
marks Mr. MITCHELL, 
' it is tempting to regard 
the gut as a simple deri- 
vative of the type seen 
in Pterocles. The duo- 
denum is longer and 
narrower. The circular 
loop is enormously ex- 

FIG. Io6. INTESTINES OF Cohtmoa hvia 
pailded, but the three (AFTER MITCHELL). 

subsidiary loops seen 

in Pterocles remain. The first of these is somewhat shortened ; 
the second, that bearing the yolk-sac vestige at its end, is 
enormously lengthened ; the mesentery is folded along the 
line of the median mesenteric vessel, so that the two limbs 
of the loop are brought in contact with each other, and, 
finally, the whole folded loop is rolled into a rough spiral. 
The third subsidiary loop of the circular part of the gut has 

1 B. GERMAIN, ' N ote sur la Structure du Gesier chez le Pigeon Nicobar,' Ann. 
Sci. Nat. (5), iii. p. 352 ; GARROD, ' Note on the Gizzard and other Organs of 
Carpopliaga latrans,' P. Z. S. 1878, p. 102. 


the same arrangement and veins as in Pterocles ; but the 
caeca no longer run along it, but occur as very short stumps 
upon the rectum.' 

The cfBca are small and nipple-like ; they may be entirely 
absent. In Tympanistria bicolor I found one, an especially 
minute one, on the left side. 

The intestines of the Columbse are very short and volu- 
minous in some of the fruit-eating forms, moderately long in 
the majority of forms, and extraordinarily long in Didun- 
culus. The following are a few measurements : 

Ft. I IK. Ft. Ins. 

Carpophaga cenea . 1 6 Chryscena viridis . 1 

Chalcopkaps chnjso- Macropygia lepto- 

clilora . . .18^ grammica . . 2 10 

Phlogcenas Stairi . 2 6 Peristera Geoffroiji 10 


cruentata 2 

lanthoenas Icuco- 
losma . 3 11 

Columba maculosa . 2 8 

Geopelia cuneata . 8| Goura coronata . 5 1 

,, Jiiimilis .16 ,, Victor ice . 4 

Ptilopus jambu . 1 Phaps clialclwptera 2 6 

,, melanoce- 
phalus . . 9 
Ptilopus coronulatics 6 

,, clegans . 1 
Didiinculus K 

rostris 1 

In Goura Victoria the tensor brevis muscle is bordered 
on the patagial side for the last half af its course by a strong 
tendon, which arises from the pectoralis muscle. Below this 
tendon forms the outer and stronger part of a thin and 
rather wide tendon, in which the muscle itself ends. A 
wristward slip is given off, but there is no patagial fan. The 
biceps slip arises tendinously from the biceps, has a short 
muscular belly, and ends tendinously upon tensor patagii 
longus tendon. 

Goura coronata is much the same, but the biceps slip is 
(? exceptionally) digastric, a second tendon springing from 
tensor patagii longus muscle, and becoming muscular before 
it joins the muscular belly derived from biceps itself. 

In other pigeons and this is one of the most salient 


features of the group the tensor brevis muscle is often very 
extensive in fact, overlapping the extensors of the fore arm. 
This is the case, for instance, with Geophaps scripta, where 
I could find 110 deltoid beneath it. In Ptilopus and Pliaps, 
on the other hand, a good deltoid is present, and though mainly 
attached to the humerus is also at its extremity inserted on 
to tendon of tensor brevis. This latter insertion is wanting 
in many pigeons which have a long deltoid. In Erythrcenas 
the deltoid reaches nearly to elbow. There is very often a 
wristward slip from tendon of tensor patagii brevis, but never, 
so far as I have ascertained, a patagial fan. 

The biceps slip appears to be alw r ays an independent 
muscle, arising, as described above, by a thin tendon from 
the biceps, and it frequently has a second tendon of origin, 
as in Goura coronata. This is the ease, for example, with 
Carpopliaga, Didunculus,. and Phaps. It is absent in 

The expansor sccnndariorum is very constantly present ; 
but there are indications that it is on the wane in these birds. 
In Phlogoenas cruentata the tendon has not the characteristic 
T-shape that it has in other pigeons, but blends with a trian- 
gular fibrous slip arising from the scapula near the scapulo- 
coracoid joint. In Carpophaga paulina there is much the 
same arrangement. In PhlogoKnas Stair i the muscle appears 
to be totally absent. 

Of the two latissimi dorsi muscles the posterior is some- 
times wanting in pigeons. 1 HASWELL 2 first directed attention 
to this point, which was denied by GADOW S and FORBES, 
but reaffirmed by FURBRINGER. According to the last-men- 
tioned observer its occasional absence is a matter of individual 

The ancon,(Ens has a tendinous insertion on to the 

1 And also (? occasionally) in Ot:s,Plerodcs, and various passerines, according 

to FulUiKINGEK. 

2 ' The Myological Characters of the ColumbidaV Proc. Linn. Soc. N. S. W. 
iv. 1880, p. 303 ; ' Note on the Anatomy of Two Rare Genera of Pigeons,' ibid.. 
vii. 1883, pp. 115, 397. 

3 In his memoir upon Pterocles, quoted below, p. 315, footnote. 


The second pectoral is inserted, in an unusual way, on to 
the anterior face of the humerus, and not on to the pectoral 
ridge, as in other birds. 

In the hind limb nearly all pigeons have the complete 
formula ABXY, the ainbiens being sometimes present and 
sometimes absent. Only in Lopliolcemus apparently is the 
accessory femoro-caudal absent. 

The deep flexor tendons are gallinaceous, either blending 
or with a vinculum. In Lopliolamus au- 
tarcticus and Enjtliroenas there is a vinculum, 
and also a special slip to tendon of digit II. 

Considering the variability of so many 
muscles and organs among the Columba3, it 
is remarkable to note what slight variations 
there are in that, as a rule, rather variable 
organ, the syrinx. The accompanying illus- 
tration will serve to show the form of the 
syrinx among the Columba3 ; but in the 
species illustrated the origin of the sterno- 
tracheales is not so hi 8 hl y asymmetrical as 

latrans (AFTEK i s sometimes the case ; they occasionally 
arise more markedly from the right side of the 
trachea. The intrinsic muscles are always present, and 
generally attached, as shown in the figure, to the membrane 
between the penultimate and antepenultimate tracheal rings. 
They are sometimes continued a little further by ligamentous 
tissue. The last three tracheal rings are united by median 
bony or cartilaginous pieces. Posteriorly the tracheal rings 
are weak, or even defective in the middle line. These are 
the general characters of the windpipe and syrinx in the 
Colmnbae. A few of the genera which show some slight 
divergencies may now be mentioned. In Calcenas nicobarica 
the intrinsic muscles are continued by ligament as far as the 
penultimate tracheal ring ; the first four or five bronchial 
semi-rings are connected posteriorly by a cartilaginous bar, 
which borders the membrana tympaniformis and is continued 
up as far as the sixth tracheal ring from the end of the 


In Goura there is no union posteriorly between succes- 
sive tracheal rings, and the last two or three are quite dis- 
continuous in the middle line posteriorly. 

In Didunculus also the last few tracheal rings do not 
meet in the middle line posteriorly. 

As to the skull, the pigeons are schizognathous birds with 
a slender vomer and basipterygoid processes, absent only in 
Diclus. They are also schizorhinal, but Goura, like Cursorius, 
&c., among the Charadrii, is pseudo-holorhinal. The lacrymal 
fuses below with the ectethmoid, and, indeed, forms with it 
a nearly solid and often rather massive plate of bone. In 
Goura, at any rate, the descending process of the lacrymal 
is perforated in front by a largish foramen, as in the Rhea. 
Some pigeons e.g.Lopholccmus have amedian small circular 
foramen above the foramen magnum ; in Macropygia, &c., 
this becomes a notch upon the upper border of the foramen 
magnum. In Goura the foramen is totally absent. 

The skull of Didundulus is exceptional. The basiptery- 
goid processes are very large. The palatines, instead of 
widening out posteriorly, are narrow, solid bars throughout 
their whole extent. HUXLEY states that the internal lamina 
of these bones is ' altogether obsolete.' I find, however, in 
my specimen a pair of small downwardly directed hooks 
arising from where the palatines come into contact poste- 
riorly, which I take to be the homologues of these structures. 
Owing to the shortened and curved bill the bony nostrils are 
much reduced in extent. There is no supraoccipital foramen. 
There is, as in gallinaceous birds, a fusion between the post- 
frontal process and the zygoma. 

There are 15 cervical vertebra in Goura., Carpophaga, &c., 
14 in Columba, Phaps, &c. Vertebrae 15-17 appear to be nearly 
always ankylosed. 1 The atlas is notched for the odontoid pro- 
cess. The hypapophyses begin in Goura upon the eleventh 
cervical and end upon the first dorsal. Four ribs reach the 
sternum in Goura Victoria, of which the three first bear 
uncinate processes. Only three reach the sternum in some 

iili- NEWTON and GADOW (loc. cit. on p. 314). In Lcucosarcia picata I 
found four fused, and in Geotrygon violacca only two. 



pigeons. The sternum has both spina externa and interna ; it 
has two pairs of notches (Goitra), or one pair of notches and 
one pair of foramina (Phaps, Carpophaga). 

The furcula is U-shaped without hypocleidium. The 
coracoids may meet, but they do not overlap. 

The following table gives the variations of the principal 
variable organs in a number of genera : 

Ambiens C*ca G ^ d 





+ + + 


Turtur + + + 

Old World 

Leptoptila + + + 

South America 

Macropygia . . . + + + 

Australia, East 



+ + + 

North America 


+ + + 

Japan, Fiji, Samoa 

Chamaepelia + + 


Geotrygon . . . + + 

South America 

Metriopelia + + 

South America 

Peristera + + 

South America 

Zenaida + + 


Zenaiclura + + 

North America 

Caloenas + + 

New Guinea, 

China, Nicobar 

Chalcopelia + 



Chalcophaps . . . + + 

Australia, East 


Tympanistria . . . + or 1 + . 


Ocyphaps +0 


Leucosarcia . . . + + 


Phaps . . . + + 


Didunculus + 


Carpophaga . . . + 

+ + 

East Indies, Pacific 


Ptilopus .... 


Australia, Fiji, 


Erythroenas . 




Lopholeemus . 





Treron . 

India, East Indies, 

Africa, Mada- 

gascar, Celebes, 

New Guinea 


New Guinea 



South Africa 

Geopelia . . . : 


Australia, East 


Starncenas ... 

+ 00 


Pblogcenas ... 



East Indies 

Columbula . . . + 


Turacoena . . . + 



Celebes, Timor 


? + 

1 Fiji 

COH iMB/E 313 

GARROD included with the Columbse Pterodes, which I 
treat of separately. Accordingly, in quoting his classification 
of the group, 1 I term families what he termed subfamilies. 
For him the group contained two families, viz. Columbidse 
and Pteroclidae. The following is, with the alterations 
referred to, GARROD'S scheme of division of the group : 

Family I. Columbidae. Columbae with ambiens, caeca, oil 
gland, no gall bladder, and 12 rectrices. 

Genera : Columba, Turtur, Macropygia, Ectopistes, 
L ep top til a , la ntJi'ii as. 

Family II. Phapidae. Columbae with ambiens and no 

Division A. Oil gland present, no gall bladder. 
Genera : CTiam&pelia. 
Zeiiaida, Zenaidura. 

Geotrygon. A^T^ 

Catenas. I* (LIBRARY 

Ch a Ico-pclin. \5^ 




Division B. The oil gland and gall bladder present. 

Genus Carpophaga. 
Division C. The accessory femoro-caudal absent ; oil 

gland and gall bladder present. 
Genus Lopholcemus. 

Division D. Oil gland and gall bladder absent. 

Family III. Treronidse. Columbas without ambiens. 

Division A. With caeca and oil gland ; no gall bladder. 
Genus Phlog<ri///x. 

1 ' On some Points in the Anatomy of the Columbse,' P. Z. 8. 1874, p. 249, 

and ' Notes on Two Pigeons,' &c.. ibid. 1875, p. 307. 


Division B. With cseca, no gall bladder, no oil gland. 

Genus Starn&nas. 
Division C. With oil gland, without gall bladder and 


Genera Geopelia, CEna. 
Division D. Without oil gland (or rudimentary) and 

caeca ; tarsi scutellate. 
Genera Treron, Ptilopus, Erijthrcenas. 
Division E. With casca, oil gland, and gall bladder; 

tarsi reticulate. 
Genus Goura. 

To these must in any case be added another family to 
include the flightless dodo (Didus) and the equally flightless 
solitaire (Pezophaps), 1 the former from Mauritius, the latter 
from Bodriguez. 

The dodo at the time of its description by Messrs. 
STRICKLAND and MELVILLE - was only very imperfectly 
known. Subsequently OWEN 3 gave an account of and figured 
the greater part of the skeleton. Later still Sir E. NEWTON 
and GADOW ' supplemented this account by further details, 
and published a figure of ' the first correctly restored and 
properly mounted skeleton.' There are naturally many 
other notices of this much- written- about bird. 

The skull has the median supra-occipital foramen of 
some pigeons, but not the basipterygoid processes. The 
nostrils are schizorhinal. The palatines have, contrary to 
what is found in Didunculus, with which Sir K. OWEN 
specially compares Didus, an internal lamina. Neither do 
the postfrontal process and zygoma meet and fuse, as they 
do in Didunculus. The interorbital septum is thick and 

There are 15 cervical vertebrae, and the atlas is notched for 
the odontoid process. The last cervical and first two dorsals 

1 A. and E. NEWTON, ' On the Osteology of the Solitaire,' etc., Phil. Trans. 
1869, p. 327 ; E. NEWTON and J. W. CLABK, ' On the Osteology of the Solitaire 
(Pezophaps solitarius),' ibid. vol. clxviii. 1879, p. 438. 

The Dodo and its Kindred. London, 1848. 

:i ' On the Osteology of the Dodo,' Tr. Zool. Soc. vi. p. 49. 

1 ' On Additional Bones of the Dodo,' etc., ibid. xiii. p. 281. 


are fused. Four ribs reach the sternum, of which the last 
belongs, as in Pezopliaps only, to the first pelvic vertebra. 

The scapula and coracoid are, as in theRatites, 1 ankylosed, 
and the angle between them approximates to that of the 
Ratites in its wideness. 

The sternum has a fair keel, neither spina interna nor 
externa, and the coracoids do not nearly meet. The clavicles 
have no hypocleidium. 

I discuss the affinities of the Columbae under Pterocletes. 


Definition. -After-shaft small ; aquincubital. Oil gland nude. A crop 
present. Cseca long. Muscle formula of leg, ABXY -f . Biceps 
slip and expansor secundariorum present ; plantar tendons of 
type IV. Skull schizognathous, holorhinal, with basipterygoid 
processes. Both, carotids present. 

This group contains but two genera, Pterocles and Syr- 
rhaptes, both of which are Old- World in range. Pterocles is 
more widely spread than Syrrhaptes, extending southwards 
to Africa and Madagascar. Syrrhaptes is confined to Central 
Asia. These birds have a pigeon-like aspect, though NEWTON 
has pointed out that Syrrhaptes has a plover-like flight. 

The feet are peculiar for the feathering, which extends to 
the claws ; and in Syrrhaptes the three toes (the first is 
altogether aborted) are encased in a common ' podotheca,' 
which presents the appearance of ' a fingerless glove.' This 
is not the case with Pterocles, which, moreover, has the first 
toe. The sand grouse have a small aftershaft and a nude 
oil gland. The rectrices vary in number from fourteen to 
eighteen. Contrary to what is found in the pigeons the 
newly hatched sand grouse is covered with down, and in the 

1 MOSELEY (' On the Structure and Arrangement of the Feathers in the Dodo,' 
Reji. Brit. AKS. for 1884, p. 782) notes that the feathers are disposed in threes, 
a feature which is, he says, apparent in pictures of the bird. 

2 M. BOGDANOW, ' Bemerkungen iiber die Gruppe der Pterocliclen,' Hull. Soc. 
Imp. St. Pctersb. xxvii. 1881. p. 1(14; H. GAI>O\V, 'On some Points in the 
Anatomy of Pterocles,' &c., P. Z. ,S. 1882, p. 312 ; D. G. ELLIOT, ' A Study of 
the Pteroclidse, or Family of the Sand Grouse,' ibid. 1878, p. 233. 


adult there is down upon the apteria. The pterylosis, as 
figured by KITZSCH, is almost exactly like that of the pigeons, 
but there are no neck spaces. The beak has no soft ' cere,' 
such as exists among the pigeons. 

GADOW has contrasted the crop of the sand grouse with 
that of the pigeon. In the former it is a simple dilatation of 
the anterior and lateral walls of the oesophagus, without any 
constriction in the middle line. In the pigeons, on the other 
hand, the crop consists of two symmetrical swellings of the 
oesophagus, between which is continued the oesophagus. The 
intestinal coils have, according to MITCHELL, an ' extremely 
primitive character.' The resemblance to the intestine of the 
pigeons is great. The Pterocletes differ from the pigeons 
in the large size of the caeca. The lining of the caeca is 
marked by about six longitudinal folds, according to GADOW, 
but according to PARKER no less than twelve. In Pteroclex 
the right lobe of the liver is about three times the size of the 
left. A gall bladder is always present. Figs. 17 and 1<S 
(p. 33) show some variations in the positions of the liver and 
pancreatic ducts in the sand grouse, which are taken from 
GADOW T 'S paper upon the anatomy of this group. 

The syrinx (of Pterocles) is not in the least like that of 
the Columbae. The extrinsic muscles are perfectly symmetri- 
cal, and the intrinsic muscles are enormously developed. The 
ordinary pair present in the Columbae aue attached to what i 
regard as the first bronchial semi-ring, and are not specially 
large ; the second pair l are only visible on the posterior 
aspect of the windpipe ; they are two large fusiform muscle's 
which are inserted in common into the middle line of the 
trachea, near to its termination. 

The muscles of the hind limb have been described in some 
detail by GADOW. 

The muscle formula is complete, being thus expressible, 
on GARROD'S notation, by the letters ABXY + . There is 
only one peroneal, the longus, which has the usual attach- 
ments to the ankle and to the flexor perf or atus. 

1 Pterocles arcnarius ; it appears that P. alchata has not the second larger 


The deep flexor tendons fuse at the ankle, but no branch 
is given off to the small hallux ; this digit has, however, a 
special flexor hallucis brcvis. GADOW concludes his survey 
of the muscles of the bird with the following remarks : ' Of 
all the other muscles of the leg (excluding ambiens, peroneus, 
and absent flexor hallucis slip) there is none that shows any 
practical difference between sand grouse, pigeons, and even 
(if we include them in our comparison) the plovers. On the 
whole, however, the myology of Pterocles indicates that it is 
more nearly allied to the pigeons than to any other group of 

The general aspect of the skull of Pterocles arenarius is 
much like that of a similarly sized pigeon. The nostrils, 
however, are more distinctly holorhinal, thus leading towards 
the gallinaceous birds. They end on a level with the ends of 
the nasal processes of the premaxillaries, and do not narrow 
at all at their broadly rounded terminations. As is the case 
with Goura, Opistliocomus (a fact of possible importance), and 
some other holorhinal birds, a plate of bone underlies the ex- 
tremity of the nostrils, reducing the extent of the orifices. 

As in pigeons the ectethmoids are very solid plates of bone 
which fuse with the lacrymals, and nearly reach the jugal 
bar ; only the minutest foramen perforates this plate above. 
Syrrhaptes l agrees so far with Pterocles, but has a rather 
more vacuolated interorbital septum. 

The post-orbital and post-frontal processes are long, and 
nearly (Pterocles) or quite (Syrrhaptes) fuse at their extremi- 
ties, as in some gallinaceous birds. 

The niaxillo-palatines are not like those of pigeons ; they 
are, as in gallinaceous birds, slender curved hooks. 

The basipterygoid processes are well developed. 

There are fifteen or sixteen cervical vertebrae. 

' In almost all those respects,' remarked HUXLEY, ' in which 
the grouse differ from the fowls they approach the pigeons ; and 
an absolute transition between these groups is effected by the 
Pteroclidtp, whose popular name of " sand grouse " might fitly 

1 See PARKER, ' On the Osteology of Gallinaceous Birds,' &c., Trans. Zool. 
Soc. vol. v. 


be exchanged for that of " pigeon grouse." There is, in my 
opinion, much to be said for this view, which, however, is 
not now so generally held. HUXLEY'S view was based 
almost exclusively upon osteological characters, with but 
slight reference to the anatomy of the soft parts, which were 
indeed when he wrote (in 1868) scarcely known. The 
several regions of the vertebral column in the sand grouse 
have the same number of vertebrae as in the ' Alectoromor- 
pha?,' ' and ankylosis takes place in the same manner.' The 
skull is dove-like for the most part ; but in. certain ways it 
approaches the Galli. For instance, the maxillo-palatines are 
alike in both groups of birds ; the union between the squa- 
mosal and the post-frontal process is gallinaceous ; the 
holorhinal nostrils, which I must term those of Pt erodes, 
indicate a likeness to all the members of the group Galli. 
The remainder of the skeleton is, in HUXLEY'S opinion, 
' peristeromorphous,' but the pelvis is partly grouse-like. 
Attention may be directed to the likeness of the sand 
grouse humerus to that of pigeons. The osteological cha- 
racters, however, are not quite so intermediate in some 
respects as might be inferred from HUXLEY'S paper. The 
at least ' pseudo-holorhinal ' nostrils have their counterpart 
among the Limicolse, in Thinocorus, and in some others (see 
below). The solid ectethinoids too are also seen in that 
group, while GAEEOD'S remark that the Alcae have a hu- 
merus like that of Columbae and Pterocles is suggestive in the 
light of the unquestionable likeness of the Alcae for the Limi- 
colae, though the actual weight of this character may be 
thought by some to be discounted by the fact that it is met 
with in the Psittaci. 

Moreover Otis, which is to be placed somewhere near the 
Limicolse, has the gallinaceous union between the squamosal 
and the post-frontal process, to which I have referred as 
possibly affining the Pterocletes to the Galli. Other characters 
too, which appear at first sight to be arguments in favour of 
the position taken up by HUXLEY, may be interpreted fairly 
as marks of affinity with the Limicolse (and their immediate 
allies). Such are, for example, the long ca?ca (with folds in 


the bustards), the crop (present in Thiiiocorus), the gull 
bladder, &c. MITCHELL distinctly places both Pterocletes 
and Columbae in the neighbourhood of the Limicolae by reason 
of the arrangement of the intestinal coils. 

It is at any rate clear that the Pterocletes occupy a lower 
place than the Columbae that the// have given rise to the 
Columbse, and not vice versa. The justice of this view is 
shown by the long caeca, the existence of an aftershaft, the 
complete muscle formula of the leg, and by a few other 
equally unmistakable characters. On the whole it seems 
not unreasonable to look upon the Pterocletes as not far from 
the stock which produced the Limicolse, which itself was 
possibly not far again from the primitive gallinaceous stock. 


Definition. Rectrices, twelve. Aftershaft present. Oil gland tufted. 
Caeca long. Muscle formula of leg, A(B XY+. Skull aegitho- 
gnathous, schizorhmal, "with basipterygoid processes. Cervical 
vertebrae, fifteen. Sternum one-notched. 

This group of birds consists of the genera T/iniix and 
Pedionomus. 1 It has been confounded with the gallinaceous 
birds ; but the discovery of PAEKEE that the skull is aegitho- 
gnathous, and further investigations into the structure of 
the group of which the most important is a recent paper 
by GADOW 2 have rendered it necessary to remove the two 
genera from close association with the Galli. 

Of the two genera Tuniix (Hemipodius) is European, 
African, and Indian in range ; Pedionomus is Australian. 

Besides the external characters mentioned in the defini- 
tion, which are common to both genera, Pcdionoiniix wants 
the fifth cubital, which is present in T-unii.r ; there are four 
toes in Pedionomus ; Turnix has lost the small hallux of 
the former genus. 

1 LEGGE (P. Z. S. 1869, p. 236), from a consideration of some external cha- 
racters and habits, was impressed by possible charadriine affinities of Pedio- 

- 'Notes on the Structure of Pcclionu/nitx tur^ualtis^ Ac., Records Austral. 
J/u.s. i. 1891, p. 205. 


As to the pterylosis, there is a long spinal apterion, which 
begins on a level with the shoulder joint and reaches to a 
little beyond the level of the hip joint. Thence the two 
dorsal tracts are continued on as a single tract to the 
feathered oil gland. On the neck below there is no apterion ; 
the two tracts then divide, leaving a bare interclavicular 
space ; they divide again on a level with the anterior end of 
the carina sterni into a lateral thick patch and a median 
thinner one ; this latter swells out in its course and then 
again dwindles, being continued to the cloacal aperture by a 
few scattered feathers. 

The pectoralis I. (at any rate in T. Sykesi) is two-layered. 
The tensor patagii brevis tendon gives off a wristward slip, 
but there is no patagial fan. 

There is no biceps slip, but the expansor secundariorum 
is present (? as to both these structures in Pedionomus}. 
The muscle formula of the leg is the complete one ABXY + 
in Pedionomus ; Turnix has lost the accessory femoro-caudal 
for the most part not, however, in T. Kleinschmidti, where 
it is present. It is remarkable that in Pedionomus it is not 
B but A which is on the wane. 

Both carotids are present in Pedionomus, only the left 
in Turnix ; but in Pedionomus the left is the weaker and 
not the right, as might perhaps have been suspected. 

The alimentary canal has no crop, ' but the upper half 
of the oesophagus is very dilatable ' (in Pedionomus). The 
ccBca are well developed ; the liver in both genera is split 
into three nearly equisized lobes. The gall bladder is 

The syrinx (of Turnix lepurana) is not at all gallinaceous 
in its characters. The tracheal rings are weak and carti- 
laginous. The intrinsic muscles are thick and originate in 
close contact from the anterior face of the trachea ; they are 
inserted some way down the bronchi on to the opposite face 
of the tubes. In Hemipodius tacliijdromus the windpipe is 
very soft, and is much dilated in front of the origin of the 
intrinsic muscles, which, as in the last species, are large. 

Our knowledge of the osteology of the Heinipodes is 


chiefly due to PARKER, 1 who has described in detail the 
entire skeleton of Hemipodius varius, and also the skull of 
Ticrnix rostrata. The atlas is perforated for the odontoid. 
There are 15 cervical vcrtcbrcc, of which Nos. 10-15 bear 
haemapophyses. None of the dorsals are ankylosed. The 
sternum is reached by three or four ribs, and has one pair of 
deep lateral incisions cutting off two long thin postero- 
lateral processes. 

The skull is aegithognathous in its vomer, broad in front, 
and double posteriorly, and in the slender hook-like maxillo- 
palatines ; the latter, however, are not unlike those of many 
gallinaceous birds, while HUXLEY has compared the vomer 
with that of Tetrao urogallus. The nostrils are pseudo- 
holorhinal, and, as in pigeons, there is a considerable alinasal 
ossification, reducing the long nares, which are perfectly 
pervious. As is also the case with the pigeons, the ecteth- 
moids are large and solid, and have fused with the lacrymals. 
There are well-marked basipterygoid processes. 


Definition. Aftershaft usually present. Carotids, two. Muscle 
formula of leg, ABX(Y) + . Expansor secundariorum always 
present. Tensor patagii brevi3 without recurrent slip to tensor 
patagii longus. Caeca long. Skull schizognathous and holo- 

The rails are a group of birds of very uniform structure. 
They have as a rule a tufted oil gland, but Porzana Carolina, 
is an exception. The aftershaft is present. The rectrices 
vary in number from 10 (Aramides cay omens is) through 12 
(Porzana Carolina} to 14 (Ocydrmnt/x Earlei). 

The spinal tract encloses a long narrow apterion, which 
commences earlier in Rail its aquaticns than in Fulica atra. 
The latter bird has almost a gap between the anterior and 
posterior parts of the spinal tracts. The pectoral tract of 

1 In his papers on the osteology of gallinaceous birds and of the scgitho- 
gnathous skull in Zool. Trans, vols. v. ix. x. 



each side is double in RaUus and Ocydrowws, single in 

The following are the intestinal measurements of a series 
of species : 

Small Int. 

Large Int. 

Rallus aquaticus 
Ocydromus sylvestris 
O. lefresnayanus 

0. Earlei . 
Aramides cayennmsis 
Porzana Carolina 
Crex pratcnsis , 
Porphyrio madagascariensis 
fiaUinula chloropns . 
Fulica atra 
F. ardesiaca 
Tribonyx Morticri 



















2-25 ' 


















li, (i'75 

The folds of the intestine (fig. 159) are remarkably like 
those of the cranes (fig. 158), so much so that on intestinal 
characters only the two groups could not be separated. 

There is no crop ; the proventriculus is zonary ; the 
stomach a ' gizzard.' The right lobe of the liver is larger 
than the left, and the gall bladder is always present. 

The atlas is notched for the odontoid process ; it has no 
lateral canals. The number of cervical vertebrae is 15 in 
Fulica ardesiaca, in which there are 7 complete ribs (6 in 
Ocydromus}. On the eleventh cervical (of Fulica ardesiaca) 
the catapophyses nearly unite ; the hsemapophyses, up which 
the catapophyses do not climb, extend as far as the second 
dorsal. The sternum has very long lateral processes, with a 
larger or smaller spina externa. The clavicle comes into 
near relations with both procoracoid and scapula. In the 
skull 2 there are no basipterygoid processes, and the lacry- 
mals (in Fulica, Ocydromus, and Aramides) do not join the 

1 In Porzana notata the caca are minute -8 inch in length. Cf. Pamelas 
among Limicolse. 

- C. G. GIEBEL, ' Osteologie der gemeinen Ealle,' &c., Zcitschr. gcs. Naturw. 
v. (1855), p. 185. SHUFELDT, ' Osteology of certain Cranes, Rails, and their 
Allies,' J. Anat. Phys. 1895, p. 2; and 'Osteology of Porxana Carolina,' 
Journ. Comp. Med. Surg. 1888. 



ectethmoids, as they do in all charadriiform birds. The 
ectethmoids themselves in all rails that I have examined 
send a process upwards, which joins the frontal bone and 
leaves a foramen for the passage of nerves. The interorbital 
septum is widely fenestrate. 

The pelvis in the rails has a longer preacetabulum than 
postacetabular portion. The ilia are vertical in their plane 
anteriorly, and in Triboiujx and Fidica are completely sepa- 

FIG. 158. INTESTINES OF Cun/inui 
cristata (AFTER MITCHELL). 

.r, short-circuiting vessel. 

FIG. 159. INTESTINES OF Crex pra- 
te nsis (AFTER MITCHELL). 
x, as in fig. 158. 

rated from each other by the fused neural spines of the 
vertebrse. In Aramides and Ocydromus, on the other hand, 
the ilia reach the summit of those neural spines. In all 
these rails the pubes are fairly strong bones, which are not 
ankylosed anywhere with the ischia. 

Nearly all the Ballidae have a biceps slip. Ocydromus 
Earlei and Rallies macidatus are the only exceptions known 
to me. The tensor brevis is simpler than in many birds ; 
in Ocydromus Earlei it consists of only a single tendon. In 
Crex, as in most others, this tendon gives off a wristward 
slip. In Aramides and Porplujrio martinicus the hinder 
branch of the tendon is very feeble, and in the latter does 
not reach the fore arm. In no rail is there any distal 

Y li 


patagial fan, a fact justly emphasised by FOEBES ! in dis- 
criminating from the rails the somewhat rail-like Parr a. 
In some (e.g. Oci/droinus Earlei, Aramides} the tendon of 
the tensor brevis does not run over the arm to the ulnar 
side, but in others (e.g. Crex) it does. 

The (meaner us has a humeral attachment. 

FUEBRINGEE figures in Porphyrio an interesting con- 
dition of the biceps and of the biceps slip. The biceps slip 
arises by a tendinous head close to, but apparently inde- 
pendently of, the humeral head of the biceps. Close to the 
coracoidal head of the biceps springs a ligament which is 
inserted on to the hunierus just in front of the origin of the 
humeral head. This ligament seems to be a detached 
portion of the biceps, since in the Steganopodes (q.v. ; cf. 
also FUEBRIXGEB, pi. xxvi.) it is perfectly continuous with 
the biceps. 

In the leg both peroneals are present, with the usual 

The deep flexor tendons are often of No. I. type. But 
in Aramides there is a modification of this in the shape 
of a second vinculum, attached partly to the tendon just 
before its trifurcation and partly to the special tendon of 
digit II. In Ocydromus Earlei the second vinculum is also 
present, but feebler, not having the second attachment to 
the flexor of digit II. 

The syrinx in the rails is of a quite typical tracheo- 
bronchial form, except for the fact that the intrinsic muscles 
are attached rather far down the bronchi to the fourth 
bronchial semi-rings in Ocydromus, Aramides, and some 
others. In Ocydromus sylvestris none of the tracheal rings 
are fused ; the first two bronchial semi-rings are ossified, 
and there is an ossified pessulus. In Aramides the last 
three tracheal rings are partly fused ; the last two of these 
and the first two bronchial semi-rings are ossified ; there is 
no ossified pessulus. In Fulica (ardcsiaca and leucoptcra] 
the bronchidesmus, which is incomplete, is at first strengthened 

1 ' Notes on the Anatomy and Systematic Position of the Javanas,' P. Z. ,S'. 
isx],p. 639. 


by two yellow elastic pads of tissue springing from the 
membrana tympaniformis. 

The genera Heliornis and Podica the former American, 
the latter Old- World in distribution seem to require a sepa- 
rate family for their reception. The structure of these two 
birds has been mainly investigated by myself ; l the skeleton, 
however, has been described by BRANDT also. 2 

In neither bird is there an aftershaft, though the oil 
gland is tufted. Unlike other rails they are quintocubital. 
The ptenjlosix is essentially ralline. In Hcliornis the neck 
is nearly continuously feathered, there being only a short 
vental apterion. The dorsal tract is strong between the 
shoulder blades, and is forked ; the hinder parts of the tracts 
scarcely join the anterior ; they become fused some way 
in front of the oil gland. In Podica senegalensis, but not in 
Hcliornis, the ventral tracts are undivided. The less degree 
of specialisation is seen in other features of the anatomy of 
the smaller American finfoot. 

The muscle formula of the leg in both genera is ABX + , 
the Y of the rails not being developed. The chief peculiarity 
of the leg muscles, however, concerns the biceps. This is a 
very large muscle ; in Podica it has no less than three 
separate insertions on the leg. First there is the ordinary 
insertion through a perfectly normal sling ; just before this 
tendon a branch is given off which is inserted independently 
on to the leg some way further down. In addition there is 
an extensive insertion on to the fascia covering the calf of 
the leg. In Heliornis the muscle is somewhat simplified, 
only the first and third insertions being present. The 
complications of the biceps may have some relation to 
swimming; for in certain auks (q.v.) there is a similar 
gastrocnenial attachment. 

1 ' On the Anatomy of Podica senegalensis,' P. Z. S. 1890, p. 425 ; ' On the 
Osteology, Pterylosis, and Muscular Anatomy of the American Finfoot (Heliornis 
surinamensis),' Ibis, 1893, p. 30. 

2 ' Beitrage zur Kenntniss der Naturgeschichte der Vogel,' etc., Hem. Ac. 
Sci. St. Petersburg, 1840, p. 81. See also GIEBEL, ' Zur Naturgeschichte des 
surinamischen Wasserhuhns Podoa surinam-ensis,' Zeitsclir. yes. Natunr. xviii. 
1861, p. 424, and NITZSCH'S Ptcrylograplnj. 


The deep flexor tendons of Heliornis are remarkable (see 
fig. 160) for the fact that both tendons split into three branches 
for the three digits before they unite, the slip to the hallux 
being given off from the flexor hallucis previously. The 
conditions which characterise Podica are unknown. In the 
fore limb some of the muscles are characteristic, and differ 
from those of the rails. The patagialis brevis consists in 

OF Heliornis (AFTER BEDDABD). 

A, fl. hallucis : B, fl. communis ; 1-4, 
slips to digits. 

Heliornis (AFTER BEDDARD) 

Tp, tensor patagii ; Hi, biceps ; Bi.s, 
biceps slip ; x, tendinous slip. 

both of a simple undivided tendon, which has not more than 
an indication of a patagial fan. This indication is seen only 
in Heliornis (fig. 1(31) in the shape of an upwardly directed 
tendinous slip, to which, a little before its termination 
apparently upon the patagium a well-developed biceps slip 
is attached. In Podica the biceps slip ends freely upon the 
patagium. Both these conditions are different from what is 
met with in the rails, but are to some extent paralleled 
among the grebes. In Podiceps the biceps slip ends freely 
upon the patagium, as in Podica. In ^Eclimopliorus (p. 387) 
the biceps slip is connected directly with the patagial fan. 
The likeness between that bird and Hclioniix appears to me 
to be unmistakable. 

The expansor secundariorum is present, and the anconams 



is anchored to the humerus. The finfoot agrees with the 
divers in the extensive origin of the posterior latissimus dorsi 
from the front end of the ilium. The largely tendinous 
origin of the rhomboidei appears to point in the same direc- 
tion. As to the alimentary viscera, there are long caeca ; the 
right lobe of the liver is the larger ; a gall bladder is stated 
by GIEBEL to be present in Heliornis. I did not find one in 
Podica, but the matter doubtless requires re-examination. 

FIG. 162. SYRINX OF Podica 
xenccjalensis (AFTER BEDDAKD). 

FIG. 163. STERNUM OF Heliornis. 

The syrinx is typically tracheo-bronchial, and in no way 

The osteology of the Heliornithidae is not very decisive as 
to their affinities. In Podica there are seventeen cervical 
vertebra, an advance upon the fifteen of the rails and an 
approach to the twenty-one of Podiceps. Six ribs reach the 
sternum in both genera of Heliornithidae. The sternum has 
but one pair of notches, and in Podica, at any rate, the spina 
externa is well developed. The skull is, on the whole, rail- 
like, bearing, perhaps, a greater resemblance to Aramides 
than to any other genus of rails. The clavicles, contrary to 
what is met with in the rails, reach, and are firmly attached 
to, the carina sterni. 

/.s (fig. 164) of the Heliornithidae is in some respects 




unlike that of the typical rails. As in Fill lea, the ilia are 
widely separated by the fused neural spines of the dorsal 


vertebrae concerned ; but the ischia are broader and directed 
more downwards (their position is, in fact, more primitive) 

FIG. 16(5 SKULL OF Heliornis. 


I'"i<i. 107. SKULL OF 


than in the rails, while the pubes are ankylosed at least at 
one point with the ischia. 

There are a considerable number of extinct rails, many 
of which were flightless, thus showing an exaggeration of a 


tendency of many existing rails which either do not fly 
much or are hut feebly fitted for flying. 

It is particularly upon small islands that these flightless 
rails have been discovered, both living and fossil, and in 
islands where loss of flight may be regarded as having been 
of less importance as a disadvantage in the struggle for 
existence. Thus in New Zealand there was (until recently) 
the large Notornis, whose skeleton has been described by T. J. 
PARKER. 1 The last living specimen was taken in 1879. From 
the Chatham Islands are known Paltzolimnas cliatliamensis 
and Nesolinmas Dieffenbachii. The skeletons of these rails 
have been described by H. O. FoRBES, 2 MiLNE-EDWARDS, 3 
and ANDREWS ; 4 the latter by ANDREWS. Nesolimnas ap- 
pears to be not yet extinct ; the former species is. Another 
form from the Chatham Islands was originally described 
under the generic name of Aplianaptenjx, and supposed to 
be congeneric with A. Broecki of Mauritius. Both of these 
birds have been dealt with by MILNE-EDWARDS and ANDREWS. 
Diaphorapieryx Hawkinsi was a largish rail, with the keel of 
the sternum much reduced, being about half the height of 
the keel of the flying Hypotcenidia celebensis and slightly 
less than that of Ocydromus. The scapula and coracoid 
make an exceedingly wide angle, as in all flightless birds 
about 130 degrees. The resemblances of Diaphorapteryx to 
Aplianapteryx are set down by ANDREWS and GADOW to 
parallelism of development, and not to real affinity. 

Palceolimnas chatliamensis is not, as was at one time 
thought, identical with Fulica Newtoni of Mauritius ; it may, 
however, be the same as Fulica prisca of New Zealand. 
The bird is much like Fulica in osteological characters, the 
principal difference being the large size of the impressions 

1 ' On the Skeleton of Notornis Mantclli; Tr. N. Zeal, List. xiv. (1881), p. 

- In Nature, xlv. (1892), p. 410 ; ibul. p. 580 ; and Ibis, 1893, p. 254. 

:t ' Sur les Ressemblances qui existent entre la Faune,' &c., Ann. Sci, Nat. 
(8), ii. 1896, p. 117. 

4 ' On the Extinct Birds of the Chatham Islands,' parts i. and ii. ; Novitates 
Zool. iii. p. 73 ct .sr^ ., p. 260 ct seq. ; and ' Note on the Skeleton of Diapltor- 
apteryx Haickinsi,' Gcol. Mag. 1896, p. 337. 


for the supra-orbital glands. The keel is reduced as com- 
pared with living coots, its height being 12 mm. as compared 
with 17 mm. and 15 111411. in F. atra and F. cristata. The 
wing is short in proportion to the leg, shorter than in F. atra ; 
but ANDREWS thinks that the bird ' may still have been 
capable of heavy flight for short distances.' 

Ncsolimnas is a more aberrant form in some particulars. 
It may be still living, but the only specimen was obtained in 
1840. The most striking feature in the osteology of this 
bird appears to be the schizorhinal nostrils, which do not 
occur elsewhere in the rails (as defined in the present volume). 
The wings are reduced, but the scapula and the coracoid do 
not make a wide angle (forty-five as against sixty for Ocy- 

Of the extinct European rails described by MILNE-EDWAKDS, 
from the Eocene and Miocene, a number of species have been 
described and referred to the genus Rallus. Gypsornis is believed 
to be most nearly akin to Aram ides. 


Definition. Three-toed birds. Oil gland absent. Aftershaft present. 
Aquincubital. Skull schizognathous, holorhinal, without basi- 
pterygoid processes. Muscle formula, BXY + . Caeca long. No 
biceps slip. Expansor secundariorum present. Syrinx without 
intrinsic muscles. 

THE bustards are undoubtedly a much-specialised group, not 
(in my opinion) distinctly nearer to the charadriiform birds, 
where they are placed by FURBEINGEE, than to the cranes, 
with which they are associated by GADOW. 

They are distinguished from all their allies by the total 
absence of an oil gland. The feathers have an aftershaft. 
There are twenty rectriccs in Otis and Houbara, sixteen in 
Eupodoiis Denhami, eighteen in Tetrax. 

In both Otis and Tetrax the lateral neck spaces are 
reduced to a rudiment on each side close to the shoulder. 
The dorsal tract is divided high up on the back of the neck ; 
the two halves come nearer together, and at the same time 


get broader in the lumbar region ; they finally completely 
fuse to form a broad and uninterrupted tract. 

The ventral tract is undivided in the neck in Otis ; it is 
very broad in the pectoral region, where it divides into two 
narrow bands, with an indication of a third in the form of a 
slightly divergent outer group of particularly strong feathers. 
The two pairs of narrow tracts unite round the cloaca. 

In Eupodotis australis the dorsal tract is divided high up 
on the neck, the ends of the tracts at ends of scapula are 
particularly strongly feathered, arid the two halves of the 
posterior region of the dorsal tract run in anteriorly between 
the anterior forks. The ventral tract has a fainter indication 
than in Otis of the outermost branch. It does not appear 
again to divide into two. 

A striking peculiarity of the bustards is the variability of 
the carotids. In Otis and Honiara there are two ; \i\Eitpo- 
dotis only the right ; in Tetrax only the left. 

The following are intestinal measurements :- 

Small Int. 

Large Int. 

i 8BC8 




Otis tarda 



Eupodotis a list ml in 



13-5 and 14 

E. kori . 




E. arabs . 



9-5 and 8-75 

E. Denhami . 



7-5 and 9 

Houbara undiilata 




18 4 


H. Macqueeni . 

36 4-5 


Tetrax campestris . 



The caecum of Otis tarda is highly remarkable. The 
median third of the gut is much dilated, and is lined by a 
smooth mucous membrane, which is marked by about seven 
slightly raised longitudinal folds, which are visible externally, 
but connected with no sacculation. Scattered about are 
numerous circular glands, of the size of hemp seed. The 
terminal part of each caecum, which is some two inches in 
length, has villi, like intestine. The apical region has not, 
but there is a close retiforrn disposition of mucous membrane, 
which gradually passes into longitudinal folds of the middle 

< HIDES 333 

region. The ca?ca of Eupodotis Denhami appear to be much 
the same. 

In Eupodotis a astral is the caeca are not dilated only in 
the middle, as are those of Otis tarda, but for the apical 
eleven or twelve inches or so. The dilated region is lined 
with irregular folds. E. Denhami agrees with 0. tarda. The 
liver lobes are equal, or (Tetrax) the right is the larger. A 
gall bladder is present. 

The bustards (at any rate Eupodotis Denhami) are remark- 
able for possessing a rudimentary penis, as does the perhaps 
nearly allied (Edicnemus. It is a short blunt cone, grooved 
above, with a row of glandular pores on each side below. 
Internally there are two oval spongy bodies, attached to 
anterior part of sphincter muscle, and external to these on 
each side are retractor muscles, attached to back of cloaca. 

The bustards have a tracheo-bronchial syrinx, but the 
intrinsic muscles are either absent or, if present, are but 
feebly developed. 

Eupodotis australis has perhaps the least modified 
syrinx. The first two or three bronchial semi-rings seem 
really to belong to the tracheal series, on account of their 
greater depth and slighter dividing membranous intervals 
than those which follow. The intrinsic muscles are reduced 
to a narrow ligament, fanning out somewhat below. 

In E. kori the ligament representing the intrinsic muscle 
of each side is even feebler, and in E. Denhami it has abso- 
lutely vanished. 

In all the above species the rings and semi-rings preserve 
their independence, and are not fused, except one or two 
ventrally to form the pessulus, which is strong and ossi- 

In Otis, on the other hand, the pessulus is slender and 
cartilaginous, being formed by one ring only. There are no 
traces of intrinsic muscles. 

The genus Honiara has a rather peculiar syrinx, which, 
however, like the last, is without intrinsic muscles. It is 
compressed from side to side just before the bifurcation. 
But the ' waist,' thus formed does not correspond to the 


boundary line between trachea and bronchi ; it lies between 
the penultimate and antepenultimate tracheal rings. 

In front of the last three tracheal rings there is no par- 
ticular modification of the trachea. The antepenultimate ring 
is strongly ossified in front, where it is convex downwards, 
thus leaving a considerable membranous interval between 
itself and the ring in front. The next ring is of the same 
size, and also ossified in front ; the tough and elastic mem- 
brane uniting the two can be easily stretched. The last 
tracheal ring is much narrower, but also ossified in front ; it 
passes into the cartilaginous pessulus. Posteriorly these 
rings are incomplete, but are joined by a particularly tough 
membrane. The bronchial semi-rings are delicate, and not 
so long (from before backwards) ; they naturally diminish 
successively in length. The above description refers to H. 
Macqueeni, but A. undulata hardly differs. 

The bustards exhibit a phenomenon known as ' showing 
off,' which is associated with certain anatomical peculiarities. 
The appearance of the male bird, when indulging in this 
display, is illustrated by a plate which accompanies Dr. 
MUEIE'S paper * upon the subject. The neck is immensely 
puffed out, so as actually to trail upon the ground. This 
singular behaviour on the part of the cock bird during the 
breeding season is not confined to the European Otis tarda ; 
it has been observed in both Eupodotis australis and E. 
Denhami. It is curious that, though the result to all out- 
ward appearance is much the same, the mechanism which 
produces the inflation of the neck differs in the two cases. 
In Eupodotis the anterior section of the oesophagus becomes 
dilated. In Otis there is a special pouch 2 developed between 
the two halves of the lingual frenum, which extends for a 
considerable way down the neck. 

The tensores patagii are fairly characteristic. No bustard 

1 P. Z. S. 1868, p. 471 ; Sir W. ELLIOT, 'Notes on the Indian Bustard,' &c., 
P. Z. S. 1880, p. 486. See also FLOWER, P. Z. S. 1865, p. 747 ; NEWTON, Ibis, 
1862, p. 107; MUKIE, P. Z. S. 1869, p. 140; GARROP, ibid. 1874, pp. 471, 
673 ; FORBES, ibid. 1880, p. 477. 

2 Cf. similar pouch in duck Biziura, v. p. 458. 


has a biceps slip. In Eupodotis Denliami the brevis tendon is 
a broad fibrous band spreading out after the ulnar muscles 
and inserted on to humeral tubercle. In Ei/podotis <utxtr<ilix 
and Houbara Macqueeni there is, in addition, a broad wrist- 
ward slip which does not cross the fore arm. In Otis tarda 
the extreme degree of complication is reached, for there is, 
in addition to the structures described, a slight patagial fan 
joining the longus tendon in the usual way. 

The anconcBits has a tendinous humeral head (at least in 
Eupodotis australis and Otis tarda). Otis has no latissi- 
mus dor si posterior. 

As in other three-toed birds, the deep flexor tendons are 
completely blended. 

The glutens max i in its is large and quite covers the 

The number of cervical vertebrce is sixteen, seventeen 
(fide GADOW and FURBBINGER), or eighteen (Houbara Mac- 
queeni). The atlas is notched. In Houbara Macqueeni at 
any rate the eleventh to thirteenth cervicals have closely 
approximated haemapophyses. The last haemapophysis is on 
the Dl ; in the two vertebrae in front these processes are trifid. 

Five ribs reach the sternum, all of them with uncinate 
processes. The sternum has two notches. There is neither 
spina extema nor spina interna. 

In the skull the margins of the orbit are very sharp, as in 
(Edicnemus and Rhinochetus. The interorbital septum is 
riot greatly fenestrate. The descending process of the lacry- 
mal just comes into contact, but does not ankylose, with the 
pref rental process of the ethmoid. The maxillo-palatine 
processes are curved and shell-like. In Houbara Macqueeni 
at any rate this bone reaches the jugal arch. The temporal 
fossa is guarded by two long and spine-like processes of the 
squamosal bones, as in gallinaceous birds, and much more 
marked than in Rliinoclietus. 

The procoracoid is of moderate size and does not reach 
the clavicle. The two coracoids are not in contact at their 
articulation with sternum. 

I place the bustards in a group by themselves, largely 


on account of the fact that they are in several respects much 
altered by modification from their allies. They show 
evidence of degeneration in the loss of the oil gland, in the 
occasional loss of one of the two carotids, in the absence of 
the biceps slip, and in the reduced muscle formula of the 
leg. GAEEOD associated with the bustards the Cariamidae, 
GEdicnemidse, Serpentarius, and possibly Phosnicopterus. 
There is, in my opinion, more to be said in favour of 
associating the first two families with the bustards than the 
last two. But, as I have pointed out elsewhere, Serpentarius 
shows more than one hint of a crane-like origin. As to the 
first two groups, they agree with the bustards in the muscle 
formula BXY, in the holorhinal nostrils, in the absence of the 
biceps slip (Cariama), the absence of basipterygoid processes, 
the absence or feeble development of intrinsic muscles to the 
syrinx ; the oil gland too, absent in the bustards, is nude in the 
cariamas, and thus shows a commencing reduction. But these 
various cases of reduction cannot be held to be necessarily 
indications of relationship. I should, however, lay some stress 
upon the holorhinal nostrils, the leg muscles, and the syrinx ; 
in this case the same conclusion as that advanced by 
FUEBRINGEE is arrived at, viz. that the Otides come nearest 
to the (Edicnernidae. The very difficulty of associating the 
Otides with either gruiform or charadriiform birds is evidence 
of the common descent of all three divisions of the class. 


Definition. Oil gland feathered. Aftershaft present. Aquincubital. 
Skull schizognathous. Both, carotids present. Cccca nearly 
always large. Ambiens- always present. Biceps slip to patagium 
nearly always present. 

This is a large group of birds which are cosmopolitan in 
range and embrace a variety of types, which may perhaps 
be arranged in six families. The type family is that of the 

1 SEEBOHM, The Geographical Distribution of the CharadrHdce > &o,,'Lond.on, 
1887. A monograph of all the species (excl. gulls). 

- Rlujnchops is alone exceptional in having no ambiens. 

L1M1COL.E 337 

Characlriiclae, which contains the largest number of genera ; 
the remaining families are not separated from it by very 
numerous points of difference, and the group as a whole is 
very near to the gulls, which I only divide as a family. The 
birds of this group, though they are generally good flyers, are 
mostly found upon the margins of the sea or of marshes and 
pools ; and their long bills are apparently constructed with a 
view to probing the mud and sand of such localities for their 
food, which is, w r ith the exception of the vegetable-feeding 
Thinocoridse, animal. The bill is usually long, and, in the 
woodcock, soft at the extremity, reminding us of the bill of 
Aptenjx, being used, indeed, for the same purpose, to extract 
earthworms. In the curlew (Numenius) the bill is curved 
downwards, as in the ibis. In the avocet (Recurvirostra 
avocetta) it is curved upwards ; in Eurynorhynchus it is 
spatulate at the extremity, and, finally, in the crooked-billed 
plover it is bent sideways. The legs are often long, and the 
toes moderately or very much so (Parridse) . There are either 
four toes, the hallux being small, as in the whimbrels, 
pratincole, &c., or the hallux and the remaining toes also are 
of enormous length, as in the Parridse only ; in many forms, 
such as the stilt plover, the hallux is absent. In Recurvi- 
rostra and Himantopus andimis the feet are well webbed. 
In the phalaropes the feet are lobate. The colour of these 
last-mentioned birds is suggestive of that of the mature gulls, 
just as the markings of the immature gulls is suggestive of 
the coloration of many Limicolae, such as the dunlin, knot, &c. 
The number of rectrices varies from ten in Rhynchcea and 
twelve in EurynorJiyncJitis to as many as twenty-six in 
Scolopax. The face in Lobivanellus is adorned with fleshy 
lobes, so often found in birds. 

The pterylosis of the Limicolae has been chiefly studied 
by NiTZSCH, 1 who figures Scolopa.c and Charadrius. The dor- 
sal tract, single on the neck, bifurcates between the shoulders 
into two strong bands, which either are (Scolopax) or are not 
(Charadrius) continuous with the anteriorly bifurcate pos- 
terior section of the dorsal tract. The ventral tract divides 

1 See also ANDKESON as quoted on p. 343. 





at the beginning of the neck ; on the breast each half gives off 
a strong lateral branch. 

All the Charadriidse, and indeed all the Limicolae, 1 are schi- 
zognathous ; but many of them have not the typical condition 
of the vomer which accompanies as a rule the schizognathous 
skull. In the woodcut (fig. 169) a few exceptions to this are 

given, which range from the typical 
condition observable in Sterna to an 
excavated extremity, such as charac- 
terises Recurvirostra. In Chionis the 
vomer ends in the typical manner, i.e. 
in a point ; but it is exceedingly broad 
before its termination, and therefore 
quite unusual. 

In TJiiuocoms and Attagis the 
vomer is short and broad, and almost 
passerine in form. 

The maxillo-palatines are, as a 
rule, thin and scroll-like plates, which 
are bent downwards and often defi- 
cient in ossification, leaving holes here 
and there. The palatines have a 
spout-like process, extending upwards 
towards the base of the skull, which 
is especially well marked in CEdic/n - 
m-us. The Limicolse are nearly all of 
them schizorhinal, the delicate bar of 
i- the premaxilla being inserted at a 
different plane from the attachment 




Pm.r. premaxilla; V<>. vomer ; MX, 
maxilla: I'l, palatine; R. ros- of the nasals to the fl'Olltals. Ill the 
triiiu ; Xa, nasal ; M.rp, ruaxillo- 

, eetetimioia. Characlriidse proper there are a pair of 

largish occipital vacuities, one on either side of the foramen 
magnum. These same birds have basi-occipital processes, 

1 SHUFELDT (' Contributions to the Comparative Osteology of Arctic and 
Subarctic Water Birds,' J. Anat. Phys. 1889 and 1890) figures a few skulls of 
Limicolffi. See also the same, ' On Apliriza virgata,' Joiirn. J/or/i/;. ii. p. 311. 
See also 'Osteology of Niunenius,' &c., Journ. Anat. I'Jn/s. 1885, p. 51; 
' Observations on the Osteology of Podasocys mmtan/is,' ibid. 1884, p. 86. 


\vliichvary somewhat in the degree of their development. The 
lacrymal bone articulates with the ectethmoid, and makes a 
complete arch of bone in the anterior region of the orbit. The 
bones are particularly slender in Himantopus, leaving, there- 
fore, a large vacuity. In one or two types the foramen of the 
arch is almost obliterated by the thickness of the ectethmoid. 
The upper margin of the orbits in the Charadriidae is marked 
with conspicuous grooves for the supra-orbital glands. 

There are fifteen cervical vertebra in Htematopus, Nu- 
menius, &c. 

The atlas is perforated by the odontoid (Numenius, 

1. Sti'nid Idnindo. 2. Htematopm ostralegus. 
3. N/i men iits arquatus. 4. Rccurvirostra avocctta. 
5. Chionis alba. 

Limosa) . The Limicolae differ from the Grues in the fact that 
only one or two vertebrae (the ninth in Limosa) are furnished 
with two fairly closely approximated hsemapophyses for the 
reception of the carotids. In Numenius these exist on the 
ninth and tenth, and there are traces on the eleventh. The 
first dorsal vertebra, or the first two, has a large distally 
expanded haemapophysis. an indication of a state of affairs 
which is carried much further in the allied group of Alcae (q.v.) 
In these points the gulls may be contrasted with the more 
typical Limicolse. There are no paired hsemapophyses borne 
by the centra. 

The unpaired liEemapophyses extend (inLestris) from CIO 
to C15 ; on Dl-3 there are slightly bifid hsemapophyses. The 
atlas is notched. 

In Chionis CIO has a slightly excavated haemapophysis, 

z 2 


a trace of the double one of other birds. The haemapophyses 
of Dl, D2 are the longest, and the former is slightly trifid ; 
this trifidity is very marked^m the case of the two last cervicals. 
The atlas is notched. 

In Parra, as in the typical Limicolse, the atlas is perforated 
by the odontoid process. On the eleventh vertebra only do 
the two processes, which form an incompletely closed canal 
for the carotids, approach each other markedly in the middle 
line. The first dorsal vertebra has the largest haemapophysis, 
which is flattened slightly distally. 

(Edicnemus, with a notched atlas, has paired ventral out- 
growths for the carotid, closely approximated only on CIO. 
The three following have median blade-like hsemapophyses. 
On the fourteenth to sixteenth there are lateral outgrowths 
of these. The first dorsal has the last and the strongest 

In Attagis the atlas is perforated ; the hsemapophyses 
are very feeble. 

In Limosa and in other genera the clavicle is attached 
by ligaments to the acrocoracoid, procoracoid (which is mode- 
rately developed and curved upwards) , and scapula. The two 
coracoids are not in contact at articulation with sternum ; 
the latter is two-notched and has the spina externa only ; six 
ribs reach it. The same statements may be made about 
H&matopus and Nuiiit'iii-us, Eudromias (all examined by 
myself), as well as other genera. 

The pelvis of Numenius may serve as a type for that of 
the Limicolse. 

The preacetabular portion of the ilium is about equal 
in length to the postacetabular portion of that bone; the 
two bones are excavated horizontally, and are just prevented 
from coming into contact by the fused neural spines of the 
vertebrae. The pubes are stroiigish bones and not fused with 
the ischia. The ischia end in long thin processes which 
extend back beyond the ilia and nearly as far as the ends of 
the pubes. Hamatopus and Limosa are much the same. 
Chionis hardly differs. In Parra the pelvis has rather more 
the look of that of a rail. The pelvis of Attagis is wider 


than that of other Limicolae, and the ilia are rather further 
apart. In (Edicnemus bistriatifs, but not in (E. grallarius, 
the ischium has a well-marked pubic process, which reaches 
the pubis. 

As to the muscular anatomy, there is great uniformity in 
the tensores patagii of this group. Charadrius pluvialis 
may serve as a type. In that bird there is a biceps slip ; the 
tensor brevis early divides into two, of which the anterior is 
again divided not far from its ending ; a recurrent slip runs 
to the longus. 

The same disposition of tendons is found in Glareola, 
Niiiiieiiius, Scolopax, Himantopus, Vanellus, Machetes, Parra, 
Eecurvirostra, Totanus, Limosa, the only differences being 
that in some (e.g. Parra jacana) the middle only of the three 
tendons which are inserted upon the fore arm is continued 
over the muscles of the fore arm to the lower border of the 
ulna, while in others (e.g. Numenius) both the principal 
tendons are thus continued. In Glareola the middle and 
westward tendons are thus continued. 

Vanellus cristatas has been recorded with two separate 
biceps slips, which both run to the tendon of the longus. In 
Tringa canutiis I found a second biceps slip, largely but not 
entirely tendinous, which is attached to the outer of the two 
main tendons of the tensor brevis, this latter tendon indeed 
only dividing into two near its insertion on to the ulna. 

In Gallinago the recurrent slip connecting the two 
tensores seems to be absent. 

In Gambetta flavipes GAEEOD found no biceps slip at 
all, and it seems also to be absent in Metopidins afri- 

cdnus. 1 

Tlunocorus rumicivorus has the typical pluvialine arrange- 
ment of the tensores tendons that has been already described, 
but the biceps slip is remarkable for the fact that it has a 
tendinous band running along it. The characters of the 
tensores patagii in this group are fairly distinctive and at 
any rate serve to distinguish the Limicolae from the Ralli. 
They do not, however, permit of the enforcement of any 

1 Founds, P. Z. S. 1881, p. 


views respecting the families into which the Limicolae have 
been divided. 

An expansor secundariorum is almost universally present, 
but is often feebly developed. It appears to be absent in 
Tringa canutus and Chionis. 

The biceps is so far noteworthy that only the coracoidal 
head is present in Rliynclicea and in Parra sinensis. 1 The 
condition of this muscle in Himantopus is extremely 
interesting. It is stated by Dr. GADOW (on the authority of 
MECKEL) to be a double muscle. In H. nigricollis I find 
the following arrangement : There are two distinct portions 
(1) a part which may perhaps correspond to the entire biceps 
of other birds, with two heads, a coracoidal and a humeral ; 
from the former of these arises the biceps slip : (2) in addi- 
tion there is a distinct coracoidal portion, with a fleshy 
belly, which has, however, a common origin from the coracoid 
with the coracoidal head of the main muscle. Something of 
the same kind appears to occur in Chionis and Scolopax ; it 
may obviously be compared with the gulls (q.v.) In Cur- 
sorius the biceps was also double, though the division only 
commenced a little below the level of the humeral attachment. 

In Lobivanellus there were indications merely of the same 
division in the lower part of the belly of the muscle. Glareola 
has a biceps which is double for the greater part of its course. 

The anconceus appears to invariably possess the tendinous 
humeral head. 

My remaining notes upon the myology of the group are 
scanty. In Lobivanellus atronuclialis, the semimembranosus 
and semitendinosus are inserted by a common tendon, and 
the latter gives off a branch to gastrocnemius. There is but 
one peroneal muscle (the longus). The latter is alone 
present in Chionis alba and Himantopus niijricollis. The 
pectoralis primus in these birds does not appear to be 
divided into two layers. 

There is some variation in the deep flexor tendons of this 

1 In this bird the biceps slip arises (as figured by FURBKINGEK) from the 
humerus itself. 



In Tut anus calidris there is a slender vinculum, and, in 
addition, the flexor hallucis gives off a special slip to the 
branch of the flexor commnnis which supplies the second 
digit. The arrangement, in fact, is like that in Scopus, in 
many Accipitres, &c. In the Parridse, on the other hand, the 
tendons blend early upon the ankle, and in those that have 
been examined no branch to the hallux has been discovered. 

Chiunis alba has the deep flexor tendons of Totanus. 

In all Charadriidae the ambiens is present. The genera 
Charadriiis, Calidris, Gambetta, Gallinago, Limosa, Ma- 
chetes, Scolopax, Strepsilas, Totanus, Triuga, and VaneUnx 
have the reduced formula AXY + . 

In Hcematopus, Hiinantopus, Recurvirostra, JEgialitis, 
Nitmenius, there is the full formula ABXY+ ; so, too, in the 
representatives of the remaining families of the Charadrii, 
with the sole exception of the Australian thickknee, Bnr- 
rliinus, and some (Edicnemus, which have the formula BXY + . 

With the exception of the Parridae (q.v.) the charadrii- 
form birds have well-developed caeca. The intestinal mea- 
surements of a few types are as follows : 

Small lur. 

Large Int. 





Hcematopus ostralegus 




(Edicnemus crepitan-s 

22 (32) 


2i (3) ' 

girillaiiiis . . 16 (17) 

2ir (3) 


,, superciliaris . . 1ft 


2 2 

,, bistriatus . . 25 


2 f 

Himantopus brasiliensis . 



Ninneiiiiin t/r<[H(itits . . . \ 30 



Strepsilas intcrprcs . 




Gambetta flavipes 




Glareola pratincola . . . s. 1 , 


Si-ulo/itu- nuiticola ... 4s 



Gallinago /jnllni/ila . 



I 2 

Tringa citimt/is 




Limosa rufa .... 



Numenius phesopus . 




\'niicllttt< cayennensis 


2f i 

Reciirrinifftrn /irnrrtta 

41 (30) 

Hydrophasianus cJtirurga . 



Thinocorus rumicivorus 


2-J- and 2^- 

Chionis alba .... 



8-25 and 9 

Attagis ..... 



Eury-norhynchus /ii/giHn'/ix - 




1 The brackets roiit:iinu r ineiisurrnii'iit-i uf ;i srccunl 

- ANDKKSON, -MU the Pterylosis, &c.,of the Spooil-billed Smuli.iiier,' Tr.l.h.n. *'><-. (2), i. p. 213. 


Comments upon the facts set forth in the above table must 
obviously be discounted by the variations (quite considerable 
in amount) which occur in one or two of the species. The 
table given by GADOW l increases the number and extent of 
the individual differences in intestinal length. 

In the liver the right lobe is, as a rule, larger than the 
left. Sometimes it is only slightly so ; but in Charadrius 
pluvialis the right lobe is twice as large. In Scolopax rusti- 
cola and in (Edicnemus crepitans the lobes are equal. The 
gall bladder is nearly always present. GADOW did not find it 
in a specimen of Numenius arquatus and of Tringa arenaria 
and T. alpina. This, however, appears to be individual. 

In all the Charadriidse the syrinx is of the tracheo-bron- 
chial pattern. 

In Vanellus cayennensis 2 (fig. 45, p. 66) the last twenty- 
five tracheal rings are narrower than those which precede them, 
and of equal diameter throughout. The first two bronchial 
semi-rings are like split tracheal rings ; the next two are very 
closely applied together ; the remainder are normal bronchial 
semi-rings with no modification. The most remarkable fact 
about the windpipe is the enormous size of the intrinsic 
muscles, of which, however, there is only a single pair. The 
muscles end in a tendon, w T hich is inserted on to the second, 
third, and fourth semi-rings. In Vanellus cristatus the intrin- 
sic muscles are certainly large, but not so abnormal as in the 
other species of the genus. Only the last four tracheal rings 
are modified, and in front they are all fused in the middle 
line to form a bony box ; behind the last three tracheal rings 
are semi-rings, the pessulus being attached to the fourth. 
The muscles are attached to the first bronchial semi-ring. 

In Himantopus nigricollis there w r as no trace what- 
ever, that I could detect, of these muscles ; nor in H. 
brasil'iensis. Charadrius pluvialis, Hamatopus ostra- 
legus, and Squatarola helvetica are also without these 
muscles. On the other hand they are present in Totanna 

1 In Bronn's Thicrreich, ' Aves,' p. 624. 

2 GAKEOD, ' On the Trachea of Tantalus loculator and of Yandlits caijoi- 
neiisis,' 1 P. Z. S. 1878, p. 625. 


can-ntus, Tringa cinchix, Xtiiiicitiiix arquatus, 
hiaticula, Limosa r-itfa, L. cegocephala, Machetes pugnax, 
Streps-Has interpret, and Scolopax ritsticola. But although 
the muscles are present in the birds included in the second 
list they do not, in all of them at least, reach as far as the 
bronchi, though they may possibly be continued by fibrous 
tissue to a more normal point of attachment. Thus in 
Limosa cegocephala the muscles stop three or four rings 
before the end of the trachea. In Scolopax rusticola, on the 
other hand, the rather broad intrinsic muscles reach as far as 
the first bronchial semi-ring. 

The windpipe of the Indian painted snipe (liliynclicea 
capensis) is peculiar in that it is convoluted slightly in the 
female, not in the male, as might be expected in view of this 
frequent difference between the sexes in other birds. As has 
been pointed out, the female is the larger and more richly 
coloured of the tw r o, a fact which is in harmony with the 
more complicated trachea. 

This is not seen in young females. The same condition 
is stated by GOULD to characterise the Australian R. ans- 
tralis. 1 

The large group of the Limicolse has been variously 
divided. I follow GADOW in the families, but include also the 
Laridae, which I separate from the auks. I should define 
the Charadriidae, to which most of the foregoing refers, as 
schizorhinal birds, with occipital fontanelles, furrows for 
supra-orbital glands and basipterygoid processes, and fifteen 
cervical vertebrae. 

The family (Edicnemidse has been instituted for the genus 
(Edicnewws, which includes the Norfolk plover and a number 
of other species closely allied ; these range widely, being only 
absent from North America, Central Asia, and New Zealand. 
The Australian (E. <jr<tll<iriux has been separated as a distinct 
genus, Bnrliinus, which, as also (E. crcjiitu ns, instead of 
possessing the complete muscle formula (ABXY + ), as in 

1 WOOD-MASON, 'On the Structure and Development of the Trachea in the 
Indian Painted Snipe (Rliyncluea capcnu'iK)' P. '/.. N. 1*7*. p. 74.",. 


the other members of the genus, has the formula BXY + . 
The ptcrijlosis is as in the Charadriidae, but the number of 
rectrices may be as numerous as fourteen. There is no 
liallux. In the skull the absence of basipterygoid processes 
(sometimes indications of them are present, according to 
GADOW) and the holorhinal character of the nostrils distin- 
guish this family from the Charadriidae. The depressions 
for the supra-orbital glands are well marked. The lacrymal 
bones are nearly (occasionally quite, though by suture) 
united with a process of the frontals, and form a canal, 
through which the gland apparently passes ; this is seen in a 
more exaggerated way in Chionis and in Vanellus (v. de- 
scription of Chionididae) . The post-orbital angle is not 
distinct from the post-frontal process. There are no occipi- 
tal foramina, as in Charadriidae. There are sixteen cervical 
vertebra, the last three of which have ribs of progressively 
increasing length ; five or six ribs articulate with sternum. 
Contrary to what is found in the Charadriidse, the coracoids 
slightly overlap at their external articulation. 

Some of the visceral characters have been already dealt 
with above in the general description of the Limicolse. 

There is a tendency for the ambiens not to cross the knee ; 
this occurs individually with specimens of several species. 

The syrinx has not, except as an occasional variation, 
any intrinsic muscles. 

(Edicneinus bistriatiis has on the anterior wall of the 
cloaca two hardly elevated ridges, which end by slightly free 
points, and seem to represent a rudimentary penis. 

The surgeon birds and ja9anas of the tropical regions of 
the Old and New Worlds, with their enormously elongated 
feet and their somewhat rail-like aspect, are now known to 
belong to the Limicolae (and not to the Ralli), of which they 
may be regarded as forming a distinct family, Parridae. 
Their anatomy has been chiefly studied by FORBES.' As 
with the Thinocoridae so with the present family there is a 
character of the alimentary canal which immediately distin- 

1 ' Notes on the Anatomy of the Ja^anas,' P. Z. S. 1881, p. G39. 


guishes them from all their allies. In this family the caeca 
are mere passeriform nipples, measuring from '15 to '2 of an 
inch in length. 

The muscular anatomy has already been to some extent 
treated of in connection with the structure of the entire group 
of birds of which the present genera form a family. In all 
of them, Pa rra, Hijdropliasianus, and Metopidius, the mus- 
cle formula is complete, i.e. ABXY+ . The condition of the 
deep flexor tendons of the foot is very singular. As FORBES 
justly pointed out, the peculiarly large size of the hallux (as 
of all the digits) of the foot in these birds seems to be un- 
reconcilable with the entire absence of a special slip from the 
conjoined tendon of the long flexors. ' This fact,' he thinks, 
' seems to indicate that the Parridse may have been developed 
from some form with a more normal-sized foot, and a small 
hallux which had no special long flexor, the great size of 
their feet having been developed in accordance with their 
peculiar habits.' 

The syrinx has a pair of intrinsic muscles. 

The skull has well-formed basipterygoid processes, but 
no occipital foramina or supra-orbital impressions. In 
Metopidius the radius is extraordinarily enlarged (see fig. 70, 
p. 125). In the remaining genera there is no such modifica- 
tion of the bone, but there is a metacarpal spur, which may 
be of the same use, i.e. for fighting. In Parra the clavicle 
is at its articulation further from the procoracoid than in the 
Charadriidse, and the sternum has only one pair of notches. 
Five ribs reach it. 1 

The single genus Chionis,- of antarctic range and some- 
what gull-like form, makes up the family Chionididse. 

There are twelve rectrices. 

The skull is peculiar in that the grooves for the supra- 
orbital glands end in a large foramen on each side, which is 

1 The bones of Parra albin nclia are described and figured by MILNE-EDWARDS, 
Hist. Madagascar. 

- The peculiar sheath which covers the base of the bill and the nostrils 
(whence ' Sheathbill ') is declared by STUDEK to be developmentally different 
from the tube of the Tubinares. 


formed by the union of the lacrymal with a process of the 
frontal, forming a continuous bony bar. As has been already 
mentioned, (Edicnemus shows a very considerable approach to 
this state of affairs. So too, as I interpret it, do the quite 
typical charadriid Vanellus and Eudromias. In these birds 
the grooves for the supra-orbital glands do not, as they 
do in Limosa and Recurvirostra, border the margin of the 
orbit. They are situated at some distance from it, and each 
ends in a very small foramen, bordered in front by the 
ankylosed lacrymal, which I take to correspond to the large 
foramen of Chionis. 

There are 110 occipital foramina or basipterygoid pro- 

The cervical vertebra are fifteen, of which the last three 
bear discrete ribs ; six ribs reach the two-notched sternum, 
and there is one behind. The clavicles have no hypocleidium, 
and end a long way in front of carina. The coracoids are 
not in contact at sternal articulation. 
The muscle formula is ABXY+. 

The syrinx of Chionis is not widely different from that 
of other Limicolse, and exhibits, as will be seen from the fol- 
lowing description, no particular resemblance to the Galli. As 
is the case in so many Limicolae, the intrinsic muscles end as 
such some way in front of the bifurcation of the windpipe, 
though they are continued on to the bronchi by fibrous tissue. 
They end in Chionis upon the fifth tracheal ring counting 
from the last. 

The last four tracheal rings are more or less closely united 
to form an ossified box. The first bronchial semi-ring to 
which the fibrous continuation of the intrinsic muscles is 
attached is the widest (from before backwards) of the rings 
of the windpipe, and is deeper than the bronchial semi-rings 
which follow. 1 

1 The following are the principal memoirs dealing with the anatomy of this 
bird : EYTON, ' Note on the Skeleton of the Sheathbill,' P. Z. S. 1858, p. 99 ; 
A. REICHENOW, ' Osteologie von Chionis minor,' Ac., J. f. O. xxiv. 1876, p. 84 ; 
SHUFELDT, ' The Chionidse : a Review of the Opinions on the Syctematic 
Position of the Family,' Auk, 1893, p. 158, and in ' Contributions to Comparative 
Osteology,' &c., J. Anat. 1'lii/s. 1891, p. 509 ; KIDDEK and COUES, ' A Study of 



It is usual to separate the two genera Thinocorus and 
Attagis, both South American birds, into a family, Thinoco- 
ridae. In contradistinction to their allies they are graiii- 
eating birds, connected with which habit is the presence of a 
crop, an absolutely distinctive character so far as the present 
group is concerned. Their anatomy has been chiefly studied 
by GAEROD.' The differences which distinguish them from 

FIG. 170. SKULL OF Attar/is Gayi (AFTEB GARBOD). 

other Limicolae are neither great nor numerous. In the skull 
the basipterygoid processes are absent, and the vomer is broad 
and rounded in front as figured by PARKER ; 2 the skull, in fact, 
as has been mentioned on a previous page, is segithognathous 
rather than schizognathous. There are no occipital foramina, 

Chionis minor,' Bull. U.S. Nat. Mus. iii. 1876, p. 85 ; E. 0. CUNNINGHAM, ' On 
Chionis alba,' J. Anat. Phys. 1870, p. 87 ; BLAINYILLE, ' Memcire sur la Place 
que doit occuper le Genre CJiionis,' Ann. Sci. Nat. 1836, p. 97 ; STUDEB, 
Forschitnysn>i.<i<> S. M. S. ' Gazelle,' Bd. iii. ' Zoologie u. Geologie,' p. 107. 

1 ' Notes on the Anatomy and Systematic Position of the Genera Thinocorus 
and Attagis,' P. Z. S. 1877, p. 413. 

'- ' On JEgithognathous Birds,' Zool. Trans, vol. x. 


but the supra-orbital impressions are present. The maxillo- 
palatines are very far apart, and besides being short are 
swollen instead of being leaflike plates. The same pseudo- 
holorhiny that characterises the Glareolidse is also found in the 
present family. Five ribs reach the sternum. The coracoids 
are quite separate at their insertion on to sternum. The 
muscular formula of the two genera is the complete one 
ABXY + . 

In Tliinocorus rumicivorus the syrinx has a pair of 
lateral muscles, which are attached to the fourth incomplete 
ring when seen from in front. This is probably the first 
bronchial, the last three tracheal rings being thus incomplete. 

Glareola, Cursoriiis,Plnvianus,a.ndDroinas, all Old-World 
genera, are included in a separate family, Glareolidae, which 
GAEEOD regarded as very near akin to the Thinocoridse. 1 As 
in them, basipterygoid processes and occipital fontanelles are 
absent, and impressions of supra-orbital glands present. The 
muscle formula too is complete, and the syrinx has intrinsic 

Except in Cursorius there is a hallux. FORBES, with 
some reason, has united into a group Pluviales, equivalent to 
the rest of the Limicolae, this family, together with the Thino- 
corida3 and Chioniclidse, mainly on account of the above 
combination of skull characters found in no other Limicolae. 

As in them also there is pseudo-holorhiny, the bony 
nostrils, though extending back beyond the nasal process of 
the premaxilla, being distinctly rounded off. 

Pluvianus, however, has typically holorhinal nostrils. 

In Cursorius the grooves for nasal glands are converted 
into elongated foramina by a fusion between adjacent pro- 
cesses of the skull. 

The gulls form another distinct family, Laridae, containing 
the genera Larus, Lestris, Sterna, Rhynchops, Aiwus, and 

The skull is schizorhinal and schizognathous, without 

1 LINN.T.US placed Glareola in genus Hirundo, SUNDEVALL in Caprimulgidoe. 


basipterygoid processes or occipital fontanelles, 1 but with 
well-marked grooves for supra-orbital glands. In Lestris 
antarcticus at any rate there is a foramen formed round 
anterior end of supra-orbital groove, as in Chiotiis (q.v.) In 
Lestris there is a distinct tendency towards the pseudo-holo- 
rhinal condition of the Thinocoridae and Glareolidse. 

In Larus the relations of the clavicles, scapula, and cora- 
coids are as in the Charadriidse ; but the clavicles provided with 
a hypocleidium come into nearer relations with the carina 
sterni, to which they are attached by a ligament. The cora- 
coids are in contact behind the spina (externa) sterni. The 
same statement may be made of Lestris. 

The cervical vertebra are fifteen in number. The dorsal 
vertebrae are, as in Alcidse and plovers, opisthoccelous. Six 
(Lestris antarcticus) or seven ribs articulate with the sternum. 
The first phalanx of the second is commonly perforated. 2 

As to the pterylosis, the plan is that of the Limicolae, but 
the ventral tract does not divide until some way down the 
neck. The feet are webbed, the hallux is small or absent, 
and there are twelve rectriccs. 

The muscular formula of Rissa tridactijla is AX+ ; of 
Larus, Stercorarius, and Gygis, AXY + . The other genera 
have the complete formula ABXY + . Rhynchops has no 

In Lestris crepidatus and L. antarcticus the semi- 
membranosus is sometimes two-headed, one arising from 
ischium and one from postacetabular ridge of ilium. 

The tensores patagii 3 (see fig. 171) are on the plan of 

1 These are present in the young : see PARKER, Linn. Trans. (2), i. p. 142. 

: This does not, however, as it has been stated to do, distinguish the Larida? 
from other Limicolee ; though apparently universal in the gulls (including 
Anous and Gygis), the same perforations are found in Glareola. The value of 
this anatomical fact may be judged by the perforation of the same bone in such 
varied types as Pteroclcs, Coracopsis obscura, (not in C. nigra), Psittacula 
passcrina, Machcerhamphus Andersoni, Heliodilus, Caprimulgus, Phactlion 
candidus (not P. rubricauda), and Fregata minor. These instances are taken 
from the osteological plates illustrating MM. GKANDIDIEK and MiL\E-En\VAr,i>s's 
Histoire Naturellf. de Madagascar. 

:! For muscular anatomy of Lamias see GIEBEL, ' Beitriige zur Anatomie 
d. Moven,' &c., Zeitschr. /. d. ges. Naturw. x. (1857), p. 20; BEI>DAHI>, 'A 
Contribution to the Knowledge of the Anatomy of Rliynchops,' 1 P. Z. S. 1896, 
p. 299. 


those of other limicoline birds, but are apt to be a little more 
complicated. In Larus argentattis, of which the tendons are 
represented in the annexed cut, the anterior stronger branch 
of the brevis tendon gives off a forwards and downwards slip 
to the extensors of the fore arm, from which arises the usual 
connection with the long us tendon. There is a second con- 
nection between the two tendons. At the origin of the 


t.p.l, teusor longus ; t.p.b, tensor brevis; B, wristward slip ; F, patagial fan ; 
A, tendons to ulnar side of arm. 

patagial fan is a bony nodule, as in the petrel. The tensor 
longus tendon also gives off a slip (A, fig. 172) to the inside 
of fore arm which is also present in the auks (g.r.) In 
Rliyncliops (fig. 171, A) there are two such tendons. 

Lestris antarctica, Sterna, and Larus marinus (according 
to FUEBEINGEE'S figure, PI. xix.) are much the same, but are 
without the additional slips A and B. Of these A is present 
in Eissa tridactijla. 

The expansor sccundariorum is frequently absent, but it 



is present in Larus argentatus, marinus, and glaucus, not 
infuscus. It is absent in Sterna and Lestris ; present in 
Anous ; absent again in Ehynchops. 

Tbe biceps is peculiar in some members of this family. 
In Larus, Sterna, and Anous the biceps has, as usual, the 
two heads, humeral and coracoidal ; but they form two per- 
fectly distinct muscles, of which one, the coracoidal, soon 
divides into two distinct muscles again, which are inserted 



it, osseous nodulu. Other letters as in fig. 171. 

respectively upon the radius and ulna. The biceps head 
goes to the radius, and as a rule gives off the biceps slip to 
the patagium ; but in Larus ridibundus, according to FUE- 
BEINGER, this slip arises from the coracoids, an anomaly 
observable also in the petrel, Thalassiarche. Ehijn chops 
has no biceps slip at all. 

The anconceus is generally attached by a tendon to the 

A A 


The cfBca vary considerably in their development, as the 
following table of measurements shows : 

Small Int. 


Larire Int. 

Larus argentatus . 

56-25 (36) 

5 (-25) 

3-25 (1) 

, glaiicus 




, ridibundus . 



, Jamesoni 




Lestris pomatorhinus 




, antarcticus 




, crepidatus . 




Sterna cantiaca 



Rissa tridactyla 




The lobes of the liver are subequal in Sterna, Anous, 
Gygis, and Rissa. In Larus and Lestris the right is the 
larger. A gall bladder is always present. 

In Anous the cceca are quite short ; in Gijgis long and 

The syrinx of this family is typically tracheo-bronchial 
(at any rate in Lams'), with well-developed muscles, which 
always reach the bronchi. The family thus differs from 
other Limicolae where there is a tendency towards a retro- 
gression of the intrinsic muscles, sometimes culminating in 
actual disappearance. 

In Lams marinus the last six or seven tracheal rings 
are rather narrower from above downwards than those 
which precede them, and are more or less firmly attached 
(except the last ring, which is incomplete both in front and 
behind) to form a box. When the syrinx is viewed from 
behind, a broad three-way piece is seen, into the formation of 
which 'the penultimate tracheal ring and the four or five in 
front of it appear to enter. This piece, however, is only 
really solid at the edges, the bars being a continuation of 
the penultimate tracheal rings. In the middle it is so thin 
as to be little more than a membrane. The first bronchial 
semi-ring (to which the intrinsic muscles are attached) is 
bow-shaped and in close contact with the last tracheal ring. 
The remaining semi-rings are narrower and run in a straight 
direction across the bronchi. 

The membrana tympaniformis is distinguished by its 


thinness and transparency from the thick yellowish mem- 
brane which unites the edges of the greater number of the 
bronchial semi-rings. 

In Larus fuscus and L. glaucns the differences are but 
slight, and chiefly concern the greater solidity of the three- 
way piece. 

In Larus argoitatus the edges of the three-way piece 
are most solid ; but they are connected by a series of four 
or five bars which divide up the central region of the three- 
way piece into alternate thicker and thinner portions. This 
is a peculiar specialisation of the three-way piece which I 
have not observed elsewhere. There is, furthermore, a thin 
bony curved rod, closely applied to the outside of the three- 
way piece, which arises from, or is at least connected with, the 
last tracheal ring. 

Lestris antarcticus has a syrinx which is rather different 
from that of Larus, and which points in the direction of the 
Charadriidse, owing to the fact that the intrinsic muscles do 
not reach the bronchial semi-ring except as a fibrous band. 
The three-way piece, which is solid, is formed by two or three 
tracheal rings ; there is no differentiation in membrane 
closing bronchial semi-rings. 

It is evident from the foregoing account that, while there 
are a few differences between the various genera of Laridae, 
Bkyn chops is quite the most anomalous form in structure as 
well as in external appearance, as seen in its remarkable bill, 
with its scissor-like edge and projecting mandible. Still, 
the differences are, in my opinion, not sufficient to place 
Rhy licit ops in a family by itself opposite to the remaining 
Laridse. It seems that the usually received division of the 
family, making an additional one for Rliyncliops, will serve to 
divide naturally the Laridse ; we may term these divisions 
subfamilies. They will be thus characterised: 

Subfamily I. Larinse. 

Muscle formula of tin- lt'<j, J.YY+. Ca'ca rudi- 
nii'iitanj. Biceps slips and c.i-panxor st 
dariorum present. 

\ A -2 


Subfamily II. Stercorariinae. 

Muscular formula, AXY +. Caca long. Biceps 
slip present. Expansor secundariorum absent. 

Subfamily III. Sterninge. 

Muscular formula, ABXY+. Caca rudimentary. 
Biceps slip present. Expansor secundariorum 

Subfamily IV. Rhynchopinse. 

Muscular formula, ABXY-. Cfsca rudimentary. 
Biceps slip and expansor secundariorum absent. 

Gygis and Anous require further investigation before they 
can be placed in this system ; they are usually regarded as 
terns by HOWAED SAUNDEES, for example. Anous has the 
complete muscle formula. Gygis has the formula of the 
gulls; but then the tern, Sternula (sp.), wants the accessory 
femoro-caudal, and in this approaches the gulls. Anous has 
the expansor secundariorum. 

Of undoubted or reputed extinct limicolous birds a number 

have been described. If PaUeotrinya (with three species), from 

North America, is rightly referred by MARSH to this group, it goes 

back to Cretaceous times. Milnea, from French Miocene, is known 

by the humerus. Elorius, known by an imperfect coracoid, a tarso- 

metatarsus, and parts of the skull, seems to belong here. JEijial- 

ornis is considered by LYDEKKER to have been a gull-like bird, 

largely on account of the perforated first phalanx of the second 

digit. 1 Haley ornis has been described from the extremity of a 

humerus and the back of the skull. 





i ce 
ce = 







*% 3vS 








O pa 

& S-i 

Charadniilir . 
CEdicnemidsa . 





1 -Hutched 





A(B)XY + 
(A)BXY + 

Tliinocoridae . 




1 -notched 





AI;XY + 

Laridre . 



In cotit. 




1 A character which we have seen (supra, p. 351, footnote) to be of no 
account in fixing affinities. 


The foregoing table shows some of the principal points 
in which the several families of the Limicolae differ from 
each other, and will afford a justification for the divisions 
adopted in the present work. AVhatever is the relation 
between the other families, we can clear the ground by 
removing the Laridse from competition for the basal place 
in the series. As was discovered by PARKER,, the young of 
these birds have basipterygoid processes and occipital formina, 
the persistence of which, therefore, in the Charadriidse and 
Parridas (basipterygoids only) places those two groups lower 
in the series than the more specialised gulls. That the gulls 
are rightly placed here, and therefore as rightly removed from 
a closer association with the Alcee, can hardly be disputed. 
GADOW, who does the reverse in his scheme, enumerates only 
the following points in which the gulls differ from the 
Limicolae :- 

In the Laridae 

Down feathers are thicker. Coracoids in contact. Haam- 
apophyses mostly (not in Lcstris) wanting to the dorsal ver- 
tebrae. Hypotarsus simpler. In muscle formula of leg dis- 
appearance of B instead of Y. Webbed feet. 

As a matter of fact, the crossing of the coracoids in 
(Edicnemus destroys the second of these, at best very slender, 
grounds, and, as GADOW admits, the webbing is almost as well 
developed in Recurvirostra. 

On the other hand the differences from the Alcae are 
more pronounced. 

These latter birds have 

A much longer sternum. 

Largely developed dorsal ha3inapophyses, of which in- 
dications only are to be found in the gulls and in other 

The biceps slip is peculiar. 

The leg muscles are always reduced, the formula being 
in Phaleris only AX . 

It may be mentioned in addition that the expansor 
secundariorum is always absent in the Alcse and only some- 
times in the Laridge. The auks are, in fact, so far as we 



can see, a further modification of the gull type, but further 
from them than they are from the Limicolae. 

We may, therefore, probably regard the gulls as a slightly 
modified offshoot of the typical Limicolae (Chionis, perhaps, 
being slightly intermediate), leading towards the Alcae. The 
arrangement inter se of the remaining families brings us to 
the broader question of the affinities of the Limicolae as a 
group. It appears to me that the only other groups which 
need be specially considered in this connection are the 
Grues, Otides, and Ralli. It may be convenient to preface the 
discussion with a tabular statement of the actual points of 
likeness between these several groups. 





Cerv. vertebra 





"Baslpt. pr. . 
Nares . 

- or + 
Schiz. or hoi. 

Schiz. or hoi. 



Occip. for am. 
S termini 

+ or - 
1- or 2-notchecl 

+ or - 
1- or 2-notched 


1- notched 

Leg muscles 
Patagial fan 
Siceps sli/i . 

A(B)XY + 


(AB)XY + 

+ or - 

BXY + 



The very difficulty of finding any characters, greatly 
noteworthy, in which the groups in .question vary is an 
index of how closely allied all four are. There can, to my mind, 
be no doubt of their common origin. The Limicolae on the 
whole come nearest to the Grues, and especially to the true 
cranes, whose distinctive characters are a little swamped in 
the above table, owing to the aberrant cranes (e.g. Ehinochetus, 
Psopliia), whose peculiarities have naturally modified that 
table. Taking this point of view, it seems to follow that of the 
Limicolae the most primitive section is that of the Chara- 
driida3 proper ; for it is among them that the forms with 
the greatest number of points of resemblance to the cranes 
occur. I should consider these birds to be slightly lower in 
the scale than the cranes. 

ALCLE 359 


Definition. Oil gland tufted. Aftershaft present ; aquincubital. 
Skull schizognathous and sch.izorh.inal. Occipital fontanelles 
present, but no basipterygoid processes. Two carotids. 1 Ten- 
sores patagii and biceps slip distinctive. No expansor secuii- 
dariorurn. Dorsal vertebrae opisthcocoelous. 

This group of birds comes nearest to the Limicolae, but 
differs more from any of them than they do among them- 
selves. The group is entirely confined to the northern 
hemisphere, and is mainly Arctic. 

The oil gland is invariably tufted and the feathers have an 
after shaft. 

In the pterylosis the dorsal tract divides between the 
scapula, and there is (? universally) a well-marked spinal 
space. But there is no break between anterior and posterior 
parts (as in Limicolse). The ventral tract, contrary to what 
we find in the Limicolse (incl. Laridae), does not divide early 
in the neck. The rectrices (see table, later) vary in number 
from twelve to sixteen. The Great auk (Alca impennis) is 
said to have possessed eighteen. 

The oil gland has often many apertures ; there are only 
two in Brachyrhamphus marmoratus, but four in Lunda 
cirrhata, six in Synthliborhamphus antiquus, and eight in 
Uria columba.' 1 

The skull is schizorhinal 3 and schizognathous, with well- 
marked occipital fontanelles 4 and impressions for supra-orbital 
glands ; the latter nearly meet in the middle line, leaving but 

1 Synthliborhamphus antiquus has only one (the left). 

: The remarkable shedding of the beak of the puffin (stated also to occur in 
the penguin ; cf. P. Z. S. 1880, p. 2) has been described by BUREAU in Bull. 
Soc. Zool. Fr. ii. 1877, pp. 377, 432. See also ibid. iv. 1879, p. 1, for the same 
phenomenon in other auks. 

3 For osteology of auks see OWEN, ' On Alca inqwiinis,' Tr. Z. S. v. p. 317 ; 
SHOFKLDT, J. Anat. Phijs. vol. xxiii. ; and PARKER, ' On the Morphology of the 
Duck Tribe (Anaticlse) and Auk Tribe (Alcidae),' Cunningham Memoirs E. Irish 
Ac. No. 6, 1890. 

1 These are sometimes obliterated with age. I find them present in a young 
Uria troile, absent in an old one. In a specimen of Fratercula arctica there 
was only one present. They also may be present or absent in IBrachyrhamphus 
and Synthliborhamphus. 



a thin median line. The interorbital septum is very imper- 
fect. Basipterygoid processes are absent in the adult, but 
present as rudiments in the young. The lacrymals are 
firmly united with the prefrontals, as in gulls and Limicolse. 
The vomer is distinctly double in the young. Alca tor da 
has an os uncinatum. In Uria troile and in Alca tor da and 
A. impennis there is on each side a foramen at anterior end 
of supra-orbital grooves, as in Chionis, &c. There are fifteen 
cervical vertebra ; the first few dorsal vertebrae have very 
conspicuous haemapophyses, bifid at their free ends. Seven 



o, slip to ulnar side of fore arm. 

FIG. 174. THE SAME OF Synthll- 
borliainpli us antiq it its. 

ribs reach the very long and narrow sternum, which has one 
notch on each side posteriorly (Uria), or in addition a fenestra 
on each side and a median notch (Fratercula) or no notches 
at all (Alca impennis). 

As to the muscular anatomy of the auks, the simplest 
form of the ten sores pa'tagii tendons l is seen in Alca tor da, 
where there are two brevis tendons, both of which pass over 

1 For muscles of Alca (and Spheniscus) see A. CARLSSOX, ' Beitriige zur 
Kenntniss d. Anatomie d. Schvfimmvogel,' Bill. K. Svensk. Vet. Ak. Handl. 
ix. 1884, No. 3. 



the extensors of the fore arm to be inserted on to the ulna ; 
the anterior of these bifurcates in the limicoline fashion 
just in front of its insertion, and here a very faintly marked 
fan l (FiJEBEiNGEE) connects it with the longus. 

Synthliborhamphus antiquus is in some respects even 
more simple. 2 

There is but one brevis tendon, from which just as it 
passes over the extensor of the fore arm the merest apology 
for a forward branch exists ; 
from this branch arises a 
special muscular belly of 
the extensor metacarpi (cf. 
Petrels). There is no 
patagial fan, but a special 
slip, found in all the other 
members of the family (and 
also occurring in Lanix 
argent at us), runs from the 
longus tendon to the op- 
posite side of the fore arm 
to that upon which the 
patagial fan, when present, 
is inserted. 

In Bracliyrliamplius 
marmora tus three separate 
and parallel tendons arise 

rliinu inoiiocerata (AFTER BEDDARD FROM 

from tensor patagii brevis muscle, of which the anterior is 
the strongest and alone passes to the ulna. From a small 
wristward slip arise a few fibres of the extensor metacarpi, as 
in the last species. There is no patagial fan, but an ulnar slip, 
which gives off a branch running back to the humerus. In 
Uria columba there is the same. 

. In Lunda cirrhata there are but slight differences ; the 
two most anterior of the brevis tendons cross the extensor 

1 Not figured at all by GARROD in a MS. sketch. 

- For various details in anatomy of soft parts see BEDDARD, ' On the Anatomy 
of a Grebe (JEclimopliorus ii/rijor), with Remarks upon the Classification of 
some of the Schizognathous Birds,' P. Z. S. 1896, p. 538. 


muscles and reach ulna, but they cross it as a single diffuse 
band formed by their fusion. 

Ceratorhina inonocerata has a slight patagial fan, as well 
as ulnar slip ; otherwise it is like the last. 

Fratercula arctica has only two brevis tendons, upon the 
anterior of which, at the origin of the patagial fan, is an 

The biceps slip is present in Alcidse, but is generally, if 
not always, peculiar. Thus in Alca torda it is inserted partly 
on to patagial membrane, partly on to inner of two brevis 
tendons (not long us tendon). 

In Lunda cirrliata it arises tendinously and joins inner 
of three patagialis brevis tendons. The biceps slip (as in 
some petrels) is all that is left of the humeral head of the 

The biceps slip of Fratercula arctica is inserted on to 
middle of three tendons. 

]n Brachyrhamphus marmoratus the biceps slip ends in 
a long and fine tendinous thread, which is inserted on to the 
innermost of the three tendons of the brevis. 

In Plialeris psittacida the biceps slip is firmly adherent 
to the single brevis tendon ; but from it just at the lower end 
of the line of the brevis tendon an obliquely running strand 
is found, which reaches the tendon of the longus. 

The biceps, as already mentioned, consists merely of the 
coracoidal head, the humeral head being represented only by 
the biceps slip. 

The muscle is not large, and in Plialcris its muscular 
belly is largely divided into two. 

The humeral head of the anconceus seems to be nearly 
always present. 1 

There is no expansor secundarioruin. The deltoid has, 
as a rule, no scapular slip, but there is one in Uria. 

The muscles of the leg which are invariably present are 
the femoro-caudal and the seinit<'iulinoxnx. The accessory 
head of the latter is never present. The ambiens and the 
accessory femoro-caudal may be present, and, except in 

1 It is (? individually) absent in Brachyrliamplius marmcratus. 



Phaleris psittacula, are never both absent; the formulae are 
the three following : 

ABX- Uria. 

AX+ Ceratorhina. 

AX Phaleris. 

The biceps femoris of Braclujrliamphus gives off a fleshy 
slip to the outside of the thigh, to the gastrocnemius (cf. 
Podica). This slip is not to he found in Phaleris. 

The least modified form of syrinx is seen in Alca tor da, 

FIG. 176. SYRINX OF Loinrin 

troile (AFTER BEDDARD). 

, intrinsic muscle. 

FIG. 177. SYRINX OF Cerutorlinnt 
monocerata (AFTER BEDPAKI>). 
i, intrinsic muscle. 

Lomvia troile, Synthliborhamphus antiquus, and Uria 

In Lomvia troile the last four or five tracheal rings enter 
into the formation of a three-way piece, which is deeply 
excavate medianly. The intrinsic muscles are attached to 
the first bronchial semi-ring, which hardly differs from those 
that follow ; it and the one which immediately succeeds are 
slightly more broad than the rest. 

Alca tor da is rather more gull-like, there being no 
marked depression in the pessulus medianly, and the trims- 
parent membrana tympaniformis being sharply marked off 
from a thicker yellowish region behind. 


Synthliborhatnphus antiqmis has also a perfectly typical 
tracheo-bronchial syrinx, the first bronchial semi-ring (to 
which intrinsic muscles are fixed) being longer and deeper 
than those which follow. Phaleris is similar. 

Uria columba distinctly differs from Lomvia troile (with 
which it is often considered to be congeneric), and is an 
approach towards a type of syrinx to be described imme- 
diately, but with certain peculiarities of its own. 

A dozen tracheal rings in front of the last are very 
thin (more particularly in front), and have, therefore, wide 
membranous intervals. The last tracheal ring, however, is 
stout and ossified ; it appears to be composed of two closely 
adjoined ; posteriorly three rings enter into the formation 
of the tracheal box. The first bronchial semi-ring is very 
much arched, so much so that laterally it conceals the last 
tracheal rings. To it the intrinsic muscles are attached. 

In Ceratorhina monocerata this state of affairs is ex- 
aggerated. Not only the first, to which muscles are attached, 
but the second bronchial semi-ring is very convex upwards, 
forming, indeed, the half of a rather elongated ellipse, as 
shown in the figure (fig. 177). The last two tracheal rings 
are ossified and closely connected. The last twelve tracheal 
rings are shallow vertically and leave considerable mem- 
branous interspaces. 

In Lunda cirrhata there is an almost identical syrinx, 
but the last tracheal rings are not particularly thin in front. 

Fratercula arctica and F. corniculata are sufficiently 
similar to need no special description. 

The tongue is generally fleshy, elongated, triangular, and 
spiny only at the base. 

The relative proportions of the Uver lobes not only vary, 
but the absolute size of the organ varies greatly. 

A gall bladder is always present. There is no crop. 

The cccca are usually mere nipples, -35--25 inch in length, 
but in Alca tordd one inch. The length of the small 
intestine in Fratercula arctica is 28'5 inches, in Alca tunla 
49 '5 inches. 

From the characters displayed in the accompanying 



table it seems possible to divide the group into two families, 
Uriidae and Fraterculida 3 , which maybe thus defined :- 


Rectrices, twelve or fourteen. Lobes of liver equal, 
or left larger than right. Muscle formula, 
ABX- (or AX-). 


Hcctrices, sixteen. Lobes of liver equal, or right 
larger than left. Muscle formula, AX+. 
Syrinx peculiar. 


Access* in 




t'riu tiiliimba ... 14 
Synthliborhamphus antii/iius 14 
Ceratorhina itionucerata 16 




J_j i\, 

Lin/i/a cirrhata ... 16 

Good . R>L 

Brachyrhamplms ni(trm<iritti/.< 14 
Phaleris psittacula . . 14 

I'nili.Tcnla tirct/i'ii . . 10 




Alcn ton/a .... 12 Present 
Urin trnili' .... Good 




With U-shaped first 
broiichial semi-ring 
With U-shaped first 
bronchial semi-ring 

With U-shaped first 
bronchial semi-ring 

There appears to me to be no doubt that the Alcae are 
best placed in the neighbourhood of the Limicolae, though, 
as FUEBEINGEE justly states, ' at first sight the relations 
between the two groups do not appear to be intimate.' 
These differences, however, merely concern outward form, in 
which it is perhaps reasonable to compare the auks with 
the grebes. But an anatomical study shows plainly that 
the grebes are much further away from the auks than are 
the Limicolae. Such points of likeness as there are with the 
Colymbi are largely, if not entirely, due to the similar life ; 
thus the elongated sternum, which is also shared by the 
aquatic ducks, and possibly the muscle formula ABX + . 
With the Limicolae are many positive points of likeness, to 
which no such explanation seems to be applicable. In the 
skull it is hard to find points of difference ; but the most 
remarkable point of similarity is the presence in both groups 
of those additional tendinous slips upon the patagium on the 


ulnar side (in Charadrius as well as Laridse) which have been 
duly described in the foregoing pages. Nor are there any 
salient facts, save such as are evidently associated with loss 
of the power of flight, which contradict such a placing. 


Definition. Oil gland present and tufted; ' feathers with an aftershaft. 
Rectrices, twelve. A quintocubital or quintocubital.'-' Ambiens, 
semitendinosus, and accessory always present. Expansor secun- 
dariorum present. Cseca large. 3 Skxill schizognathous, sehizo- 
rhlnal, without basipterygoid processes. Two carotids. 

Among the typical cranes of the family Gruidae I include 
not only the nearly cosmopolitan Grus and the African 
Balearica, but also the South American Aramus. 

There are no particular remarks to be made about the 
pterylosis, which NITZSCH states to be precisely like that of 
PsopJiia (see below, p. 374). 

The muscular system is fairly uniform in its characters, as 
will be seen from the length of the above definition. 

The tensores patagii of the demoiselle crane (G. virgd) 
are furnished with a muscular biceps slip, which is reinforced 
by a tendon springing from the biceps below the origin of the 
biceps slip. There is also the usual fibrous junction with 
the deltoid crest of the humerus. 

From the pectoralis 4 springs a broad flat tendinous slip, 
which joins the undivided tensor patagii. The tensor brevis 
divides at once into two thin broad diffuse tendons, of which 
the anterior sends forward a wristward slip, from whose 
junction with extensor metacarpi a slight patagial fan pro- 
ceeds to the ]ongus tendon. 

In Grus leucogeranos the tensor brevis tendon widens out 
shortly after crossing biceps slip into a wide diffuse band, 
composed of many strands, but not distinctly separable into 
two or three tendons. There is a patagial fan. 

1 Except in Mesitcs, Cariama, and Rliinoclietus. 

- Rhinochctus, Cariama, Psopliia. 3 Not in Eitri/pyga. 

4 The pectoralis I. is usually stated to be single. It appeared to me to be 
distinctly double in Grus carunculatus and in Balearica pavon ma, especially 
in the latter. 



A ravins scolopaeeus has the same thin diffused tendons ; 
but they are distinctly divided below into a main tendon and 
a wristward slip. There is no patagial fan. 

The anconcBus has generally, if not always, a well-marked 
broad humeral slip. 

The typical formula of the leg muscles for the cranes is 
ABXY + . This is the case with all the members of the 
genus Grus, excepting G. leucogeranos, where I could find 
neither A nor B. In Aram/is and Balearica pavonina the 
formula is BXY + , and in B. reguloruin, as in G. leucoge- 
ranos, XY+ only. In G. americana the femoro-caudal is 
minute and has but a feeble accessory. 

The deep flexor tendons are united by a strong vinculum. 

Both peroneals appear to be present ; but the only notes 
at my disposal on this matter refer to G. leucogeranos. 

The left lobe of the liver is much smaller in B. pavonina, 
a little smaller in G. ant i gone and G. virgo. The proportions 
are reversed in Aramits. 

The gall Madder is present ; there is a good gizzard ; the 
proventricidus is zonary. The following are intestinal mea- 
surements : 

,-Mlilll Tilt. 

Large Int. 





Grus antigonc 



7 and 8 






S05 (<J)86( , i 



mncricana 3 




ctnuith'tisis $ 




australasiana $ 

84 (73) 

6 (3-5) 

7-5 (7 and 8-25) 




2-5 and 3-2') 

r><ilcarica pavonina 9 



6-8 and 6-1 

,, reguloruin $ 


5 - 5 

Animus scolopaeeus 9 


2, 2| 

The intestinal coils in the crane tribe are very character- 
istic and quite unlike those of any other birds except the 
rails and bustards. The figure of Car-lama shows the 
characters of the Grues generally and may be compared with 
that of Crex on p. 323. 

The genus Grus has the most typical syrinx. In G. lenco- 
geranos the first bronchial semi-rings are firmly attached to 


each other, and the first two are ossified and somewhat arched. 
To the first of these are inserted on each side the two flat, rather 
broad intrinsic muscles, which run side by side, and which 
appear to be continued by fibrous tissue on to the second 
semi-ring. There is a normal pessulus. The membrana 
tympaniformis gets narrower from above downwards (having, 
therefore, a triangular form), and finally ends opposite the 
thirteenth semi-ring ; but the rings remain semi-rings after 
this point, though their ends are very closely approximated, 
until close to their opening into the lung. G. australasiana 
shows no special differences. In G. canadensis the two mus- 
cles, though distinct above, appear to fuse below ; they do not 
quite reach the bronchial semi -ring as muscle, bat are 
attached to it by a short ligamentous ending. Grus carun- 
culata agrees with the last. 

A peculiarity found in many cranes is the convoluted 
trachea. 1 This state of affairs is not found in Balearica or 

In both males and females of the following species the 
trachea is convoluted : G. cinerea, G. antigone, G. carun- 
culata, and G. leucogeranos. The males of G. australasiana, 
and G. canadensis are known to be the same, and the female 
of G. americana. In the female of G. leucogeranos and in 
the male of G. carunculata the trachea, though convoluted 
more or less, does not enter the substance of the sternum, as 
it does in the others. This too holds good for Tetrapteryx 
and Anthropoides. 

The trachea has the usual pair of extrinsic muscles, which 
in Balearica pavonina arise not from the costal processes, as 
is the rule, but from the angle of the first rib. 

I have myself examined syringes of the following species : 
Grus canadensis, G. australasiana, G. leucogeranos, G. 
carunculata, Balearica pavonina, and B. regulorum. 

The syrinx of Balearica is rather different and less 

The two intrinsic muscles are present, but they end in a 

1 See ' A Natural History of the Cranes,' by W. B. TEGETMEIER, and FOKBES, 
P. Z. S. 1882, p. 353. 

GKUES 369 

fibrous band fourteen rings above the end of the trachea. 
The first tracheal ring is not so strongly modified as in 

13. pavomna hardly differs. 

There are nineteen cervical vertebra in G. carunculata, 
twenty in Balearica. 

Seven ribs reach the sternum in both. The clavicles in 
the former and in Tetrapteryx are ankylosed with the sternum, 
but not in Balearica. Some of the dorsal vertebrae are partly 

The skullh&s occipital fontanelles, as in mostcharadriiform 
birds. This holds good also of the slightly aberrant Aramus. 
The impressions for the supra-orbital glands are slight, and 
largely concealed when viewed from above. The lacrymal 
bones do not blend with the ectethmoid. The interorbital 
septum is much fenestrated, but not so much so as in the rails. 
In Tetrapteryx and Balearica l the palatine bones do not 
appear to come into contact posteriorly, and at any rate the 
inner lamina is continued right to the end of the bone. 
This is not the case with Grus, where the bones do come 
into contact posteriorly and the inner laminae are not 
continued to the end. 

The pelvis of the t} r pical cranes (Grus, Balearica, Tetra- 
pteryx) hardly differs from that of such a rail as Ar amides. 

An outlying member of this group is usually included in 
the family Rhinochetidae. This family is represented by but 
a single species, the kagu (Bliinoclietus jubatus), of New 
Caledonia. The bird is not unlike a heron in appearance ; 
but BARTLETT, who made a careful study 2 of the habits 
of specimens at the Zoological Society's Gardens, compared 
its quick active movements rather with those of a crane 
than with the slow motions of a heron. The anatomy of 
the bird has been chiefly studied by PARKER (osteology), 3 

1 So too apparently in Aniliropoides stanlcyanus (PARKER, Tr. Z. S. x. pi. 
liv. fig. 0). 

- P. Z. S. 18(52, p. 218. 

3 ' On the Osteology of the Kagu,' Zool. Trans, vi. p. 501. 

B B 



MUEIE ('dermal and visceral structures'), 1 and myself 
(syrinx and muscular anatomy). 2 Others, however, particu- 
larly FUKBBINGEB and GAEROD, have contributed details of 
importance to our knowledge of this bird. 

The powder-down patches, which were originally dis- 
co'vered by BARTLETT, exist as scattered groups of feathers 
of the kind ; there are not the regular patches found in the 

=:*. st 

St, semitendinosus; A, its accessory ; Si, seniimembrauosus. 

near ally of Rhinochetus, Mesites. The oil gland is present 
but nude. The feathers have an aftersliaft. There are 
twelve rectrices. The pterylosis, imperfectly described by 

1 ' On the Dermal arrl Visceral Structure of the Kagu,' Zool. Trans, vii. 
p. 465. 

- ' Contributions to the Anatomy of the Kagu,' P. Z. S. 1801, p. t). See also 
W. MARSHALL, ' Quelques Observations sur la Splanchnologie de Rliino- 
clietus jubatus,' Arch. Neerl. 1870, p. 402. 

GRUES 371 

MUEIE, has been rather more fully dealt with by FORBES. 
The dorsal tract is double on the neck, and continues so 
until its termination about on a level with the scapulae. 
The posterior portion of the dorsal tract is not continuous 
with the anterior portions ; it terminates with a slight bifur- 
cation anteriorly and is widely dilated mesially. The ventral 
tract is broken into two by the intervention of powder-downs, 
and the pectoral branch is perfectly separated from the 
main tract, a unique feature, save for Mesites. It is the 
scattered powder-downs which are apparently responsible 
for much of the breaking up of the pterylse of Bliinochctus. 1 
The semitendinosus,its accessory, the femora-caudal, and 
the ambiens are all present in the kagu. 
As in Psopliia and some other birds, the 
semitendinosus is inserted in common with 
the semimembranosus. The relations of the 
last muscles and of the gastrocnemiiis are 
illustrated in the accompanying figure (fig. 
17iS), which w r ill explain itself. Both pero- 
neals are present, and have the typical 
arrangement seen when both muscles are 
developed. The deep plantar tendons are as 
shown in the figure (fig. 179). The flexor FIG. no. DEEP 

7 77 v .1 i in -i FLEXOR TENDONS 

liuUucis supplies the hallux alone, and is OF 

tied to the flexor communis by a strong ( AFTER BEDDARD). 
vinculum before the trifurcation of the 
latter. The mode of insertion of the tensor patagii brevis 
is complicated ; the tendon divides into three branches, 
the two inner of which are prolonged some way beyond the 
tendon of the extensor metacarpi radialis longi, to which 
they are first of all attached. There is a biceps sJi}).~ 
The anconceiis long us has a flat tendon of origin from the 
humerus, as well, of course, as its scapular head. A muscle 
apparently peculiar to Bhinochetus (see fig. 180) is what 
has been termed by me an ' accessory biceps.' This arises 

1 In his paper on Mcsitcs, P. Z. S. 1882, p. 267. 

2 I wrongly asserted the absence of this in my paper upon the anatomy of 
the bird. 

I! H 2 



from the humerus just below the insertion of the deltoid, 
and is inserted near to the insertion of the biceps. The 
e.cpansor secunda riorum is present. 

The syrinx of Bhinochettis is tracheo-bronchial, and 
presents us with no features of special interest. The 
accompanying drawing (fig. 181) shows its lateral aspect. 
The bronchidesmus is incomplete ; the intrinsic muscles are 
attached to the third bronchial semi-ring. 

The number of cervical vertebra is sixteen. Four of the 


liJiinochetus (AFTER BEDDARD). 

I), D2, deltoid ; Ldl, LAI, latissimus dorsi ; Bil, biceps ; 
Jii2, accessory biceps ; Jf, nerve. 

Rhinochetus (AFTER 

last dorsal vertebra are ankylosed. Five ribs articulate with 
the sternum. The sternum is unnotched. The skull is 
schizorhinal ; there are no basipterygoid processes. There 
is a partial bony internasal septum not to be found in the 
cranes. The interorbital septum is more fenestrate than in 


them. The palatines are abruptly truncated posteriorly, as 
in the herons. 1 There are small occipital foramina. 

An abnormal member of the crane 2 group is the South 
American seriema, of which it is usually considered that there 
are two genera, Cariama and Chung a, of the family Cariamidse. 
These birds agree with the cranes 
in possessing an aftershaft and in 
the number of their rectrices 
(twelve). Theo'iZ gland, however, 
is nude. In the pterylosis (which 
has been described by NITZSCH) 
there is a marked break between 
the posterior forks of the anterior 
section of the dorsal tracts and the 
anterior fork of the posterior sec- 
tion of the same tracts. 

The dorsal tract is single on 
the neck and divides interscapu- 
larly. The posterior parts of the 
ventral tract are formed of two 
rows about two feathers wide. 
Each joins the outer branch above 
by one row of feathers merely. 

The skull (fig. 182) is des- 
mognathous, but the two maxillo- 
palatines, though they come into 
contact in the middle line, are not 
fused. The nasals are appa- 
rently holorhinal, really schizo- p IG 132. SKULL 
rhinal (see p. 144), and there are no VENTRAL VIEW. (AFTER BEDDABD.) 
basipterygoid processes. There are p ' palatine ; 6 ' 8U P raorbital ri ' l * c - 
fifteen cervical vertebra ; five ribs articulate with the sternum, 
which is one-notched ; it has the spina externa. 

1 Stress has been laid upon this fact and comparison, but a posterior trun- 
cation of the palatines, nearly as marked, is to be seen in Fratercula arctica 
and not in some other auks. 

The skull is described by PAUKER, 2V. Linn. Soc. (2), i. p. 128 ; the 
general osteology and to some extent the visceral anatomy by BTRMEISTER, 

OF Chung a 


The muscle formula of the leg is BXY-f, as in bustards. 
The accessory femoro-caudal muscle (of Chimga) is peculiar 
in that it becomes reduced in the middle to a thin tendon, 
being muscular at both extremities. Both peroneals are 

The tensores patagii spring from a single muscle. There 
is no biceps slip, another point of likeness to bustards. The 
brevis tendon spreads out into a broad aponeurosis, but there 
is no patagial fan. The anconceus has a humeral head. 
The intestinal measurements of the two birds are as 
follows :- 

Cariama cristata. Chung a Burmeisteri. 
Small intestines, 33 inches. 33 inches. 

Large intestine, 3 ,, 3*5 ,, 

CSBC-A, 8-75 ,, 8-5 and 1O5 inches. 

The family Psophiidse is represented by the single South 
American genus Psophia, including some four species. 
These birds have the outer aspect of a rail rather than of a 
crane, and PAEKEE has commented upon their ' phasianine ' 
expression of face. Nevertheless their nearest alliance 
seems to be with the crane tribe, and perhaps more 
especially with Cariama. 

The pterylosis has been described and figured by NITZSCH. 
There are apparently ten rectrices (not twelve, as NITZSCH 
stated), and the oil gland, as in Gr-us, is tufted. The dorsal 
tract is single on the neck and forms a strong bifurcation 
between the shoulders ; from the two ends of the fork a 
single row of feathers descend and unite to form a weakly 
feathered but widish posterior part of the dorsal tract. The 
ventral tract bifurcates early in the neck, and each in the 
pectoral region gives off a strong band on the outside ; the 
main portion of each tract is continued on to the cloaca by 

' Beitrage z. Naturgeschiehte der Seriema,' Abliomll. not. Ges. Halle, i. (1854), p. 
17 ; the viscera also by GADOW, 7. /. O. xxiv. (1876), p. 445, and by MARTIN, 
P. Z. S. 1836, p. 29. See also BEDDARD, ' On the Anatomy of Burmeister's 
Cariama,' P. Z. S. 1889, p. 594, and literature there quoted. 

GRUES 375 

a very narrow band of feathers, which is only one feather 
wide to begin with, and afterwards only two feathers wide. 

The tensor patagii muscles are distinctly grume ; the 
biceps slip is present. 

In the hind limb the muscle formula is BXY + , as in 
Cariama and the bustards. 

The syrinx, shown in the accompanying woodcut (fig. 
381), presents no remarkable features. It is quite typically 
tracheo-bronchial, and has, as will be observed, an incom- 
plete bronchidesmus. It has been stated 
(by TRAIL) that the windpipe communicates 
with an air space, apparently after the fashion 
of the emu. But there is no doubt that this 
statement was based upon some imperfection 
of the example studied. It has been also 
stated that the windpipe in the male is con- 
voluted ; this requires confirmation also. 

The skull of Psophia ' is schizognathous Fm 183 _ SYHINX 
and holorhinal (fig. 81, p. 143). AsPAEKEE OF Psopitia leu- 
first observed, the orbital margin is furnished BEDDAED). 
with about five smallish supra-orbital bones, 
a feature which reminds us of certain archaic birds, as the 
tinamous, Arboricola, and Menura. The lacrymal has a de- 
scending process, which is swollen and nearly comes into 
contact with the ectethmoid. The maxillo-palatines are 
comparatively large and swollen bones ; as in Cariama 
these bones are convex on the outer side, and not concave- 
as in Grits. There are no occipital foramina. It may be 
remarked that the holorhinal nostrils of this bird show no 
such approach to schizorhiny as is displayed by Chunga. 

From the anterior part of the maxillo-palatines, on a 
level with a point just in front of the commencement of the 
bony nostrils, a stoutish knob of bone 2 projects inwards on 
either side. Of this there are traces in the cranes, parti- 
cularly in Tetrapteryx. If these processes were to be 
increased in size and to meet a bony internasal septum, we 

1 P. E. BEDDARD, ' On the Structure of Psapliia,'' &c., P. Z. S. 1890, p. 329. 
-' Duly referred to by PAEKEE, ' Osteology of the Kagu,' Tr. Z. S. vi. p. 507. 


should have the ' desniognathous ' skull of the American 

P sophia has seventeen cervical vertebra, of which the 

last bears a rudimentary rib. Five dorsal vertebrcc are 
ankylosed, there being two free ones behind. The stern it in 



eight ribs 


(fig. 184) is entire and unnotched 
with it. 

The atlas is notched for the odontoid process. From 
the fourteenth cervical vertebra to the third dorsal there are 
blade-like median hypapophyses. In front of the fourteenth 
the catapophyses nearly enclose a canal ; they get further 
apart and die away anteriorly. The following table shows 
the number and character of the hypapophyses in various 
Grues : 








Last on Oil 













The family Eurypygidse contains but one genus and 
species, Eurypyga helias, native of South America. It has 
an oil gland, which is generally nude but occasionally tufted, 
and twelve rectrices. Eurypyga, like Rhinochetus and 
Mesites, has powder-down patches, but their arrangement is 
very different from those of Rhinochetus. Dorsally there is 
on either side of the dorsal tract a compact dense triangular 
patch ; in front it continues over scapula as a band which 
runs on to the sternal surface, and there forms a sparsely 
feathered patch more or less continuous with pectoral tract 
of contour feathers. There are a few scattered powder- 
downs on axilla and along neck. 

The tensor patagii brevis is broad and rather diffused, 
stronger at the two edges ; it sends off a wristward slip. 
The tensor longus is reinforced by a strong biceps slip. 

The expansor secundariorum is strong and 'ciconiine.' 
The ancoiiicus has a humeral attachment. The insertion of 
the deltoid extends halfway down the humerus. I have 
noticed in the pectoralis primus a vertical septum dividing 
the muscle into a right and left half. 

The muscle formula of the hind limb is complete, i.e. 
AJBXY+. The glutceus I. extends well over the biceps. 
Both peroneals are present. 


The liver is equilobed, with a gall bladder. The intestines 
are 18 inches long, the short caeca ( inch) being li inch 
from cloaca. 

Both carotids are present. 

The shidl of the sun bittern has been described 1 and 
figured by PAEKEE. 

It presents several points of likeness to that of Rkino- 
chetus, notably in the ardeine character of the palatines, 2 
which are cut off squarely behind and are of approximate 
length throughout ; each palatine, moreover, has a fenestra, 
as in Tigrisoma leucolophum (and also in Numenius phceopus 
and Anous stolidus). The interorbital septum is widely 
fenestrate ; there are no occipital foramina. As in Psophia 
(q.v.), Rliinoclietus, and cranes, there is a rudimentary ' snag' 
from the anterior part of maxillo-palatine. 

The nostrils are schizorhinal, and the curves of the 
various surfaces of the bones are such that if the very 
narrow anterior chink were closed a well-rounded and quite 
typical holorhinal skull would be the result. 

Eurypygah&s a one-notched sternum with well-developed 
spina externa. There are eighteen cervical vertebrae, and 
three dorsals are fused. 

The pelvis is a little less rail-like than in Grits, Rliino- 
chetus, Psophia, &c., in being wider, and in the more hori- 
zontal plane of ilia, which do not meet. 

Family Aptornithidae. 3 The two species of Aptornis, A. 
defossor and A. otidiformis, from New Zealand quaternary 
deposits, were originally referred to the Dinornithidse, and 
more lately to the rails. F QEBEINGEE has, however, advanced 
certain reasons for relegating them to the neighbourhood of 
Rhinochetus, and I follow him in placing them in the present 
group. The chief reason which persuaded FUEBEINGEE to 

this conclusion was the schizorhinal nostrils, quite evident 

1 ' On the Osteology of the Kagu,' Tr. Z. S. vol. vi., and ' On /Egithogna- 
thous Birds,' ibid. vol. x. p. 307, pi. liv. figs. 7, 8, 1). 
- See, however, footnote, p. 373. 
3 OWEN, ' On Dinornis,' pt. xv. Tr. Z. S. vii. p. 353. 

({RUES 379 

in OWEN'S plates, 1 and showing the inward curvature so often 
found in the schizorhinal nostril, and quite apparent in Gnt.s 
(though Jioiin. Rhinochetus) . A special point of resemblance 
to Rhinochetus among the crane-like birds seems to me to 
be in the partial ossification of the nasal septum. The 
solidity, posteriorly at any rate, of the interorbital septum is 
like Psophia so far as gruine birds are concerned, while the 
spout-like process upwards of the palatines is quite in 
harmony with FURBRINGER'S views of the affinities of Ap- 
tornis. The junction of the zygoma with the post-frontal 
process is not crane-like ; it occurs among gallinaceous birds, 
and there is a near approach to it in Otis. 

In the view given by OWEN of the under surface of the 
skull is a bone described, though not figured, which ap- 
pears to me to correspond to the desmognathous palate of 
Cariama. The union of the bones and their divergence 
posteriorly are precisely like what is to be seen in Cariama. 

Large basipterygoid processes are present, but OWEN 
failed to find upon them an articular surface. 

So that while the outline of the skull of Aptornis is very 
like that of some of the large rails its affinities have been 
probably more correctly diagnosed by FURBRINGER. 

Besides the Aptornithidae already mentioned other forms 
referable to the Grues have been obtained from Tertiary 
strata. Of these Aletornis (with a number of species) is placed 
among the Gruidae and Gcranopsis of LYDEKKEB. The latter 
is known only by the coracoid, which differs somewhat from 
that of Grus. LYDEKKEK does not admit the genus Pal&ogrus of 

The family Mesitidae is represented by the Madagascar 
Mesites, a genus containing but a single species, which has 
been investigated anatomically by MILNE-EDWARDS 2 and by 
FORBES. 3 One of its principal characteristics was originally 

1 Loc. cit. pi. xl. figs. 1, '2. 

2 ' Remarques sur le Genre Mcsitcs,' &c., Ann. Sci. Nat. (6), vii., and in Hist. 
Nat. dc Madagascar. 

3 ' Description of the Pterylosis of Mcsitcs,' &c., P. Z. S. 1882, p. 267. 


discovered by E. BARTLETT/ who found and described 
briefly the powder-down patches. 

The bird has sixteen rectrices, and apparently but there 
is some little doubt about the matter a nude oil gland. The 
contour feathers have 110 aftershaft. There are five pairs of 
powder-down peddles. The most anterior pair lie in the 
iriterscapular region, and are enclosed by the dorsal tracts. 
The second pair are upon the rump, the third pair at the 
commencement of the pectoral region ; the fourth pair lie 
also on the ventral region, but posteriorly ; the fifth pair, 
finally, are axillary. The number of these pairs is greater 
than in any known bird, and their definition and complete 
separation as distinct patches contrasts with the diffused 
arrangement characteristic of Rhinochetus and Eiwypyga. 

There are four apteria on the neck, since both dorsal and 
ventral tracts divide early. The dorsal tracts converge inter- 
scapularly, and then become much feebler, and are continued 
on to the Y-shaped posterior part of the tract. The ventral 
tracts cease altogether at the commencement of the pectoral 
region, but recommence behind the powder-downs. The 
outer branch is present, but is quite unconnected with the 
main stem. 

The muscle formula is complete, ABXY + . Both carotids 
are present. 

As FURBRINGER removed Aptornis from the rails and 
placed it in the present group largely on account of its 
schizorhinal nostrils, it is remarkable that he did not also do 
so with the present bird. The bony nostrils are, in fact, of 
the type that has been termed pseudo-holorhinal. They are 
rounded at their end, but elongated and curved inwards ; they 
are exactly like those of Glareola. 

Mesites is schizognathous, with delicate maxillo-palatines. 
The descending process of the lacrymal abuts upon, but 
does not fuse with, the very stout square ectethmoid. This 
part of the skull, again, is more like Glareola than any grume 
form ; but it is also like Pterocles and various other birds. 

There are seventeen cervical vertebra, and four ribs reach 
1 ' Remarks on the Affinities of Mcsitcs,' P. Z. S. 1877, p. 292. 



the one-notched sternum. 
plete ribs are ankylosed. 

The dorsal vertebrae with com- 
The furcula is quite degenerate. 





"H i __ g 







3 - ~- 

~ ! 


^- -^ 


P be 



S S : - 

- i 










: : 
i C 

t, ^ 

~* " 

O o 
: ^f 

t> 2 

















Gruidae . 





+ + 







BXY + 

XY + 








15 None 


BXY + 


XY + 

Rhinochetidae . 







lucompl. 1C 



AXY + 


Eurypygidae . 


N., t. 




18 1-3 











17 1-4 











17 2-4 

Sol. BXY + 



+ i " 


+ i 



15,14 None 









Incompl. 16,17 



BXY + 





n Hoi. 


Incompl. 1G 







I1XY + 

1 Not completely fused, and not always. 

It is clear from the accompanying table that the seven 
families which I here include with the Grues are a tolerably 
divergent series of birds. Yet it does not appear to me 
possible to locate any one of them elsewhere. The bird 
concerning whose position I am most doubtful is naturally 
Mesites. It is placed near the hemipodes among the galli- 
naceous birds by GADOW, and by FURBEINGEE near the 
hemipodes but among the rails. SHAEPE takes the view 
which is urged here, while FOEBES and some other recent 
writers are impressed by its likenesses to Eurypyga and 
Rhinochetus, FOEBES, indeed, having associated all three in 
a separate group. There are, unfortunately, so many lacunae 
in our knowledge of this form that a strict comparison is as 
yet hardly possible. Allowing the characters of the deep 
flexor tendons, not mentioned by FUEBEINGEE, the agree- 
ments with the Turnices do not appear to me to be more 
numerous than those with Eurypyga, while the powder- 
down patches are unknown in either rail or gallinaceous 

There are difficulties too with other genera of crane-like 


birds. For PARKER the inclusion of Cariama in the present 
group is impossible. Yet I find myself in this matter in 
accord with most recent writers. If Ca riant a is not allied 
to the cranes, where are we to put it ? The only alternative 
seems to be the rails, which are in any case not far removed 
from the present group. MITCHELL has pointed out the 
very close resemblance in the intestinal tract of cranes, rails, 
Cariama, Otis, and (as yet unpublished) Psopkia. PARKER, 
on the other hand, has emphasised the likeness between 
Cariama and the Accipitres, a likeness which has even im- 
pressed itself upon their external physiognomy. FORBES 
went so far as to include in the same group with Cariama 
the secretary bird. 1 It appears to me, in fact, that the 
origin of the Accipitres is to be traced to some crane-like form, 
and that the very varied characters of the Grues point to 
their being a basal form connected with more groups than 
one. As is pointed out elsewhere, unless we are to regard 
the Accipitres as here treated as diphyletic, or even triphy- 
letic, we must assume that the Cathartse and Serpentarius 
are the most primitive forms w r heiice the typical Falconidee 
have been derived by a loss of two mtiscles of the leg. In 
these birds the formula BXY or AXY must come before A 
alone. But little change is required to convert Cariama into 
Gypogeranus ; and if it be objected that this is because 
Cariama is an accipitrine, the quite similar skull, so far as 
the desmognathism is concerned, of Aptornis may be pointed 
to ; it will hardly be contended that Aptornis is anything but 
a crane or a rail. Along another line probably the peculiar 
desmognathism of the American vultures may be derived 
from conditions obtaining in the crane group. PARKER has 
pointed out the ossification and fusion in the middle line of 
alinasalsin Cariama ; but in that bird the fused bones, though 
quite analogous with, cannot be the exact homologues of, 
the fusion of bones which has occurred in Catkartcs, &c. ; 
for the position is totally different, being much further 
forward in Cariama. There is, however, in Psopkia and in 
others of its allies particularly well developed in Psophia 

' ' A raptorial isomorph of the Cranes ' (PARKER). 

GRUES 383 

a small snag of bone, already referred to as projecting out- 
wards from the maxillo-palatine in precisely the spot required ; 
if this projection were to grow more extensively, we should 
have a palate exactly like that of Catliartes and Gypagns. 

To consider another quite different group of birds, the 
Herodiones, I have urged on another page (p. 441) the low 
position of PI a talc a among the Herodiones. In more than 
one particular these birds recall the cranes. The occipital 
fontanelles, the complete muscle formula are among these 
characters. FURBRINGER has hinted at, but not accepted, 
the diphyletic origin of the Herodiones ; were it not for Scopus 
and Balceniceps, this might be fairly assumed indeed ; and in 
view of this possibility the strongly ardeine character of the 
palatines in Rhinochetus and Eunjpyga may be commented 
upon. 1 This character, too, is coupled with the existence of 
powder-down patches. 

Though the charadriiform birds form, in my opinion, a 
group distinct from the Grues, there are many points of simi- 
larity between them. This is practically shown by the fact 
that (Edicnemus has been much bandied about between the 
two groups. The occipital fontanelles, the supra-orbital 
impressions are among the points of likeness. In view of 
these facts the likeness in some matters of the skull of 
Mesites to that of Glareola is perhaps significant. It is 
really mainly the form of the intestinal coils as described by 
MITCHELL that leads me to dwell upon the separateness of 
the two groups. But this question is dealt with more fully 
under Limicolse (on p. 358). 

As an appendix to the Grues we may perhaps consider the 


This singular group of birds was originally discovered in 
the lower Tertiary strata of Patagonia by the well-known 
palaeontologist AMEGHINO. Of the several genera assigned 
to the family, many of which are very imperfectly known 

1 See, however, p. 373 footnote, where a similar conformation of these bones 
in the auk, Frati-mtla, is described. 


and doubtfully referable to the group, 1 Phororhacos is the 
best known, almost the entire skeleton being now in the 
museum at Buenos Ayres. A convenient summary of AME- 
GHINO'S work, with criticisms, has lately appeared in the 'Ibis,' 2 
by Mr. ANDREWS, from the whose paper the information here 
has been extracted. When first discovered (the anterior por- 
tion of the lower jaw) the creature was referred to the mam- 
malia. As with the leg bone of the Dinornis its ornithic 
nature was doubted, though the lesson afforded by the 
Dinornis might have tempted naturalists to be bolder than 
was perhaps reasonable in the forties. The bird was formerly 

FIG. 185. SKULL OF Phororhacos. LATERAL ASPECT. 

held to be ratite. But LYDEKKER, who examined the qua- 
drate, found that, as in carinates, it articulated by two heads 
instead of the one which characterises that bone in the 
ratites. 3 

The skull of Phororhacos longissimus is two feet in 
length ; the hooked beak has two or three serrations at the 
commencement of the hook, which remind one of the Eocene 
Odontopteryx, in which bird, however (see p. 418), the serra- 
tions extend along the entire length of both jaws. Seen 
from above certain resemblances between this skull and that 
of Phalacrocorax or Plotus will be apparent. ANDREWS 
thinks that the lacrymal was united with the jugal by a 

1 E.g. Bmntoniis, Dryornis, Pelecyornis. 

2 Jan. 181M3. 

3 Except Apteryx. 


separate bone which exists in Cariama (and of course 
Chunga), a bone which is found in other birds, though its 
articulation with the palatines often makes its identification 
with that of the seriema doubtful. 

The ventral surface of the skull is crane-like. 1 
The vertebral column is interesting on account of the 

fact that many of the dorsal and all 
of the caudal s have their centra per- 
forated for the remains of the noto- 
chord. There is no pygostyle. The 
pelvis is remarkably like that of Hes- 
perornis and the grebes. It has the 
length and narrowness of that limb 
girdle in the birds mentioned. It 
has, however, been more satisfac- 
torily perhaps compared with the 
pelvis of Cariama. The slenderness 
and length of the coracoid, together 
with the absence of a procoracoid, 
prevent a comparison in this particu- 
lar with the ratites. It is typically 

FIG. 186. SKULL or Plio- 
rorliacos. DORSAL ASPECT. 

FIG. 187. PELVIS OF Pliororliacos. DORSAL, 

carinate in fact, and shows perhaps special resemblances to 

The ulna shows well-marked tubercles, which indicate the 
insertion of the secondaries. Though reduced in size, and 
therefore possibly useless for purposes of flight, it seems 

1 See letter from Dr. GADOW, Ibis, Oct. 1896, p. 586, where it is pointed out 
that Stcreornis is synonymous with Pliororliacos, and that therefore the name 
Stereornithes cannot stand. It is, however, difficult to compose a name out of 
Phororliacos, and it is not proved that the birds in question are definitely 
cranes. I therefore leave the name. 

C C 


likely that the wings of this bird were more efficient than 
those of the ratites on account of the apparently well- 
developed remiges. 


Definition. Oil gland tufted. Afiershaft present. Aquiiicubital. 
Accessory semitendinosus absent Biceps slip present. G-lutseus 
maximus large, extending behind acetabulum. Caeca long. 
Skull holorhinal, without basipterygoid processes, schlzogna- 
thous. Tibial crest strongly developed. 

This group of birds contains two very well-marked families, 
the divers (Colymbidse) andthe grebes (Podicipedidse). In view 
of their numerous and important points of similarity I have 
not thought it desirable to separate these two families quite 
so widely as has GADOW. The Colymbidse contain but one 
genus, Colymbus, with four species. Of the grebes there is 
perhaps also only one well-marked genus, Podicipes. But 
the Central American Centropelma (of SCLATEE and SALVIN) 
has some claims, on account of the complete loss of flight, to 
generic distinction, while in the course of the following pages 
it will be seen that there are certain, if small, reasons for 
distinguishing jEclimopliorux and Tac/iybaptes. Podilyinb/tx, 
another alleged genus, has not been dissected. 

Our knowledge of the anatomy of this group of birds is 
chiefly due to NiTZSCH, 1 BEANDT,* CoiiES, 3 SnrFELDT, 4 and 
myself. 5 

As to external characters, there is a close agreement 
among the Colymbi. The number of rectrices in Colymbtis 
glacialis I find to be twenty. NITZSCH gives eighteen to 
twenty for the genus. Specialised rectrices are not recognis- 
able among the grebes. The inferior tract of feathers is 

1 Loc. cit. 

' l ' Beitriige z. Kenntniss d. Naturg. d. VSgel,' Mem. Ac. St. Peterxb. 1840, 
p. 197. 

a ' On the Osteology of Colymbtt*,' Ac., Mi-m. Bost. Soc. Nat. Hist. i. 1866, 
p. 131. 

4 ' Concerning the Taxonomy of the N. American Pygopodes,' Ac., J. Anat. 
1' lii/ft. 1892, p. 198. 

5 ' Notes upon the Anatomy of a Grebe (JEchmoplwrus major),'' etc., P. Z. S. 
1H%, p. 538. 


divided into two about halfway down the neck ; the two 
tracts are not again divided upon the trunk, where they are 
broad. The spinal tract divides high up on the neck (Pocli- 
ct'px crixhthis) or only between the shoulders (Columbus 
glacialis). The anterior part of the spinal tract is stronger 
than the posterior part, and is separated from it ; the latter is 
solid, enclosing no space. 

The patagial tendons of Colymbus glacialis are rather 
simple. The tendon of the brevis is a rather broad undivided 
band. There is no patagial fan, but, as in ^Echmophorus, a 
delicate tendon arises from the fore arm near the insertion of 
the tensor brevis and ends in the biceps slip. In CoUjmbus 
arcticus, according to FUBBRINGER'S sketch, a broad diffuse 
tendon arises from biceps slip, and ends freely upon the 
patagium. The first description given is from my own dis- 
section of C. glacialis, and agrees with a manuscript note 
of Mr. FORBES upon C. septentrionalis. But C. glacialis 
apparently sometimes approaches C. arcticus. I have a manu- 
script sketch by Professor GARROD showing a broad band of 
fibres arising from the biceps slip, but ending on the fore arm. 

The anconceus has a humeral head. The expansor sccun- 
(1 a riorum appears to be absent. The biceps is single-headed. 

The leg muscle formula, in contradistinction to what we 
find in the grebes (where it is BX ), is ABX+ . 

There is a peculiarity in the gastrocnemius of C. septen- 
trionalis which is worth calling attention to : one of the heads 
of that muscle arises from the tendinous end of the glutens 

The femorocaudal is tendinous for half its length ; its ac- 
cessory is a very large muscle. The ambiens in Colymbus 
glacialish&B two heads of origin. 

The combined plantar tendons give off a small slip to the 
hallux in C. septentrionalis, as in Podicipes minor. 

The divers have two carotids, the more modified (cf. leg 
formula) grebes only the left. 

The lobes of the liver in the Colymbi are equal, and there 
is a gall bladder. 

The following are a few intestinal measurements :- 

c c 2 
































1-1, 1- 









The ccfc> of C. septentrionalis are conical and saccular in 
form, with irregular but distinct transverse rugae. 

The syrinx in the divers (Colymbus septentrionalis) is not 
in any way remarkable in form. The last tracheal rings and 


i, intrinsic muscles. 

Km. 189. SYEIXX OF 

i, intrinsic muscles. 

the first bronchial are ossified and firmly attached to each 
other. The pessulus is also ossified. The succeeding bron- 
chial rings are soft and cartilaginous. The powerful intrinsic 
muscles are inserted partly on to the last tracheal ring and 
partly on to the first bronchial, the line of insertion being 

The syrinx of JEchmophorus (fig. 188) has a very incom- 
plete bronchidesmus, a very wide space between the two 
bronchi existing above its anterior edge. The last two 
tracheal rings are fused to form a long box, into the compo- 
sition of which it appears to me that the first bronchial semi- 
ring enters. In any case, if that be not so, the first bronchial 


semi-ring has the unusual relations shown in the drawing, 
which are perfectly consistent with the belief that the ring is 
the second bronchial. The intrinsic muscles are attached to 
the third tracheal ring in front of the tracheo-bronchial box. 
The bronchial semi-rings are fairly ossified, but have rather 
wide membranous interspaces. 

In Podicipes cristatus there is the same failure of the 
intrinsic syringeal muscles to reach even the end of the 
trachea. A box is formed by fusion at the end of the trachea, 
into which it appears to me the first bronchial semi-ring does 
not enter. The bronchial semi-rings are deeper and closer 
together, and the whole bronchus is more ossified, than in 
the last genus. The bronchi, too, are longer. 

In Podicipes cormttus the syrinx is much the same, but 
of course smaller. The first free semi-ring of the bronchus 
seems to be No. 2. There is a wider membranous interval 
between it and the antecedent tracheo-bronchial box than in 
the last species. 

Tackybaptes ji.uviatilis (fig. 189, p. 388) has a different 
syrinx. The last three tracheal rings are only fused in front, 
though they are closely united laterally. These rings are 
much ossified. The insertion of the intrinsic muscles is 
remarkable. They run obliquely forward, converging, to be 
inserted into the last three tracheal rings. The first bron- 
chial semi-ring is arched, and ossified in front, where it is 
fused with the tracheal box ; otherwise it and the succeeding 
rings are cartilaginous. It is clear, therefore, that the 
syringeal characters justify the generic distinction here 

The cervical vertebra are more numerous in the grebes 
than in the divers ; they are only fourteen or fifteen in the 
latter, twenty-one in ^Echmophorus and Podicipes conni- 
tus. Cervical vertebrge 10 to 16 in ^Echmophorus ' have 
a catapophysial canal; on 17 and 18 the hypapophyses 
are blade-like and enormous ; these processes extend to 
the very end of the dorsal series. The last cervical and 
the first three dorsals are fused. There is no catapophysial 

1 A canal is nearly formed in Podiceps, 



canal in Colymlus ; but the hypapophyses are greatly deve- 
loped, and in the dorsal region Y-shaped, with widely diver- 
gent flattened and expanded limbs. None of the dorsal 
vertebrae are fused. Six ribs reach the sternum in the two 
grebes mentioned ; eight or nine in the divers. The stern inn 
is one-notched ; in the grebes it has in addition a median 
triangular notch. There is no anterior spine to the sternum 
in the grebes. The procoracoid is moderately large and 
hooked in the divers, absent in grebes. The pelvis is very 
elongated and compressed, as in Hespcroniis, but the ischia 
are not free from the ilia, as in that bird. 

In the skeleton of the leg the most conspicuous feature 
is the highly developed cnemial crest. The patella too is 
very large in the grebes, but not in Colymbidae, where, indeed, 
it is not ossified. In these particulars the grebes resemble 
Hesperuniis. The principal skull characters have been 
already mentioned in the definition of the group. In addition 
to these matters the strongly marked temporal fossae may be 
mentioned, which nearly meet on the upper surface of the 
skull. They are not so well marked in the small Podicipex 
i/iiiwr. In the divers are strongly marked furrows for the 
supra-orbital glands ; this is not the case with the grebes. 
The former have also a single median occipital foramen 
above the foramen magnum. The ectethmoids exhibit the 
bullate form so characteristic of the Anseres. 

It may be useful to state in a tabular form the principal 
characters of the grebes and divers. 1 


( 'iirotids 


Dors. Vert. 

No. of 


mil* nosus 


ABX + 




Ankylosed 4-f> 
Not mik. 8,9 


These characters appear distinctly to point to the more 
modified structure of the grebes. There are reasons (p. 165) 
for regarding a small number of cervical vertebrae and a 

The extinct Colijmboidcs (C.minnfun, M.-Ei>. ; C. anr/Ucus, LYDEKKEB), of 
which the former is known by a humerus, the latter by a coracoid, and possibly 
a piece of sternum, is said to combine the characters of grebes and divers. 


large number of complete ribs as more archaic characters 
than the reverse ; the other features as evidence of degenera- 
tion require no comment. 

In considering, then, the affinities of the Colymbi the 
Colymbida? are to be chiefly taken into account. But any 
comparisons bristle with difficulties. The late Mr. FORBES 
in his final scheme of classification definitely associated the 
Colymbi with the Heliornithidae ; in this course I supported 
him by reason of certain muscular characters of both groups 
of birds. The outward appearance, too, of both birds is not 
at variance with such an affinity. The muscular formula of 
the leg is the very unusual one of ABX + . This reduced 
formula is found in Podoa and in the Sphenisci, Anatida?, 
Tubinares, Plialacrocorax, and certain Alcidae. It is to be noted 
that all of these are at least largely aquatic in their habits, a 
fact which must, of course, discount the value of the character ; 
but still ABX is an unusual formula, and there are other 
grounds for regarding the birds mentioned (with the excep- 
tion, perhaps, of the Alcida?) as having some relation to each 
other. The insertion of the biceps slip on to the patagium 
allies Colymbus and Podoa ; though this also occurs in other 
birds, it is, again, in some that are presumably not far from 
the Colymbi : for example, of the forty characters selected 
by GADOW for the comparison of the various groups of birds 
the Colymbidae agree with the Sphenisci in twenty-eight ; 
with the Tubinares in twenty-seven ; with the Steganopodes in 
twenty-six ; with the Anatidae in twenty-three ; with the 
Heliornithidaj in twenty-three. 

On the other hand there are twenty-five points of likeness 
to the Laridse, as deduced from the same tables, and no less 
than twenty-nine to the Alcse. As might indeed be imagined, 
it would be rash to lay much stress upon the proportions of 
these numbers. It is not to my mind clear with which of 
the groups mentioned the Colymbi are most nearly allied. 
Their undoubted relationship to the Hesper or niches is treated 
of on another page (p. 395) ; and it is perhaps this very fact 
which prevents us from detecting likenesses to more modern 
(?) groups. 



Definition. Large extinct birds of a diver-like form. Skull holorhinal, 
with, supra-orbital impressions. Vomers paired. Quadrate 
single-headed. Sternum without keel. Shoulder girdle platy- 
coracoidal ; clavicles fully developed. Fore limb represented by 
Immerus only. 

The principal source of information concerning the 
remarkable genus Hesperornis, from the Cretaceous of North 
America (to which three species, viz. H. regalis, H. crassipes, 
.and H. gracilis, are assigned), is naturally the magnificent 
treatise on this bird (and on Ichthyornis) by MAESH. ' 

The bird stood about six feet high, and presented the 
general form of the diver, to which bird it is now held to 
come nearest. MAESH, however, compared it more especially 
with the Struthiones, and it is spoken of by him and by 
others as an ' aquatic ostrich.' The former view now holds 
the field. 

There are, nevertheless, various struthious features in 
the skull, and in other parts of the skeleton, some of 
which may be held to be due to its loss of the power of 
flight ; others are not so explicable from the point of view of 
our actual knowledge of bird structure. 

The skull has the general contours of that of the diver, 
and, like that bird, has very well-marked furrows for the 
supra-orbital glands. The nasal bones produce a holorhinal 
nostril. The figure given by MAESH of the upper surface of 
the skull is a little suggestive of there having been an ap- 
pearance of the ossified ethmoid on the surface of the skull, 
as in some struthious birds and tinamous. The interorbital 
septum is fairly ossified with a large foramen. The lacrymal 
is large with a large descending process, which nearly, if not 
quite, reaches the jugal bar. The pitting of the premaxil- 
laries led MAESH to the inference that the beak was present, 
in addition to the teeth, which will be presently referred to. 

The skull has been described as ' saurognathous,' on 

1 Odontornithes. Washington, 1880. 


account of the paired vomers. These were, however, not 
particularly like the presumed vomers of the woodpeckers 
(see p. 187). Each bone is broad behind, where it may, as 
in struthious birds, have articulated with the pterygoids, and 
tapers in front to almost a point. It is not clear whether 
the bird really had, as MARSH is disposed to infer, a ' dromseo- 
gnathous ' palate. In any case the basipterygoid processes 
are present, and the articulations on the pterygoids are 
towards the posterior end of these bones, as in the skull of 
the Struthiones. The palatines are longi&h bones, and are 
compared to those of the ostrich. They taper in front. As 
in most Struthiones the articular head of the quadrate, for 
articulation with the skull, is not divided into two facets. 
The rami of the lower jaw do not appear to have been 
ankylosed together, but to have been connected by a possibly 
merely chondrified or ligamentous tract, which would have 
allowed a gaping of the mandibles seen partly in the 
pelican, and obviously useful to a fish-eating bird, as we 
may presume Hesperomis to have been. Both lower and 
upper jaws have teeth, which are implanted in a continuous 
groove, widened at the implantation of each tooth. In the 
upper jaw the teeth are limited to the maxillas, and there 
were fourteen to each maxilla. The lower jaw had teeth 
along its entire length, and the number given is thirty- 

The vertebra are saddle-shaped. The number of cervical 
vertebras is seventeen. The entire vertebral column con- 
sisted of forty-nine vertebrae. None are ankylosed, except, of 
course, the sacral series, and some at the end of the tail. 
Pneumatic openings were not discoverable in any of the 
vertebra?. The atlas has not been found. In no vertebra 
do the catapophyses seem to have united to form a ventral 
canal, a state of affairs which is occasionally met with in 
the Struthiones and is characteristic of the Herodiones and 
Steganopodes. The fourteenth cervical is extraordinary by 
reason of the enormous size of the two catapophyses, which 
are approximated and nearly parallel. The following vertebra 
has the first median hypapophysis, which is bifid at the free 


tip ; the hypapophyses die away on the fourth dorsal 

With regard to the caudal vertebrae, the most remarkable 
fact is that while there is no pygostyle there is no rudi- 
mentary state of affairs observable ; for the last few vertebrae 
have greatly expanded transverse processes, which would not 
move independently. MARSH thinks that ' the end of the tail 
would move mainly as a whole. This would give great 
power, similar to that in the beaver's tail or the flexible blade 
of an oar.' 

As to the shoulder girdle, the scapula is in the same 
straight line, or nearly so, with the coracoid ; it belongs, in 
fact, to what FURBRINGER has termed the ' platycoracoidal ' 
type, seen also in the Struthiones. But, contrary to what we 
find in those birds, there is not a fusion between the cora- 
coids and scapula in Hesperoniis, and there is, moreover, a 
complete pair of clavicles. The coracoid has a strong pro- 
coracoidal process, and also a supra-coracoidal foramen, as in 
various Struthiones. 

The two coracoids are widely removed at their articula- 
tion with the sternum. The clavicles, though complete, 
appear to have joined each other ventrally by a joint ; they 
arise from the procoracoid. The sternum, which has articu- 
lar surfaces for five ribs, has no keel. It is notched in the 
middle line posteriorly, and is wider in front than behind. 
The elongated form of the sternum is compared by MAI:SII 
to that of Uria. The ribs have uncinate processes. 

The fore limb of. Hesperornis appears to have consisted 
only of the humerus, as no other bones were discovered, and 
as there were no distal facets for articulation with a radius 
and ulna. 

The pelvis of Hes'perornis in its general form resembles 
that of Podicepx. The constituent bones are, however, 
entirely free distally. The acetabulum is closed by bone, a 
state of affairs only seen in the emu and to some extent in 

The ischium has no processes tending upwards to the 
ilium and downwards to the pubis, as in Struthiones. 


The prepubic process is large, but not larger than in certain 
recent birds, and not so large as in Geococcyx and Tiiuuimx. 

' The femur of Hesperornis is remarkably short and stout, 
more so than in any known bird, recent or fossil.' The tili/i 
has an enormous cnemial crest, as in divers, and the patella 
is a huge bone ; the latter is perforated by a foramen for the 
ambiens muscle. The feet have four toes. 

Nothing is known about the soft parts of this bird, save 
the feathers, which have been lately l stated to be quite 
ostrich-like. But casts of the brain have brought to light 
the interesting fact that it has smaller cerebral hemispheres 
than are found in any existing bird. 

It seems clear that, as D'AncY THOMPSON and others have 
argued, the nearest affinities of Hesperornis are with the 
Colymbi. That they are more nearly related to the ratites 
was the opinion of MARSH. The likeness to the ratites, 
however, seems mainly to be based upon the degenerate 
structure of the wings in both. The degeneration of the 
wings, however, has not proceeded along precisely similar 
lines. Although the angle between the scapula and the 
coracoid is nearly as open as in the ratites, the two bones 
have not become ankylosed ; and moreover the clavicles are 
retained. The fact that of the fore limb only the humerus 
remains (at least in an ossified condition) may be compared 
with the fact that in the Dinornithidse traces of the same 
bone have been met with. On the other hand there are 
more positive likenesses to the Colymbi, which are not so 
clearly due to the immediate action of environment. The 
long and narrow pelvis, and the huge cnemial crest with the 
relatively enormous patella, are among the more salient points 
of resemblance. 

The characters in which the approach the 
Colymbi are divided between the two families of the latter 
group ; in some points Hesperornis is more of a diver ; in 
others it comes nearer to the grebes. The following table 
of comparison may help to make its ' mixed ' characters 
clear : 

1 In a letter to Nature, April 8, 1897, by Prof. MARSH. 


Hi >|irriiniis 

( 'erv. vert. . . 17 

14, 15 17--21 

Catapopli. canal .00 

lln'inap. . . - With expanded With expanded Not so expanded 

ends ends at ends 

Dors. vert, fused . . None None Several 

Patella .... Large Very small Large 

S/tjira-orbit. groan:* . Large Large Not marked 

With the grebes Hesperornis agrees in only two of the 
characters, and in the remaining four with the divers. It 
presents in fact, as might be expected of so ancient a form, a 
compound of the grebe and the diver. 

We cannot however, in my opinion, put it down defi- 
nitely as the ancestral form whence both divers and grebes 
have branched off ; but it seems to approach that form, 
agreeing as it does in most points with the more generalised 
divers. 1 


Definition. Moderately sized to large birds, with, wings modified to 
form a swimming paddle. Aftershaft present. Feathering 
continuous. Oil gland tufted. Muscle formula of leg, ABX + . 
Wo biceps brachii or expansor secundariorum. Two carotids. 
Skull schizognathous, holorhmal, without basipterygoid pro- 
cesses. Dorsal vertebrae markedly opisthocoelous. Scapula 
flattened and expanded. Metatarsals short and incompletely 

The anatomy of the penguins has been mainly investi- 
gated by MENZBIEE 2 and WATSON. 2 The group, which is 
entirely antarctic, contains the genera Eudyptes, Aptenodytes, 
Pygosceles, and Spheniscus. 

1 For discussion of the affinities of the Hesperornis see, besides MARSH and 
FUKBRINGER and GADOW in their large works. FURBRINGER, ' liber die syste- 
matische Stellung der Hesperornithidse,' Orn. d. deutscJi. Vcr. z. 
Sclmtz. Vogel, xv. 1890, p. 488 ; F. HELM, ' On the Affinities of Hexjiernniin,' 
Nature, xliii. 1891, p. 368 ; SHUFELDT, ' On the Affinities of Hesperornis,' ibid. 
p. 176 ; D'A. W. THOMPSON, ' On the Systematic Position of Hesperornis,' Html. 
Mus. Univ. Coll. Dundee, i. 1890. 

Since WATSON'S 'Report on the Penguins collected by the Challenger' 
the following papers have appeared : M. MENZBIER, ' Vergleichende Osteologie 
der Penguine,' &c., Bull. Soc. Imp. Nat. Mosc., 1887 ; H. SCHAUINSLAND, ' Zur 
EntwickelungdesPinguins,' Ycrh. Ges. DeutscJi. Natitrf. Leipzig, 1891, p. 135 ; 
STUDER in Zoology of S. M. S. ' Gazelle.' 


The penguins have a continuous feathering, the feathers 
acquiring upon the paddle-like wing a scaly aspect ; ' they 
have an aftershaft. The oil gland is tufted. 

In the myology of the fore limb the most remarkable 
fact is the entire absence of the biceps, a muscle which is 
wanting nowhere else among birds. There are also absent 
the expansor secundariorum, scapulo-humeralis anterior, and 
serratus metapatagialis. The tensor patagii loiigus is pre- 
sent, and its tendon is inserted on to the whole length of the 
bones of the arm as far as the extremity of the last phalanx. 
The latixxiinus dorsi is peculiar in that its two parts, ending 
in thin tendons, pass side by side through a pulley arising 
from the scapula. The pectoral-is major (of E. clirysocome) 
meets its fellow in the middle line over the carina sterni, as 
in tinamous, &c. In the hind limb the muscle formula is 
ABX+. The semimembranosiis is remarkable for the fact 
that it has, besides the usual head of origin, a second from 
the aponeurosis of the abdominal muscles. The accessory 
femorocaudal (in E. chrysocome) sends a muscular slip to the 
tendon of the femorocaudal. The tibialis anticus divides in 
Spheniscus mendiculus into two heads of insertion ; in other 
species its insertion, as is usual with birds, is single. There 
is but one peroneu-s present, the longus. The deep flexors 
are as in birds with but three well-developed toes, i.e. they 
blend. In S. dc-mersus there is a slip to the hallux ; in 
Pygosceles papua, as in Heliornis, the flexor hallucis splits 
into three tendons, one for each branch of flexor profundus, 
there being 110 slip to hallux. The glutaeus maximus is 
absent in Eudyptcs, limited in extent in the other genera. 
The ambiens grooves the patella. 

In the vascular system the most remarkable fact is the 
breaking up of the brachial artery into a rete in the arm. 
There are two carotids, which do not fuse. In the leg the 
sciatic artery is practically absent. 

1 It is stated that when moulting the scale-like feathers of the wing are 
detached in a continuous piece, as with reptiles (BAKTLETT, P Z. S. 1879, 
p. 6). 


The tongue and the roof of the mouth are covered with 


The provcntriculus has a patch of glands upon the right 
side, heart-shaped in Eudyptes chrysocome and Spheniscus 
demersus ; the gizzard is small and not gizzard-like. In 
Pygosceles the proventricular gland, however, is zonary, and 
the same state of affairs was found in one of four examples 
of Aptenodijtes. 

In Eudyptes chrysocome the gut is thrown into a vast 
number of primitive irregular folds. The duodenal loop is 
excessively complicated, more so than in Haliaetus, to which 
it bears some (probably a convergent) resemblance. 

The following are the intestinal measurements of a series 
of species : 


Small Int. Large Int. 

( 'WCM 

Fr, In. 



E/nJi/jitcs chrysocome 

11 8 (14 ft, 8 in., 


i* ! 

23 ft.) 

,, chrysolophus . 

21 3 (20 ft.) 

3i (3 in.) ? 1 

Sphenisctts magellanicus 

Aptenodijtes longirostr^ . 

30 6 (19 ft.) 
21 8 (17 ft. 10 in.) 

4 H 
41 ii 

*2 A 2 

From these data, which are extracted from WATSON'S 
memoir already referred to, it is clear that there is some 
individual variation in the length of the small intestine 
among the penguins, which is, indeed, greater than anything 
that has been recorded in any other group of birds. That 
the ca3ca are so small is a curious fact in the structure of a 
fish-eating bird. 

In the liver the right lobe is larger than the left in all 
species. The gall bladder is large and extends a long way 
down the abdominal cavity, as in the toucans. 

The syrinx is not especially divergent in structure. The 
trachea has a septum down the middle, as in certain 
petrels (q.v.) The intrinsic muscles are attached to the 
tracheal rings a considerable distance above the bifurcation 
of the tube, the distance varying from species to species. 
There is but little fusion between the last rings of the 
trachea. In Eudyptes chrysocome there is a thick fibrous 


pad at the commencement of the membrana tympaniformis. 
There is a tendency, as in the petrels, to the formation of ;i 
bronchial syrinx. This is especially well seen in Aptenodytes 
and Pygosceles. In the latter penguin the rings, three or 
four of them, after the bifurcation preserve the character of 
the tracheal rings, being deeper than those which follow, 
and being at the same time complete rings. 

In the penguin's skull the sutures are not so completely 
closed as in most birds ; in this they resemble the ratites 
among existing birds. 

In Eudijptes chrysocome, which may be taken as a type 
of the Sphenisci, the skull is schizognathous ; the maxillo- 
palatines are thin plates which are curved backwards, as in 
many Passeres and in some other schizognathous birds ; but, 
instead of being flatter upon the ventral and dorsal surfaces, 
they are compressed laterally. These bones gently nip the 
vomer, which is also composed of two flattened rami partially 
separated up to nearly the very anterior end. The palatines 
are large and flat with a very slightly developed internal 
lamina ; they nearly come into contact in the middle line. 
The pterygoids are unique among birds for their relatively 
immense size ; their shape is almost that of the human 
scapula, the wider region being at their junction with the 
palatines. Each pterygoid is perforated at the middle of 
the wider part. The nostrils are holorhinal, and the nasal 
bones at their posterior extremity are seen to overlap the 
frontals. There are strongly marked impressions for the 
supra-orbital glands. The lacrymals reach the jugals, where 
they expand into a flattened foot ; the descending limb of 
each lacrymal, which is flattened out in a plane parallel with 
the long axis of the skull, is there perforated by a very large 
foramen. It is extraordinary that in this bone, as in the 
maxilio-palatines, the plane of the bone is in a different 
direction from that of the same bones in other birds. There 
are no ossified ectethmoids. On either side of the foramen 
magnum are occipital fontanelles of small size. The inter- 
orbital septum is largely vacuolate posteriorly. In front of 
the occipital condyle, in the region occupied by the basi- 


occipital and the basitemporals, is a hollowed area which 
suggests the impress of the potter's thumb. The thumb of 
the writer exactly fills it. 

The symphysis of the lower jaws is of unusually limited 
extent. 1 

The vertebral column of Spheniscus Hiimboldti consists 
of twenty-one vertebrae in front of the sacrum, of which 
fifteen are cervical. The atlas is notched for the odontoid 
process. The catapophyses do not form a canal ; hypapo- 
physes commence on the eleventh cervical and continue to 
the last dorsal but one ; they are large on the twelfth and 
thirteenth ; on the fourteenth they have greatly diminished, 
but there are two large lateral catapophyses w T hich on C15 
form two large flanges (as in the diver) ; these arise from 
one base or Dl, and gradually diminish into a single median 
hypapophysis again. The first dorsal is the first vertebra 
to be opisthocoelous, but it is heterocoelous in front. The 
opisthocoelous characters of the dorsal vertebrae of the 
penguins are better marked than in other recent birds in 
which the same structure of vertebra occurs (cf. p. 111). 
In (Edicnemiis, for example, some of the dorsal vertebrae are 
opisthocoelous, but the convexity in front is by no means so 
clear as in the penguins. 

The scapula is remarkable for its great breadth posteriorly, 
which narrows towards the neck. In Pygoscdes it is wider 
actually as well as relatively than in Spheniscus, and has a 
truncated extremity ; the same is the case with Eudyptes 
and Aptenodijtes. 

The coracoid also shows some differences. In Sphcnix<--ux 
the procoracoid fuses below with the coracoid, leaving an 
oval foramen about half an inch long ; at the sternal end of 
the bone is a short upwardly directed snag of bone. In 
Pygosceles the procoracoid is not fused with the coracoid, 
but it is carried on by ligament to the snag at the base of 
the coracoid as probably also in Spheniscus. The acro- 

1 MENZBIEE describes the quadrate as single-headed. This is not accepted 



coracoid in both is very long. The fnrc.'iila is strong and 

The bones of the anterior extremity are extraordinarily 
flattened, in accordance with the paddle-like function of the 

The stcr mi in is roughly triangular, with a well-developed 
keel. There are tw r o lateral notches (one on each side), 
which are united distally by cartilage and membrane. There 
is a spina externa, but no spina interim. 

The pelvis, unlike what is found in the majority of birds, 
is remarkable for the fact that the pubes take a share in the 
formation of the acetabulum. The ilia are well separated 
from each other by the neural spines of the dorsal vertebrae. 
The pelvis is thus perfectly free from, not ankylosed to, the 
vertebral column. 

The point of chief interest in the hind limb of this group 
is the imperfect fusion between the short metatarsals, which 
closely resemble those of the dinosaur Ceratosaurus. The 
only recent bird which approaches the penguins in the 
shortness of these bones is Fregata. 

Fossil penguins are the genera Pculceeudyptes and Palce- 
xjilK'itixcus. The latter, from the Tertiariesof New Zealand, 
was originally described by HUXLEY,' and later by HECTOR. 2 
It was a large bird standing some five feet in height, the 
recent birds not being larger than three feet. The wings 
were proportionately longer than in recent birds, while the 
partly separate metatarsals were as is the case now ; hence 
this latter character is obviously an inherited one in the 
recent penguins. 

The affinities of the penguins are not clear. This is due 
to their antiquity, the existing characters of the group having 
been apparently acquired in the Tertiary period. The main 
facts of structure in which the penguins differ from other 
birds are 

(1) Continuous feathering (except Chauna). 

1 ' On a Fossil Bird . . . from New Zealand,' Q. J. (iwl. Sue. LSoK, p. 070 
- ' On the Remains of a Gigantic Penguin,' Trans. New Zealand fnxt. 1871, 
p. 341. ' Further Notice of Bones of a Fossil Penguin,' &c., ibid. Ls7'2, p. 438. 

D D 


(2) Scale-like feathering of the flattened wing. 

(3) Absence of biceps brachii muscle. 

(4) Presence of a vascular rete in the wing. 

(5) Freedom of cranial bones (not so marked in Eatitse). 

(7) Large and flattened scapula. 

(8) Short and imperfectly fused metatarsals. 

The opisthoco3lous character of the lumbar vertebrae is 
more pronounced than in other birds, but is, as has been 
already said, a character found in many groups. 

These features, some of them, have appeared so important 
to MENZBIER that he has divided birds into four great 
groups Saururas, Eatitse, Odontorma?, Carinatas, and, finally, 
the Eupodornithes or penguins. This, however, seems to be 
a too great separation from other birds. GADOW would 
place them nearest to the Tubinares and Steganopodes, the 
Colymbi being only a little further removed. 


Definition. All four toes webbed. 1 Oil gland tufted; aquincubital. 
Skull desmognathous, holorhinal, without basipterygoid pro- 
cesses. 2 Cseca present, but small. 

Though this group shows much divergency of structure, 
its naturalness can hardly be doubted. The number of 
rectrices varies. In Phalacrocora.v 6;v/.sv7/Vy/.s/.s and graculus 
there are twelve, so also in Frcgata aquila and Plot UK 
f lulling a? Phaeton, has twelve or sixteen. P. carbo has four- 
teen. Pelecanus has up to twenty and twenty-four. 

The aftershaft is minute but distinct in Frcgata, appa- 
rently absent in Plotus and other genera. 

The skin is only slightly pneumatic in Fregata, not so at 
all in Plotus. It is distinctly emphysematous in Phaeton.* 

1 This one feature is sufficient to define the group. 
'- Rudiments exist in Pelecanus ; cf. infra, p. 409. 

3 P. melanogaster appears to have only ten. 

4 Some few details of the structure of the soft parts of this tern-like 
steganopod are to be found in BRANDT, ' Monographia Phaethontum,' Mem. Ac 



The tufted oil gland has four orifices in Phalacrocorax 
brasiliensis and in Plotus melauog aster. Phaeton has six, 
the other genera apparently two, save Pelecanus, which has 
the unusually large number of twelve. 

The pterylosis, on the other hand, is very uniform, and 
the feathering is very close. The neck and head are closely 
feathered, and there is a very narrow apterion on the breast. 
The spinal tract has a very limited apterion between the 
shoulder blades. 

The tongue is small in this group, practically obsolete in 
Plotus (see fig. 5, p. 20). The proventricular glands are in two 
large squarish patches in Phalacrocorax (brasiliensis). The 
remarkable modifications of these organs in Plotus are 
described later (p. 414.) 

The gizzard is very small in Pelecanus and Sula, the 
proventriculus being enormous. The glands are zonary in 

The following are measurements of the alimentary tract 
in a series of Steganopodes. The most remarkable fact to be 
noted in the table is the great length of the large intestine 







Fregata aqn/ln . . . 36 



PJialacrocoi'u.r l>ru\ilii //*/> . 42'75 



carbo . . Ill 



Sula bassana .... 57 



Pelecanus onocrotalnx . . 93 



,, iiiitrcttns . . 90 2'5 


rufcxcens ... 65 1'5 


Plains anliiinii: ... 54 


(One cgecum 

only in some 


,, Lei'nilliuifi ... 24 3 


,, mel&nogaster ... 30 5 - 5 

2 * 

Phartoii x/> 42 -75 -2 * 

* These measurements are derived 

The varying proportions of the liver lobes l are given in 

St. Petcrfsbiinj, Ls40, p. 239 and pi. v., and in a paper by myself, P. Z. S 
1897, p. 288. 

1 G. ALIX, ' Sur 1'Anatcmie du Pelican,' Bull. Soc. ZooL Fr. ii. 1877, p. ^_ 

D D 2 


the table on p. 415. The gall bladder appears to be always 

The tensores patagii of Fregata 1 are somewhat compli- 
cated. The tensor muscle gives off two tendons, of which 
the anterior, the longus, is much the thickest. The latter 
is reinforced by an elastic slip from the deltoid ridge of the 
humerus, whereupon it again divides into two, a branch 
going to fuse with the brevis ; where this branch joins the 
brevis tendon that tendon gives off a wristward slip, and is 
itself continued over on to the ulnar side of the fore arm. 
There is a patagial fan and a bony nodule where it arises 
from the junction of the wristward slip of the brevis and the 
extensor metacarpi radialis. 

In Sula bassana the tendons of the brevis are two from 
the very first ; the anterior one corresponds to the wristward 
slip of Fregata, and from it springs the patagial fan. 

The other genera are not very different, save that in none 
is there an osseous nodule, and that all have a patagial fan. 

The pectoralis I. is double in Fregata, Plotus, Pelecanus, 
and Sula, variable in Phaeton, single in Plialacrocorax. 

The biceps slip has been occasionally overlooked. It is 
present in Plotus, Plialacrocorax,' 1 Phaeton, and Sula. 3 It is 
absent in Pelecanus, Fregata. 

Where present, however, it is slender, and is attached 
sometimes to the tensor longus, as ordinarily, and sometimes 
to the patagium itself, as in Colymbus, Podica, &c. The 
deltoid has the scapular slip in Plialacrocorax. 

The expansor secundariorum has been commonly said to 
be absent from the wing of the steganopods. This is not, 
however, at least according to FURBRINGER, an accurate 
statement of the case. In an embryo of Plialacrocorax 
carbo unmistakable traces of it were discovered, while in 

287 ; G. L. DUVEENOY, ' Sur la Poche Mandibulaire du Pelican,' Mem. I'lnst. iii. 
1835, p. 219. 

1 The anatomy of this genus has been described by BUKTOX in Linn. Trans. 

vol. xiii. p. 1. 

2 Not always. FORBES records its absence in P. br.asilicnsis. I did not find 
one in P. africanus. 

3 Not always ; it is absent in S.fnsca (fide GAEI;<>]>). 



Sulct and Pelecanus a slender tendon, running from the arm- 
pit and ending in un striated (Sula) or striated (Pelecanus} 
fibres for the movement of the secondary feathers, was dis- 

The biccys is two-headed in the Steganopodes ; but the 
arrangement differs from what is common among birds. 
Both heads, in fact, arise from the coracoid, but the outer 
one, which corresponds to the humeral head of other birds, is 
also attached to the humerus. The two muscular bellies are 
separate in Pelecanus and Fregata, and their tendons unite 




'.'('/-, corncoM ; C, coracoidal head of biceps ; A, attachment of humeral 
B, its prolongation to coracoid. 

to divide again directly. In Fregata, indeed, the division of 
the tendon of the coracoidal head takes place before the 
junction. In Phalacrocorax africanus I found the coracoid 
head alone, and it had but one insertion. In Phaeton and 
Phalacrocorax and Sula the tendons of origin of the two 
heads from the coracoid are continuous. In Phalacrocorax 
and Plotus, at any rate, the anconaus has a humeral head. 

The muscles of the leg vary greatly among the Stegano- 
podes. The following are the formulae for the different 

genera :- 

Phaeton AXY- 1 
Plotus AX + 

Pelecanus AX 

S/ila AX + 

PJialacrocorax \ 
Fregata A + . 

f ABX + - 
1 AX + 

1 Perhaps the ambiens varies. GAKROD, FORBES, and I did not find it. 
FCRBRINGER marks it as present, as does GADOW. 

- ABX P. carto? AX P. lur/ubris, P. brasilicnsis. 


As regards the other leg muscles Phaeton has quite excep- 
tionally no tendinous loop for the biceps to pass through. 1 

The plantar tendons blend in Fregata, &c., sending slips 
to all four toes ; they blend in Phaeton, but send no branch 
to hallux. In Plotus there is no blending, but a strong 
vincnlum, which is attached to flexor perforatus, just as it 
splits up into its three branches. This too is the case with 
Phalacrocorax, but the vinculum is not strong. 

In the pelicans (at any rate in P. rufescens and P. 
mitratus) there is a curious relationship between the femoro- 
caudal and the semitendinosus. The former receives a tendi- 
nous slip not far from its insertion, which runs up from the 
middle of the semitendinosus at right angles to its fibres. 
Whether this may be regarded as a rudiment of the accessory 
semitendinosus or not is uncertain. 

Another peculiarity shared with Biziura lobata, Colymbu*, 
and the extinct Hesperornis is the perforating of the patella 
by the tendon of the ambiens in Phalacrocorax. In Plotu* 
there is a groove upon the ossified patella, and, remarks 
Professor GAEEOD, ' some of the fibrous ligament overlapping 
this groove shows traces of ossification ; so that in aged birds 
this groove may be converted into a foramen.' 

Gluiceus I. in the Steganopodes is a small muscle, not 
extending, or hardly extending, over the biceps. Glutceus 
V. is large. In Fregata glutasus I. is absent. 

The form of the syrinx of the Steganopodes varies con- 
siderably, but in all it is trach Go-bronchial ; intrinsic muscles 
may be present or absent.-' 

Of Pelecanus I have examined six species, viz. P. niitratu*, 
P. onocrotalus, P. rufescens, P. fitscus, P. conspicillatus, and 
P. crispus. In none are there any intrinsic muscles, and the 
bronchidesmus appears to be complete. 

In P. conspicillatus the organ of voice is very simple. 
The rings are but little modified. There is a bony pessulus, 
which is attached behind to the last two tracheal rings, 

1 This peculiarity is, however, to be found in the swifts. 

2 The anatomy of Pelecanus rufescens has been described by OWEN, P. Z. S. 
1835, p. 9, by MARTIN, ibid. p. 16, and by ALIX, loc. cit. on p. 40:5, above. 


which are fused and ossified where they pass into it ; in 
front the last of these and the next following are ossified and 
fused in the middle line ; the middle one of the three is con- 
tinuous with the pessulus. 

P. m it rat ux chiefly differs in the larger number of rings with 
which the broad three-way piece comes into contact in front 
and in the greater length of the membrana tympaniformis. 

P. fuscus is remarkable for the degree in which the 
syrinx is flattened from before backwards. This is caused 
by the straightening of two of the bronchial semi-rings, 
which thus come, though hardly longer than the others, to 
project out considerably beyond them. At a first glance 
these two semi-rings appear to be the first two of the bron- 
chial series which are commonly among birds different from 
those which follow. There are, however, five pairs of bars 
in front of them, only separated in the middle line in front 
by a furrow ; on slitting up the windpipe it may be observed 
that there is really a septum' between them, and that they 
are bronchial. Owing to the peculiar form of the sixth and 
seventh semi-rings the membrana tympaniformis is exposed 
behind, but not in front. 

P. ouoc ratal /(* and P. crispns have syringes which are 
very much alike. When removed from the body, at any 
rate, the bronchi stand out at right angles to the trachea. 
This is due to the very large posterior end of the pessulus, 
which is broadened into a bar at right angles to and longer 
than the median portion of the pessulus. It is not ossified. 
There is a slight dilatation of each bronchus, which is carried 
to an excess in the next species. 

In P. rufescens the characters of the syrinx of the last 
species are carried to an extreme. The pessulus is the same 
in shape, but ossified. The bronchi are greatly swollen for 
the space of about an inch, there being occasionally some 
forking and anastomosing of the individual rings. The ends 
of the rings in the swollen region are nearly in contact, being- 
separated only by a narrow membranous interval. 

In Fregata aqaila the syrinx has a pair of intrinsic 
muscles and the bronchidesmus is incomplete. The syrinx 


is flattened from before backwards, and the first two bronchial 
semi-rings are very prominent. To the first of them, and 
apparently also to the ring in front, are attached the two 
muscles into which the intrinsic muscle divides. There is 
a membranous gap separating the last of the specialised 
bronchial semi-rings from the first of those which follow, 
whose border, moreover, is concave upwards. The last few 
tracheal rings are ossified and firmly fused. 

Phakicrocorax has a complete bronchidesmus and a single 
pair of intrinsic muscles. The first three bronchial semi- 
rings are very prominent and arched, and to the third of 
these the intrinsic muscles are attached. There is a mem- 
branous gap between the last of these and the first of the 
remaining series of bronchial semi-rings, which forms, at any 
rate in P. carbo, quite a pocket. The fourth bronchial semi- 
ring is curved in the same direction as that which precedes 
it ; both, in fact, are convex. The curvature, which is slightly 
more marked in P. brasiliensis than in either P. carbo or 
P. varius, suggests very much the syrinx of certain auks 
(cf. p. 363). In P. varius the intrinsic muscles are attached 
to the second bronchial semi-ring, as also in P. brasiUetislx. 

The syrinx of Plotus does not differ greatly, but it has an 
incomplete bronchidesmus. There are two bronchial semi- 
rings, which are specially increased in length and depth ; 
they are the second and third, and are relatively stouter than 
those of Phalacrocorax ; to the first of them the intrinsic 
muscles are attached. 

The syrinx of Sula is a good deal different from that of 
other steganopods. 

There is no ossification, except in the pessulus. A square 
projection is formed by a fusion between last tracheal rings ; 
this is continuous with the pessulus and is well shown in 
GAEHOD'S figure. 1 The bronchial semi-rings are at first 
feeble with wide interval. Between the third and fourth of 
them there is, covering the insertion of the intrinsic muscle, 
a protuberant pad of elastic tissue about the size and shape 
of a pea, 

1 P. Z. S. 1870, pi. xxxviii. fig. 4. 


In Phaeton flavirostris the syrinx is typically tracheo- 
bronchial and not flattened, as in Fregata. 

The skull of. the Steganopodes ! is desmognathous. It is 
most extremely so in Pelecanus, in correlation, perhaps, with 
the long broad beak. Pelecanus, in fact, may be described 
as doubly desmognathous, for the palatines are not merely 
united but ankylosed behind the posterior nares, which are 
of limited extent. They form but one bone with a deep 
ventral median crest, and on the opposite side an equally 
pronounced dorsal crest, occupying a space left by the here 
deficient interorbital septum. Fregata is nearly at the 
other extreme, for the maxillo-pala tines are largely free from 
each other in the middle line, and the palatines are only 
united for a short distance posteriorly. Phaeton ' 2 is most 
like Fregata, but here there is no fusion between the 
palatines. The strong inferior crest of Pelecanus is repre- 
sented by two feeble ridges of limited extent. Phalacrocorax 
is intermediate. The maxillo-palatines are completely united. 
The palatines are fused for the greater part of their length 
posteriorly ; they are, however, quite flat above and have below 
but a faint trace of the median crest. The interpalatine 
space anteriorly is much more capacious than in Pelecanus. 
Plotus agrees with Plialacrocorax. 

The Steganopodes are generally (HUXLEY, FURBKINGEE, 
GADOW) said to have no basipterygoid processes. In 
Pelecanus rufescens, however, I find a pair of thorn-like 
outgrowths in the right position (cf. Platalea, p. 439), which 
I take to be the rudiments of these structures. 

The bony nostrils are holorhinal, pervious only in Phaeton, 
m others much obliterated by bony growths, as in Herodiones ; 
in Plotus, indeed, reduced to the merest chinks. As in some 
Herodiones and Tubinares they are continued forward by a 
marked groove which runs to, or near to, the very end of 
the bill, absent only in Phaeton. In Fregata there is no 

1 All the types are described and figured by BRANDT, ' Zur Osteologie di-r 
Vogel,' Mem. Ak. St. Petersb. 1840 (6), Hi. p. 81. See also for osteology of 
Plotus and Phaeton MILNE-EDWARDS in Hist. Nat. Madagascar. 

- BEDDARD, ' Notes upon the Anatomy of Phaeton,' P. Z. S. 1897, p. 288. 


hinge line separating the 
cranium from the face. 
This is present in Pele- 
canus and Phaeton, while 
in Phalacrocorax and 
Plotus the existence of 
fibro-cartilage allows of a 
free motion. 

The lacrymal in Frr- 
gata and Pelecanus is large, 
with a large descending- 
process, which in both 
reaches the jugal bar ; in 
addition Fregata has an 
uncinate bone, which 
reaches the palatine. Of 
this there is apparently a 
rudiment in Phaeton. 

In Phalacrocorax the 
descending process of the 
lacrymal completely blends 
with the ectethmoid to 
form a ring of bone, as in 
the Limicolse. In Plotu* 
the same thing occurs, but 
the lacrymal is much 
smaller. The interorbital 
septum as an ossified 
structure is almost com- 
plete in Pelecanus and 
Fregata, largely deficient 
in Phaeton, almost absent 
in Phalacrocorax and 
J'lotitx. Iii Phalacrocorax 
c<irl)o there is a small 
bonelet resting upon the 
jugal bar in front of the 

FIG. 191.- SKULL OF Fregata. VENTRAL l acryma l ; i this is also 

Vo, voiner : M.I-/J, maxillo-palatines : A', upwardly ' Ossicilltliii Sllprajurjalc of BRANDT. 
directed part of maxillo-palatines. 



well developed in Plot its. There are traces of it in Frc</t. 
Another peculiarity shared by Plialacrocorax and Plotits is 
a style-like bone l attached to the occipital, to which the 
temporal muscles are partly attached. The bone varies in 
size, is not ankylosed to the skull, and is probably to be 
looked upon as an ossification in the septum between the 
two muscles. In these birds also 
the quadrate is peculiar in form in 
that the anterior process is short 
and slender and at right angles to 
the rest of the bone. 

Sula is nearest to Phalacroeo- 
rax. It has the same peculiar 
form of the quadrate bone, and 
the equivalent of the small bone 
seated upon the jugal bone is 
apparently there, though anky- 
losed. The nostrils too are re- 
duced to a mere pinhole, as in 
Plot it x. The palatines agree ab- 
solutely with those of Plialctcro- 
corax and Plot us, but the interor- 
bital septum is not so completely 
vacuolate. It rises up, mure- 
over, in front, as in Pelcconnx, 
and a faint crest from the pala- 
tines ascends into the vacuity. 
The lacrymal is, however, 
different ; the orbital part is 
-small, but the descending bar is large and joins the jugal; 
the ectethmoids appear to be deficient as bony structures. 

1 J. A. JEFFRIES, ' The Osteology of the Cormorant,' Science, ii. p. 739, iii. 
pp. 59, 274 ; GILL, ' Osteology of the Cormorant,' ibid. iii. p. 404 : SHUFELDT, 
' Remarks upon the Osteology of Ph. bicrisidtii*,' ib'uL ii. p. 640, and ' Osteology 
of the Cormorant,' ibid. iii. p. 143; DOLLO, Bull. Mus. Roy. BcUj. iii. 18S4, p. 
130. This bone (xiphoid, YAKEELL; nuchal, MARSH ; intranuchal, DOLLO) has 
been wrongly compared with the post-occipitals of dinosaurs. It is merely a 
sesamoid. LUCAS, 'Description of some Bones of Pallas's Cormorant (Ph. 
pers2>icillaiuis),' 1'. U. S. Nat. Mus. xii. 1889, p. 88. 

AS IN FIG. 191. 

OF Phaeton 


The lacrymal, as in Pelecanus and. Fregata, is deeply notched 

MIVAET, from his investigations into the axial skeleton 
of this group, 1 set aside Phaeton and Fregata by reason of 
their possessing twelve or thirteen cervical vertebrae and 
the doubly notched (on each side) sternum of Phaeton, 
besides a number of other points. 

There are seventeen cervical vertebra- in Pelecanus, 
eighteen in Sula, twenty in Phalacrocorax. 

The atlas vertebra is notched for the reception of the 
odontoid process in Pelecanus, perforated in Sula and Plotus ; 
in Phalacrocorax the conditions are intermediate, a perfora- 
tion being only just closed above. 

The Steganopodes have catapophyses upon some of the 
cervical vertebrae which enclose a canal ; in Pelecanus this 
is found on vertebra? 8-15. In Sula the canal may com- 
mence on the same vertebra, but more usually on the ninth, 
extending to the thirteenth. In Phalacrocorax there are no 
vertebrae with a complete haemal arch formed by union of 
the catapophyses. In Plot us the canal begins on the ninth 
vertebra and extends to the fourteenth. In Phaeton there is 
no canal. The dorsal vertebrae are opisthocoelous in Phala- 
crocorax and Plotus, not in Sula and Pelecanus. 

In all these genera the sternum is at most only slightly 
notched by one notch only on each side. In Pelecanus the 
clavicles are ankylosed to its keel. They reach it in Plotus, 
Phalacrocorax, and Sula, but are only firmly connected by 
ligaments, not ankylosed. In Phaeton the clavicles are 
attached to the keel behind the extremity. 

Both spina externa and interna are wanting in Sula and 
Pelecanus ; the spina externa is present, but small, in Phala- 
crocorax ; contrary to the statement of FUEBEINGER (in 
his tables) I find a rudimentary spina externa in Plot us 
anhiiiga. Fregata and Phaeton show their divergence from 
the normal steganopod type by a better developed spina 
externa, while the latter bird may possess a rudimentary spina 
interna. Four to six ribs reach the sternum. 

1 ' On the Axial (Skeleton of the Pelecanidse,' Tr. Z. S. x. p. 315. 


The relationship of the clavicles to the scapula and to the 
coracoid, upon which FURBRINGER has laid so much stress, 
serves to differentiate some of the Steganopodes. In Plot us 
the clavicle is connected by ligament with the scapula ; this 
connection is nearly effected in Plialacrocorax and Sida, but 
not in Peli'cdinif;. In Frei/iitti the dilated end of the clavicle 
is perforated in the middle ; it is, moreover, fused with the 

The genus Plotns (consisting of the four species P. aiihiitt/n, 
P. melanog aster, P. Novce Hollandice, and P. Levailltinti) has been 
investigated by BRANDT/ EYTON,' 2 D6NiTz, 3 GABBOD, 4 FORBES, 5 FUB- 
BBiNGEB, 5 and myself. Many of its characters have been described 
in the foregoing pages. I shall bere direct attention to certain 
peculiarities of Plotns which it does not share with the other 
Steganopodes, or which it possesses in a more marked degree than 
its nearest ally, Plialacrocorax. The darters feed in a peculiar 
manner ; they pursue fishes under water with a jerky action of the 
head and neck. This 'action, as Mr. FOBBES has suggested, may 
be compared to that of a man poising a spear before hurling it. 
' Arrived within striking distance,' continues Mr. FOBBES, ' the 
darter suddenly transfixes in fact, bayonets- the fish on the tip of 
its beak with marvellous dexterity, and then immediately comes 
to the surface, where the fish is shaken off the beak by jerking of 
the head and neck, thrown upwards, and swallowed, usually head 
first.' This mechanical action is associated with a mechanism in 
the neck. 

The first eight vertebrae form a continuous curve forwards, so 
marked that the head when outstretched is in the same straight 
line with the eighth vertebra. This latter vertebra is articulated 
at right angles with the foregoing, and almost at right angles with 
that which follows ; there is thus formed a conspicuous kink in 
the neck, which is never unbent. 

1 Loc. cit. (on p. 409.) - Osteologia Arium, p. 218. 

3 ' Ueber die Halswirbelsiiule cler Vogel,' &c., ArcJi. f. .Inn I. it. Plii/s. 1873, 
p. 357. 

1 ' Notes on the Anatomy of Plotits anJiinga,' P. Z. S. 1870, p. 335, and 
' Note on Points in the Anatomy of Levaillant's Darter (Plotus Levaillanti),' ibid. 
1878, p. 679. 

5 ' On some Points in the Anatomy of the Indian Darter,' &c., ibid. 1882, 
p. 208. 

6 ' Notes on the Anatomy and Osteology of the Indian Darter (Plotus 
melanogaster),' ibid. 1892, p. 291. 


On the ninth vertebra there is on the dorsal surface a fibrous 
loop (' Donitz's bridge'), which is fibrous only in P. anliinga. 
ossified in the three other species. Through this loop passes the 
tendon of the longus colli posterior muscle to be inserted on to 
vertebrae 2, 3, 4. The longus colli anterior is a very powerful 
muscle, which ends in a long tendon attached anteriorly to the 


haemapophysis of cervical vertebras 8-10. A separate portion of 
the muscle is similarly attached to the eleventh. 

The pulling back of the head preparatory to striking is effected 
by the longus colli posterior, while the bayonetting movement 
is produced by the longus colli anterior. 

The second characteristic feature in the organisation of Plotus 
concerns the stomach. Though originally described by MACGILLI- 



VRAY in AUDUBON'S ornithological miscellany, it has been more fully 
described and illustrated by GAREOD and FORBES. One great 
peculiarity is that in P. anliinga the proventricular glands, instead 
of forming a patch, or patches, upon the inner surface of the 
proventriculus, constitute a special csecal diverticulum of the 
stomach, which is completely lined by the glands in question. 
The pyloric portion of the stomach forms a well-marked compart- 
ment, quite distinct from the gizzard region ; the opening of the 
pylorus into the duodenum is protected by a dense mat of hair- 
like processes, each of which is about half an inch in length. 
Microscopically these structures ' are much more like true hairs ' 
than like any filiform papillae which might occur in such places. 
In P. Lcraillanti there is no special compartment for the pyloric 
proventricular glands. The hairs lining the pyloric chamber have 
a more complex arrangement than in P. anhinya. There is 
(fig. 193) a dense mass of them lining the distal end of the pouch, 
but there is also a singular conical process of the mucous mem- 
brane, covered with more hairs and serving to close the pylorus. 
/ nii'/iinogaster agrees with the last species, but the plug is less 
developed, being rather a well-defined ridge than a retractile plug. 
Mr. FORBES thinks that this hairy plug of the darters is an 
exaggeration of the nipple-like valve which is found to guard the 
pylorus in many birds. The two Old-World species thus come 
nearer together, as they do in the ossification of Donitz's bridge, 
than does either to the New- World P. anhinga. 

In order to facilitate a comparison of the several genera 
of Steganopodes among themselves, the annexed table, indi- 
cating the differences in some of the muscles, viscera, and 
bones, may be of use : 




| 1 





1 i 



_r i 





^ ~ 



'S '-5 












^ z. 



A + 








1 viei-nnus 







Pun -i 1 

I'l ton 

A x y - 











AX + 



R 2 > L 



n '. 



AX + 







A b>elll 

Sula . 

AX + 


+ oro 

l!or 1 





The Steganopodes, though allowed by all to be a natural 


group, show, as is indicated in the above table, a considerable 
amount of variation among themselves. And, what is even 
more striking, the genera vary more than genera as a rule do 
-for example, the stomach of the darters, the biceps and 
carotid of Sida, and the muscle formula of the leg of Phala- 
crocorax. The accompanying table gives some of the chief 
characters of the several genera. It is interesting to observe 
from this table that there is to some extent a linear 
series to be deduced from the facts of structure. If we 
commence with Phalacrocorax, we may associate with it 
certainly Plotus and also Sula. In these genera the peculiar 
flat and largely fused palatines are associated with the 
muscle formula AX + . Sula offers a slight step in the 
direction of Pelecanus. There is a trace of the dorsal ridge 
of the palatine, and of the anterior vacuity in the interorbital 
septum for the reception of this edge. In Pelecanus these 
characters are fully realised, and at the same time the 
ambieiis has vanished. Fregata, which stands apart in the 
structure of its skull, is also unique in the group by reason 
of its muscle formula, which is only A + . But this genus 
and Pelecanus have lost the biceps slip, present in those 
referred to before, but commencing to disappear in Sula. 

The question is, In what direction has the modification 
gone ? Are we to start with Phalacrocorax or with Pelecanus, 
or with some other genus '? 

The answer to this question naturally depends upon the 
relationship of the Stegunopodes to other groups. The group 
to which they are most nearly allied appears to me (as to 
FORBES and others) to be the Tubinares. In estimating this 
affinity FORBES did not, in my opinion, lay sufficient stress 
upon the ' os uncinatum.' It is true that as he states- 
this bone appears also to be present in Cluuiga and in the 
touracos, not to mention other birds ; but in Chunga it is a 
continuation of the descending process of the lacrymal, and 
articulates with the jugal bar, while in Corythaix it articu- 
lates both with ectethmoid and lacrymal, though it ends, as 
in the birds under consideration, on the palatine. In both 
Fregata and Diomedea this bone connects the lacrymal with 


the palatine, the lacrymal itself reaching (or nearly so) the 
jngal bar. The palate of Diomedca is remarkably like that of 
Fregata, which, unlike other Steganopodes (except Phaeton), 
is not far from being schizognathous, and represents, I am 
disposed to think, the nearest intermediate form. The 
grooves, starting from the nostrils and running towards the 
end of the beak, are also found in the Tubinares (and in the 
Herodiones, with which latter group the Steganopodes share 
the very much reduced, and yet holorhiiial, nostrils) . Another 
fact which is perhaps of importance is the much-reduced 
gizzard and the correspondingly enlarged proventriculus. 
Less important likenesses are the double pectoral, the short 
colic ca&ca, which are occasionally reduced to one in the 
Steganopodes ; these points ally the group to the Herodiones 
as well as to the Tubinares. The very names Fregata and 
Fregetta, Pelecanus and Pelecanoides are an expression of 
these views. 

FUKBBINGEB, however, and GADOW place Phaeton near- 
est to the base of the steganopod series, and there is much 
to be said for this way of looking at the group. There is no 
doubt that Phaeton is very different from the other genera 
of the group ; indeed, if it were not for Fregata it would be 
difficult to avoid removing it altogether. It is really not 
desmognathous (in the sense of HUXLEY) ; for the maxillo- 
palatines do not fuse ; in front of them there is a bony plat- 
form, forming the hard palate, but this is produced from 
anterior ingrowths of the maxillae, not homologous with the 
maxillo-palatines, which are present and unfused. The 
vomer, moreover, is well developed and bifid posteriorly, 
being exceedingly like the bifid vomer of such schizognathous 
birds as the grebe, ^chmophorus. In Fregata we have a 
further step. The vomer has coalesced into a single rod ; 
the palatines have united posteriorly ; the os uncinatum, 
rudimentary in Phaeton, has increased (?), and the grooves 
running from the nostrils to the end of the beak have put in 
an appearance ; furthermore the nostrils, pervious in Phae- 
ton, have acquired the steganopodous imperviousness. The 
maxillo-palatines, however, are not united ; but beneath them 

E E 


is a slanting wall of bone (cf. fig. 191 X), present in all the 
other Steganopodes, co-existing with and lying below the 
maxillo-palatines, and therefore not comparable to them. 
True maxillo-palatines, indeed, seem to be only present in 
Pelecanus of the remaining genera of the family, where they 
are small and coalesced. The other genera have only the up- 
wardly sloping bone of Fregata. The vomer too appears to 
be at the most very small in the higher steganopods, and 
I have not been able to find it. 

All these facts point to the basal position of Phaeton. 
It has, however, in the loss of the ambiens departed from 
the primitive condition. We can derive Fregata from 
Phaeton ; but in this case the mutual relations of the 
Steganopodes and the Tubinares, and perhaps the Colymbi, 
become somewhat obscured. 

As perhaps an appendix to the present group we may 
consider Odontopterijx toliaplciis of the London clay, known l 
by a portion of a skull. 

FUEBRINGEE discusses this bird very slightly under the 
Tubinares and Anseres ; both GADOW and LYDEKKER place it 
near the Steganopodes, with which determination I associate 
myself. The most marked peculiarity of this bird, which 
has given to it its name, is the serration into longer and 
shorter teeth of the upper and lower jaw. The two jaws are 
grooved, which seems to indicate that the beak was, as in 
the Steganopodes (and other birds for that matter), divided 
into several pieces. On the right-hand side of the skull is a 
small notch, which has been identified with the bony nostril 
of that side. It has, however, in the drawings an accidental 
look, and the fact of the possible obliteration of the nostrils 
must be weighed in discussing the steganopodous affinities 
of the Odontopterijx ; for in many of those birds, especially 
in Sula and Plotus, the nostrils have been practically oblite- 
rated. It does not seem to me that the depressed form of the 
skull, or, so far as we can judge them, the shape of the 
lacrymals, is strong evidence in favour of the steganopodous 

1 OWEN, ' Description of the Skull of a Dentigerous Bird,' Ac., Q. J. GcoL 
Soc. 1873, p. 511. 


relationships of this remarkable fossil ; but the facts are at 
any rate not opposed to that placing. 

Of birds more definitely referred to the Steganopodes a large 
number have been described from Tertiary strata. If MARSH'S 
(rnic/i/nrna, from the Cretaceous, really belongs here, the group is 
as old as any existing group, and older than most. A rijillornis 
(= Lithornis) of OWEN is among the most interesting, inasmuch 
as it is known from fuller remains than others. The skull and 
some long bones have been found in the London clay. It is 
referred to the neighbourhood of the present group by LYDEKKER, 
but by FURBRINGER to the Ichthyornithes. 

Actiornis, Pelargornis are placed here by LYDEKKER, and FUR- 
BRINGER would include Remiornis (considered struthious by 
GADOW) and Chenornis, referred by others to the Anseres. 


Definition. Oil gland fsathered. 1 Aftershaft pressnt. 2 Aquincubital. 
Skull desmognathous, holorhinal, -without basipterygoid pro- 
cesses. Catapophysial canal nearly always present. Two 
carotids. Caeca present, but nearly always rudimentary. Ex- 
pans Dr secundariorum present. 

This group of birds is an extensive one, with a con- 
siderable range of structural variation. That the flamingoes 
form a group apart can hardly be doubted, though it is not 
easy to differentiate them by any very important characters 
from other Herodiones. Less easy is it to distinguish the 
herons from the storks. The extreme types, e.g. Ciconia 
and Cancroma, can be readily distinguished by the muscle 
formula and by the characters of the syrinx, not to mention 
some other points of minor importance ; but between the 
extremes are forms like AMimia, Scopus, and Baltniiceps, 
which forbid so sharp a line of division. As to Phceni- 
copterus, WELDON was the first 3 to show in a convincing 
way its likenesses to the stork, its previous association with 
the duck tribe having been in large part due to the lamellated 
bill and the webbed feet. As to the latter character, no one 

1 Except in Cancroma. 2 ? except in Lcpt/i/iti/i/s ar/jala. 

3 'On some Points in the Anatomy of Phtmiaiptfrus,'' iV:c., P. Z. ,S'. 1883, 
p. 038. 

K K 2 



would nowadays associate the gulls with the ducks for a similar 
reason, though it was, of course, done by the earlier ornitho- 
logists. The duck-like bill of the flamingo is not so exclusively 
anatiform as might be thought ; for a very decided stork, 
Anastomus, has a bill which has very much the same 
structure as regards the lamellae (hence, indeed, its specific 
name lamelligerus). 

The typical storks may be distinguished from the typical 
herons by the following table : 



Syrinx . . Without intr. muscles. 

With intr. muscles. 



C(Pca .... 2, small 

1, small 

Powder-doicns . . 





Pectoralis I. 

In two layers 


Cucullaris dorsocnta- 



Pectoralis abd. . . 


Vinculum of deep flexor Strong 

Weak or absent 

tendons . 


+ or 


2, separate 2, fused 

The set of differences may be certainly regarded as of 
family value. But it must always be remembered that there 
are tendencies to the heron-like organisation among true 
storks ; while Scopus, and possibly Balceniceps, are distinctly 

Family Scopidae. There is one genus only, containing but 
a single species, Scopus umbretta African and from Mada- 
gascar. The anatomy of this stork-like heron has been prin- 
cipally investigated, as regards the ' soft parts,' by myself. 1 

It differs from the true herons by the absence of powder- 
down patches, in having ten primaries instead of eleven, and 
in possessing sixteen cervical vertebra. On the other hand 
it differs from the storks in having an ardeiform or at 
least a ' typical ' syrinx, and (from the Plataleida?) in the 

1 'A Contribution to the Anatomy of Scopus umbretta,'' P. Z. S. 1884, 
p. 543. 


absence of a biceps slip to the patagmm. Its leg muscle 
formula too is that of a heron. Besides these facts 




Ih ^ ' /'/.'/, tlrxor hallucis : Fl.p, flexor proi'uinliis. 

which incline to both the stork and the heron side, and 
which, perhaps, are the most important of such facts it 
may be mentioned that Scopus shows a stork-like character 
in the partial division of the pectoralis primus, 
a fact which was first pointed oiit by FORBES 
incidentally, in his report upon the petrels 
collected during the voyage of the ' Chal- 
lenger ' The patag/l muscles and tendons 
are not specially distinctive of the affinities 
of the bird. In my dissection of Scopus I 
could not find the expansor sec mid a riorum ; 
but as this muscle is found in most of the 
Herodiones its absence is not in any way sig- 1/I( , 

llificailt, though it should doubtless be Pi. ANTAK. TENDONS 

verified. The CUICOHCCKS has a tendinous slip BEDDED)*. 
to the humerus. The deep plantar tendonx 
appear to vary somewhat ; in two specimens dissected by 
myself I met with the two conditions illustrated in the two 


figures annexed. In the one the flexor hallucis was united 
to the flexor communis by two distinct vincula, one before 
the tri furcation of the latter tendon, the other attached to 
the tendon supplying digit II. In a specimen of Ciconia 
nigra dissected by FORBES there was an identical arrange- 
ment. In the other Scopus (see fig. 195) only the last of 
the two vincula was present, i.e. that passing to tendon of 
digit II. 

For the skeleton of Scopus see MILNE-EDWARDS'S account. 1 

The skull is on the whole more 
stork-like than heron-like, but it does 
not show any of the extreme modifi- 
cations of the stork type. The bony 
interorbital septum, as in the storks, is 
not largely fenestrate. The inner 
lamina of the palatines does not reach 
the posterior boundary of those bones. 
In front, at about the middle of the 
interpalatine vacuity, the palatines are 
produced into a short lateral process ; 
this is well marked in many storks, 
but also in Can.cro/iia, to the skull of 
which heron that of Scopus shows 
OF another point of likeness ; in both 
these birds a deepish groove runs from 
the end of the nostril to the end of the 
bill. This groove is also found, though it is not so con- 
spicuous, in Ardea and Butorides, and (among storks only) 
in Platalea ; it is suggestive of a recently closed, more 
elongated nostril, like that of the cranes. The proeoraco'xl 
is more rudimentary than in storks, but the coracoids overlap 
at insertion. 

Family Ciconiidae. I include in this family not only the 
true storks but also the wood ibises (Tatitalttx). GARROP- 

' Histoirc Na/iircllc, itV., <lc Madagascar, ' Oiseanx,' p. 514. 
- V. E. BKW>AHI>, ' Notes on the Convoluted Trachea of a Curassow (A T o- 
itrnmntum), ancTon the Syrinx in certain Storks,' P. Z. S. 1886, p. 321. 

Leptoptilus (AFTEI; 



contrasted Ibis and Platalea on the one hand with Ciconia 
and Tantalus 011 the other, on account of the following 
differences : 

Ibis and Platalea 

Skull schizorhinal, angle of man- 
dible produced. 

Pectoralis major single. 

Accessory femorocaudal present. 

Semitendinosus muscular through- 

Biceps slip present. 

Ciconia am! Tantnliin 

Skull holorhinal, angle of mandible 

Vectoralis double. 

Accessory femorocaudal absent. 

Semitendinosus tendinous for 
distal half. 

Biceps slip absent. 

To which I add the form of the syrinx. These collec- 
tively appear to me to justify this separation. The true 



FIG. 199. SYRINX or Abdimiti x 
rJiyncha (AFTER. BEDDAED). 

storks, including Tantalus, are well characterised by the 
peculiar structure of the trachea and syrinx, there being, as 
already mentioned, an approach in these birds to the purely 
tracheal syrinx of the tracheophone Passeres. In the 
common black stork, C. iti(/r,t\\e syrinx has the form illus- 
trated in the figure (fig. 197). Its principal features are the 
absence of intrinsic muscles, the modification of the last 
tracheal rings, the existence of a rudimentary vocal process 
(see p. 69), and the closed character of the bronchial rings, 
which are rings, not semi-rings, the membrana tympani- 



formis being absent. In Xenorhynchus senegalensis, 1 how- 
ever (fig. '200), there is some approach to the more typical 
tracheo-bronchial syrinx, which is further developed in 
Abdimia sphenorhyncha and Dixsii-ra episcopus ; l the 
syringes of these two storks are illustrated herewith (figs. 199, 

FIG. 200. SYRINX OF Xenorhynchus scncyalensis (AFTER BEDDAED). 

198). In the former as well as in the latter the intrinsic 
muscles are still absent, and there is a considerable modifi- 
cation of the last tracheal rings ; but in both there is a 
partial deficiency of cartilage where the membrana tympani- 
formis is developed in other birds. In Xenorhynchus this 

1 F. E. BEDDARP, ' A Note upon Dissura episcopus,'' &c., P. Z. S. 1896, p. 231. 


is but slight ; in Abdimia and Dissnra episcopus it is well 
shown. But the latter bird has the complete bronchidesmus 
which marks the ciconiine as contrasted with the ardeine 
syrinx (cf. figs. 194 and 198). In Myctcria americana the 
bronchial rings are complete, but thinner internally, which 
is a hint of the otherwise absent membrana tympaniformis. 
The genus Tantalus is unique among the Ciconiidse in 
having a convoluted trachea. This, however, is now known 
to occur only in the male of T. ibis ; in both sexes of I\ 
loculator the trachea is uncoil voluted. 1 In the former bird 
the tube makes several intrathoracic loops, as shown in the 
figure (fig. '201). The syrinx is essentially stork-like. 

In most storks the muscle formula of the leg is AXY + ; ~ 
the only exceptions to this yet known are Xenorhynchus 
senegalensis, Dissura episcopus, &nd'Abdi>nia sphenorhyncha, 
where the ambiens is absent, and Leptoptilux crio/in/i- 
fenis and argala, where there is no femorocaudal. The 
tendency to an ardeine structure in the syrinx of these 
birds has already been remarked upon, and may possibly be 
correlated with the absence of the ambiens. Storks have 
no biceps slip, but a typical expansor secundariorum. The 
humeral head of the anconaeus is generally present. The 
patagial tendons are usually of a somewhat complicated 

The tensor patuc/ii brecis is constituted upon a similar 
plan in all storks, though there are naturally some little 
differences in detail. 

In Lcptoptilus, according to WELDON, 3 there is but one 
tendon which, widening out just before its insertion on to 
the fore arm, gives off a recurrent slip to the tendon of the 
long us. 

In Ciconia niyra, according to FURBRINGER, the tendons 

1 ' On the Trachea of Tttiitalii*: Ac., I'. Z. S. 1878, p. Ccjr, ; On the Form 
of the Trachea in certain Species of Storks and Spoonbills,' ibid. 1S7">, p. 297. 

2 A. H. GARROD, 'Note on an Anatomical Peculiarity in certain Storks,' 
ibid. 1877, p. 711. In a specimen of Xenorhynchus australis a few fibres corre- 
sponding to the accessory femorocaudal were found. 

3 Loc. cit. (on p. 419). 


FKI. '201. CONVOLUTED WixnrirK OF Taittdl/ix ibis (AFTER GARROD). 

r. L'oiacX'iil : /, furcula ; .</, sternum ; r.b, l.b, bronchi. 


are a little more complicated. The tendon of the brevis is 
obscurely divided into two from nearly its commencement ; 
the more anterior of these again divides into two, one of which 
runs forward to be inserted on to the fore arm separately 
from the hinder part, which remains continuous with the rest 
of the tendon ; there is a recurrent slip to longus. 

In Abdiuiia the tendons are much the same ; in both 
these genera the propatagialis pectoralis of FURBRINGER is 

In Tantalus leucoceplntlus the broad fascia-like tendon of 
the brevis gives off a wristward slip, from the junction of 
which with tendon of extensor inatacarpi radialis a patagial 
fan arises. 

The pterylosis of the Ciconiid* has been studied by 
NITZSCH. The neck is continuously feathered down to 
about the middle, where the spinal and ventral tracts 
respectively become divided into two. The two spinal 
tracts are narrow but strongly feathered, and cease abruptly 
at about the end of the scapula ; after a short space they 
recommence as a bifid but feebly feathered tract, the limbs 
of which unite a little way in front of the oil gland. The 
ventral tracts are broad upon the pectoral region but narrow 
towards the vent. 

In Pseudotantahis leucoccplialus NITZSCH could not find 
the aftershaft, but nevertheless one appears to exist. 

In Mycteria and Leptoptilus argala the dorsal tract has 
posteriorly no spinal space. 

In L. argala NITZSCH states the aftershaft to be absent. 

The oil gland has two apertures on each side in Cicunni 
alba, five in C. nigra ; L. argala. has no less than six. 
Anastomiis coroinnndeUciLS has three ; Tantalum /t'ltcocejjJtal/is 
has the same number. 

The deep flexor tendons of the Ciconiid* are constructed 
on the plan of type I. In Tauta/it* leucocephalus there is a 
slight variation ; a small vinculum runs to flexor communis 
before the latter divides into three, and then a broader viiicu- 
lum, chiefly going to tendon of digit II., but also slightly 
to III. and IV., binds together the two tendons. In C- 



there is the ordinary vinculum and a special slip to 
digit II. 

The lungs in the Ciconiidae are at least often distinguish- 
able from those of the Ardeidae by the great deficiency of the 
muscles arising from the ribs and attached to the pulmonary 
aponeurosis. In Cancroma there were four pairs of such 
muscles, arising from a corresponding number of ribs, in 
Nycticorax five ; but in Ciconia alia I only found one 
pair inserted on to the aponeurosis in front of the septum 
bounding the anterior intermediate air sac anteriorly. Pro- 
fessor WELDON found no such muscles in a considerable 
number of storks. 

The tongue is always small. The proventriculus is 
zonary. The liver is nearly equilobed, and there is always 
a gall bladder. The intestinal measurements of a number of 
species are as follows : 

S. I. L. I. 


Ft. In. 

Inches Inches 

Ciconia nicjra .... 4-'J 



alba . . . . I 3 (5 



may/in ri . . . 7 - <> 


boi/ciana ... 57 


/>i**/ini cpiscopun ... 33 



Abdimia sphenorhynchus . . 2 (5 



Mycteria americana . . . (14 



Xenorhynchus austral in . . -~> 



,, senegalensis . 5*5 5 


Lcptoptilns crumenifems . . 8-5 


,, arcjala . . . 6'5 


Tiintalus ibis .... 5 

,, loctilcitor . . . G'G 2 


The number of cervical vertebra? is seventeen (Lcpto- 
ptilus, Tanliilux) or eighteen (Xenorhynchus) ; the hypapo- 
physes are feeble ; there is a ventral canal formed by 
union of catapophyses of C7-C11 (Xenorhynchus), C7-C12 
(Tantalus}. Four ribs reach the sternum in Xenorhynchus, 
five in the others. The sternum has one pair of notches, and 
the spina externa is absent or small. The procoracoid is of 
fair size, but does not reach the clavicle ; the coracoids meet 

I1EROD10NES 429 

at their sternal insertion in Leptoptilits. 1 The hypoclei- 
diuni articulates with carina sterni. The skull is desmogna- 
thous, holorhinal, and without basipterygoid processes ; there 
are in Tantnhi^ rudiments of these processes in the shape of 
a minute spine. The holorhinal character of the nostrils is 
largely marked by ossifications of the alinasals ; the nostrils 
are thus much reduced in size, in a fashion suggestive of the 
Steganopodes and possibly significant, but there is no bony 
septum between them. This distinguishes the storks from 
the herons, as does also the form of the palatines. These 
bones are in the first place not cut off at right angles behind, 
as in the herons, while the internal lamina only bounds the 
interparietal space, and is at most (Xenorhynchus) carried 
back to the end of the bones as a slight median keel. This 
is absent in other storks. In Xenorhynchus and Lcptoptihi* 
the palatines again approach each other, and are only sepa- 
rated by the vomers just behind the maxillo-palatines. Oppo- 
site to this point each palatine is produced into a strong 
outwardly directed snag, large in Xenorhynchus, hardly indi- 
cated in other storks. The interorbital septum is entire. 
The large lacrymal is perforated or deeply notched for duct 
of gland. 

Family Ardeidse. The herons contrast with the storks in 
(1) the tracheo-bronchial syrinx always furnished with a 
pair of intrinsic muscles, (2) non-division of pectoralis pri/-ux 
into two layers, (3) invariable absence of ambiens, (4) pre- 
sence of powder-down patches, (5) absence or weakness of 
vinculum, ((>) presence of a single cascum only. 

On the other hand the two families agree in (1) absence 
of biceps slip to patagium, (2) presence of expansor secun- 
dariorum, in addition, of course, to the points enumerated in 
the definition of the group. 

In Cancroma the oil gland is nude. The herons have 
four or six. powder-down tracts. Six are found in Cancronni, 
Bntorides atricapillus, Anlcu cocoi, and other species; 

1 Not ill Xenorliiincltits and Dissura ; there is a trace of an overlap in 
Abdimia, Tantalum, and Platalca. 


four in A rdetta and Botau nt*. The number of rectrices differs 
much ; there are only eight in Ardetta exilis and A. involucris, 
ten in Botaiirus (not invariably), twelve in Ardea cocoi and 
( 'tuicroma cochlearia. 

The ptenjlosis differs from that of the storks by the 
narrowness of the tracts. The spinal space begins with the 
commencement of the neck, and only terminates a little way 
in front of the tufted oil gland. The anterior part of the 
spinal tract is not always more strongly feathered than the 
posterior part, and there is (according to the figures of 
NITZSCH) hardly a break between them. The ventral tracts 
also are separate early on the neck. In Cancroma they divide 
on the breast into a broader, stronger outer tract, which ceases 
just below the metapatagium, and a narrow inner tract. The 
anterior ventral powder-down patches constantly interrupt 
the continuity of the anterior and posterior sections of the 
ventral tracts. 




Ardea Goliath . 



,, sumatrana ... 57 


pur pur ca 




,, cocoi 




cgrctta . 




,, aarzetta . 



Butorides atricapilius . . 25 


ci/cnniriis ... 30 



Anlt'llii invnlucris . . . 22'5 


Nycticorax caledonicus . 40 



,, violaceus . *>0 


Tinrinoina brasiliensf 
Cancroma cochlearia 33 



Botaurns stellar is . . 51 

4 -15 

A curious absence of any apparent relationship between 
the relative length of the sections of the alimentary canal, 
and of the alimentary canal as a whole, and of the food, is 
shown by the above table, drawn up by Mr. GAEBOD, 
which I reproduce here from his notes. 

They seem, however, to ally the last four genera, which 


have a long large intestine. As a rule the left lobe of the 
liver is the smaller, while in the storks the lobes are equal. 
A gall bladder is always present except in Botaum* fitt'Hd 
The caecum is single, and, as will be seen from above 
measurements, rudimentary. 

The carotids are, as a rule, two, and separate ; but in 
Botaurus they fuse, and in Ardetta mvolucris the right only 
is present. 

Of the leg muscles the ambiens is always absent, and the 
formula is typically AXY . The feinuroeaudal is never 
strong, and is particularly slender in Ardea Goliath. 1 In .-1. 
Hum-atruua, A. ludoviciana, Nijcticorax Gardeni, Caucroina 
cochlearia, and Tigrisouia braxiliense it is totally absent. 

The deep plantar tendon* are characteristic ; there is 
almost always a very slender vinculum between the two, 
which is totally wanting in Bottii/rus stellaris, Ardetta 
involncris, and .4. ex His. 

The tensores patagii are stork-like. The tendon of the 
brevis bifurcates, and from the point where the anterior 
limb is inserted on to tendon of extensor of fore arm a 
recurrent slip is given off to lonyus. This arrangement 
holds good for Ardea pitrpurea, A. Goliath, Cancrunia 
cochlea ria, and Nijcticorax grisem; but in Cancroma the 
recurrent slip is sometimes absent. The pectoralis abdoini- 
nalis is present, and thus serves to differentiate them from 
the storks, ibises, and spoonbills, in which the muscle is 

The skull of the Ardeida? has been chiefly studied by 
SHUFELDT. 2 In the more normal forms (e.g. Ardea cinerea, 
Butorides cyanurux} the skull is holorhinal, the holorhiny 
not being obscured as it is often among the storks by 
the irregular ossification of alinasals. The vomer is well 
developed, much compressed laterally, and largely double. 
The maxillo-palatines are spongy bones, largely free 
from each other posteriorly. The palatines (see fig. '20'2) 

1 Sometimes absent in this species. 

' Osteologic-til Studies of the Subfamily Ardcimr,' ,J< urn. Cmnji. Mc<L and 
Stir,/. 1S89. 



are very straight bones, usually cut very squarely behind, 
but notched postero-laterally in Ardea bubulcus, A. mhtnta, 
A. coniata, with well-marked internal laminae, which extend 
quite to the posterior end of the bone. The interorbital 
septum is largely fenestrate. The bony ectethmoids are but 

little developed ; Cancroma is in 
several ways rather anomalous. It 
is heron-like in the fenestrated in- 
terorbital septum, and in the fact 
that the internal lamina of each 
palatine is continuous to the posterior 
end of that bone. The nostrils are 
continued forward by a deeper groove 
than that which is formed in the more 
normal herons a point of likeness 
this to Scopus and Balceniceps (qq.v.) 
The very broad palatines join the 
vomer again in front of their posterior 
junction with each other, thus divid- 
ing the interpalatine vacuity into two 
areas. It is thus to a certain degree 
' doubly desmognathous,' and is so 
far like Xenorhynchus (see p. 429). 
There is a well-marked lateral process 
of the palatines, as in Scopus and 

As in Scopus the procoracoid 
is very small and the coracoids overlap 
each other at their insertion. Like 
the storks, and unlike Scopus, the 
hypocleidium articulates with the end 
of the carina stermV The hypocleidium, moreover, projects 
backwards between the two clavicles as a narrowish piece. 

The hflemapophyses of the cervical and dorsal vertebra 
are small, those of the latter being sometimes quite absent. 
There is a catapophysial canal (in Cancroma as well as 
.Irdca), formed in the two types mentioned by cervical 
vertebrae 7-12. 

FK;. '202. VENTRAL SUR- 
ciii/'n'd (AFTER HUXLEY). 

J't, pterygoids ; /'/, palatines ; )', 
vomer : Jf.rj>, maxillo-palatines. 



Family Balsenicepidse. The great ' whale head ' of Africa, 
Balceniceps rex, requires further study before its exact 
position can be determined. It is admittedly a member of 
the present group, though its original describer, GOULD, 
regarded it as a pelican. 1 We know the skeleton through the 
labours of PABKEE, 2 while our at present scanty know- 

FIG. 203. SYRINX OF BaLcniccps. 

/>, free margin of bronchidesruus. 


ledge of the soft parts is due to myself. 3 Its powder-down 
patches were discovered by BABTLETT/ As in the herons 
also the right lobe of the liver is the largest, and the caecum 
is single. The syrinx (see figs. 203-205) is ardeine inform, but 
lacks the intrinsic muscles. These, however, are not en- 

1 Not a serious mistake in view of the admitted relationships between the 
Steganopodes and Herodiones. 

- ' On the Osteology of Balceniceps rex,'' Trans. Zool. Soc. iv. p. 209 
(abstr. in P. Z. S. 1860, p. 324). 

3 ' On certain Points in the Visceral Anatomy of Balceniceps bearing upon 
its Affinities,' P. Z. S. 1888, p. 284. 

4 ' On the Affinities of Balceniceps ,' ibid. 1861, p. 131. See also BEIN- 
HAEDT, ' On the Affinities of Balceniceps,'' ibid. 1860, p. 377, and GIEBEL, ' Ueber 
Balceniceps rex,' Zeitschr.f. d. yes. Nat. Ixi. 1873, p. 3-30. 

F F 


tirely absent ; their former presence is indicated by a narrow 
ligament on each side (fig-. 42, p. 62), which occupies the place 
that a muscle should, and is attached precisely where the 
intrinsic muscles are attached in other Ardeidae. The meni- 
brana tympaniformis is well formed, the bronchidesmus is 
incomplete, while the general form of the organ is purely 
tracheo-bronchial and thoroughly ardeine. This will be 
apparent from the annexed woodcuts. 

' The nearest relations of Balaniceps,' said PARKER, ' are 

the South American boatbill (Can- 
croma cochlea ria) and the little 
South African umbre (Scopus um- 
hri'tta)." The interorbital septum is 
stork-like in its completeness. The 
lacrymal, as in Sco'pus alone among 
Herodiones, reaches as far down as 
the quadrate jugal bar, but it is fused 
anteriorly with the walls of the 

skull. The nostrils are continued 
forward by a groove precisely like 
that of Scopus and Cancroma. In 
FIG. 205. SYRINX OF Bate- the palatine bones the fusion of the 


PESSULUS AND MEMBRANA internal laminae to form a median 

TYMPANIFORMIS (AFTER BED- keel I 3elimd tne interparietal space 

is precisely like Scopus ; so, too, is 

the lateral angle of these bones (see p. 422) . There is a firm 
synostosis between the furcula and the carina sterni. 

Cervical vertebrae 7-13 have, as in most other Herodiones 
(excluding, however, the supposed ally of Balceniceps, Scopus}, 
a ventral catapophysial canal. 

The family Plataleidae includes not only the spoonbills 
but the ibises. The name Hemiglottides was applied by 
NITZSCH to the group ' on account of the surprising small- 
ness of their tongues.' 

The pterylosis is exactly as in the storks. 1 The rectrices 

1 According to NITZSCH It appeared to me (in Platalea rosea) to be more 
like that of Tantalus loculator, in that the hinder part of the spinal tract was 
not bifid, but continuously though sparsely feathered. 



are twelve. The oil gland of Platnlcn. It'itcoroiHu lias three 
distinct orifices on each half, that of Ibis only one. The 
long downwardly bent bill of the ibises distinguishes them 
from the storks and suggests Niniii'iiiiis. NITZSCH, indeed, 
regarded the birds as intermediate between the two groups 
represented by these types. 

The tensores patagii have always a biceps slip running to 
the tendon of the longus, and there is a patagial fan. 

In Ibis (fthiopica the tendon of the tensor brevis is 
simple and rather diffuse. In Eudocimus ruber and G. mela- 
nopis the tendon gives off a distinct wristward slip, while the 
patagial fan is formed of two rather separate strands, with 
the posterior of which, rather high up, the wristward slip 
fuses in Geronticus melanopis. 

In Platalea the muscle and its tendons are much the 
same, but the brevis is very broad and fascia-like. 

The muscle formula of the leg is complete (i.e. ABXY+ ) 
in all Plataleidse. 

The plantar tendons are connected by a vinculum which 
in Eudocimus ruber extends on to the special slip to digit II. 
By the division of this vinculum may have arisen the two 
vincula of Ciconia nigra (see above, p. 427). 

The liver is equilobed (I. a'tliiopica}, or the right is a 
little larger (E. ruber). A gall bladder is present. The 
following are intestinal measurements :- 

s. i. L. I. 


Eudocimus ruber 

31 1-6 


Nipponia Teiii/iiini-l.'/i 



Plegadis falcindliis - 


J~> ' 

Ibis (Etliiopica . 



., strict ipcnnis 



Platalea Icucorodia . 





The intestinal convolutions of Platalea leucorodia are 
shown in fig. 207. The greater part of the gut has preserved 

1 HUNTER, in Essays and Observations (eel. Owen, London), isill, writes of 
this species : ' The cseca are about four inches long and very small, attached to 
the ileum their whole length.' 

I K 2 


the primitive arrangement. The duodenal loop is curved, 
and in other storks (s.s.) this (cf. fig. 208) is converted into a 
spiral. There is a tendency, in fact, among the Cicomidae 

FIG. 206. WINDPIPE OF Platalea ajaja (AFTER GAREOD). 

11, trachea ; l>, bifurcation of bronchi iu front of sternum ; r.b, l.b, bronchi ; d, oesophagus 

e, cervical muscles. 


to the formation of these spirals, which are also found in the 

The windpipe in the ibises is simple, not convoluted ; 

FIG. 207. INTESTINES OF Platalea Icncoroclia (AFTER MITCHELL). 
x, short-circuiting vessel divided. 

but in Platalea leucorodia it is convoluted. 1 In this bird 
the windpipe runs in a straight or slightly sinuous course 


x, as in fig. 207. 

into the thorax ; it then bends upon itself, and after an inch 
and a half or so reverts to its original direction, and divides 
within the thorax into the two bronchi. 

' I have a windpipe of P. leucorodia which is not convoluted, but which 
bifurcates within the thorax. 


In Phitaha ajaja (see fig. 206) the trachea is peculiar 
in that its bifurcation takes place about the middle of the 
neck, some distance at any rate before the entry of the 
bronchi into the thoracic cavity. 

In a female which I measured the bifurcation was three 
inches in front of sternum, and the distance from the bifurca- 
tion to the end of the larynx was six and a quarter inches. 
HUDSON has stated l that there are two species of spoonbill 
in the Argentine : one has a windpipe like that figured above, 
and may be considered to be the true ajaja, while the other 
has a windpipe which is not modified in any way. The 
characters of the latter apart from the question of the 
windpipe have been considered by some to be those of 
immaturity ; but this is denied by HUDSON. 

I have in my possession seven windpipes of P. ajaja, of 
which two are males and the rest females. In the two males 
the lengths of the trachea are practically identical. In the 
females the position of the syrinx varies considerably, being 
more than an inch lower in some than in others ; this may 
possibly account for the differences observed b}^ Mr. 

The extrinsic muscles of this species arise from inner side 
of coracoid, not far from the middle line of sternum. The 
intrinsic muscles of the syrinx are present ; they cease, how- 
ever, some little way in front of the syrinx. These muscles 
are also inconspicuous, and form a very thin though rather 
wide band. The syrinx is not especially stork-like, except in 
the fact that the bronchi are for the greater part of their 
extent tubes with complete rings. The first dozen or so, 
however, are incomplete, there being thus a membrana 

There is a cartilaginous three-way piece, and generally 
there is but little ossification in the windpipe. 

The windpipe of P. alba is different from both the species 
already described. 

It is not convoluted, but the bifurcation does not take 
place outside of the thoracic cavity, as in the last species. It 
1 SCLATER and HUDSON, Argentine Ornithology, vol. ii. p. 115. 


is also stork-like in the fact that there is no membraiia tym- 
paniformis ; the bronchial rings are complete rings from the 
very first. 

The main peculiarity of the windpipe, however, lies in 
the fact that from a point about three inches from the larynx 
it appears to bifurcate and to consist of two closely applied 
tubes. By cutting windows it was ascertained that this 
appearance corresponded to the reality that the windpipe 
did consist of two tubes. This arrangement, which seems to 
characterise the present species only, is, of course, suggestive 
of the median tracheal septum of the penguins, &C. 1 

Of the ibises I have examined the windpipe of Eudocimus 
ritber, E. alba, and of Ibis cethiopica. There are differences 
between the species similar to those which exist between the 
species of Platalea. In. Ibis a'thtopica there is no membrana 
tympanif ormis, the bronchial rings being complete. In the two 
other species this structure is present, and moreover the last 
few tracheal rings are defective in the middle line posteriorly. 

The extrinsic pair of muscles are attached in E. ruber to 
the inner surface of sternum not far from the middle line. 
The intrinsic muscles stop short some way in front of the 

In Platalea ajaja cervical vertebrae 7-10 have a cata- 
pophysial canal. Some of the dorsal vertebra are fused. 

The number of cervical vertebra is seventeen or eighteen 
(Platalea ajaja}. The sternum is two-notched (Platalea} 
and has a small spina externa, but no spina interna. 

The skull has, as in charadriiform birds,' 2 a pair of 
occipital vacuities. 

It is schizorhinal and has rudimentary basipterygoid 
processes, which in Platalea ajaja have the form of small 
sharp thorns. Supra-orbital impressions are feebly developed. 
The lacrymals with the ectethmoids very nearly form a com- 
plete ring of bone, the aspect of this part of the skull being 

1 GADOW, in NEWTON'S Diet. Birds, speaks of the trachea (sub voce) of Platalea 
as divided by a cartilaginous septum. I have not found this to be the case 
with P. ajaja or P. leucorodia. 

- Also, however, in geese. 


distinctly charadriine. A groove runs forward from the 
nostrils, as in Scopus, &c. (q.v.) 

Family Phcenicopteridse. 1 Phoenicopterus agrees with 
the storks in the subdivision of the prebronchial air sacs by 
many septa into smaller chambers. But, as this also occurs 


FIG. 209. SYRINX OF Pliccnicoptcrus (AFTER WELDON). 
An, frout view ; Ac, lateral view. 

in Cliauna and to a much greater extent, less weight must 
be laid upon it than upon some of the muscular characters. 

In the storks, as in Scopus, but not the Anatidse, the pec- 
toralis major is divided into two distinct layers. This is also 
the case with Phoenicopterus. The tensor es patagii are 
closely similar in the birds under comparison and diverge 
from those of the cluck. The ducks are peculiar in the origin 
of the smallest head of the gastrocnemius from the biceps ; 

1 Our knowledge of the anatomy of the ' soft parts ' of Phoenicopterus is 
mainly due to WELDON (P. Z. S. 1883) and to GADOW (Joitrn. f. Orn. xxv. 
p. 382.) 


no such connection occurs in either stork or flamingo. That 
the deep flexor tendons of Phoenicopterus are not stork-like is 
surely related to the diminutive hallux of that genus. 

On the other hand the accessory femorocaudal, though 
small, is present in the flamingo, though absent in the storks, 1 
while the syrinx (fig. 209) is not stork-like. The CIECCL are 
long (three inches), but the intestines are not duck-like. 

The atlas is notched by the odontoid process, and the 
notch is very nearly converted into a foramen. There are 
nineteen cervical vertebras. None of the catapophyses fuse 
to form a canal. The transition between catapophyses and 
hsemapophyses is more complete than in most birds ; on the 
last cervical the two catapophyses are raised on a common 
platform, and on the first dorsal is the first (and last) hgema- 
pophysis, which is flattened and obscurely bifid. The last 
cervical and the first three dorsals are fused. Five ribs 
reach the single-notched sternum. The coracoids overlap at 
their insertion. 

The skull is desmognathous, with basipterygoid processes, 
to which the anterior ends of the pterygoids are attached. 
It is holorhinal with pervious nostrils. There are lateral 
occipital fontanelles. The lacrymals are large and rather 
duck-like, notched externally ; they nearly reach the jugal 
bar. There are no ossified ectethmoids. 

In including the Plataleidse and Ibididae with the Hero- 
diones I shall have the assent of most ornithologists. Alone 
among recent observers who have occupied themselves with 
the structure of the group, GAEBOD and FOBBES placed the 
spoonbills and ibises apart. The latter included them in his 
group Pluviales with the Charadrii, Eliinoclietus, &c. These 
diverse opinions about the Plataleidse appear to me to be 
largely due to the primitive position which they occupy 
among the Herodiones. They are to my thinking not far from 
a basal ' gralline ' stock. The Plataleidge have the complete 
muscle formula, and the biceps slip to the patagium. Some of 
these muscles have been lost in the other Herodiones, in which 

1 With the partial exception noted on p. 425, footnote. 


there is traceable a gradual series of modifications. In the 
storks the accessory femorocaudal is always absent, and in 
some the ambiens and femorocaudal also ; in the Ardeidae both 
ambiens and accessory femorocaudal are gone, and the 
modification of these leg muscles culminates in the aberrant 
Cancroina, where the formula is merely XY . In the 
Plataleidae (as in Tantalus] there are just faint traces of the 
basipterygoid processes, missing elsewhere in the group. 
The ibises are schizorhinal, and in Platalea the ends of the 
nasal grooves are rounded, thus tending towards the holo- 
rhinal, while the lower part of the bony nostril is wider than 
the upper, appearing thus to show a commencing occlusion 
of the schizorhinal nostril into a holorhinal one ; in A rdea 
t-in i' re a a faint trace of a former schizorhinal condition is 
seen in a slight groove which runs back from the end of the 
holorhinal nostril, as in Cluing a (see p. 144). The schizo- 
rhinal condition, as has been before pointed out, is probably 
the more archaic. Finally the Plataleida? have, according 
to MITCHELL, the most primitive form of gut among the 
Herodiones. Significant points of likeness between the 
Plataleidse and Grues are not wanting ; the occipital foramina 
and the impressions of the supra-orbital glands were among 
the facts that led FOKBES to associate them together. The 
convoluted windpipe is common to Platalea, Tantalus, and 
Grits. Both Platalea and Gnis have the complete muscle 
formula and schizorhinal nostrils. 

It has been often asserted that there are likenesses 
between the Herodiones and the accipitrine birds. This 
largely reduces itself to a comparison between the Herodiones 
on the one hand and the Cathartidss and Serpentarius on 
the other ; for if the two latter forms of accipitrine birds are 
rightly so placed the falcons must, on account of the various 
reductions in their structure, be derived from some form near 
to these, and cannot be their ancestors. There are two at 
first sight rather striking likenesses between the Herodiones 
and these lower accipitrines. GADOW has commented upon 
resemblances in the lie of the intestines, and MITCHELL has 
still further emphasised this likeness. In both groups (the 


falcons only among the Accipitres were examined) there is a 
tendency for the intestine to be thrown into spirally twisted 
loops ; but MITCHELL is of opinion that this is really no 
more than a convergent resemblance, for as the simpler 
types are considered the spiral arrangement becomes less 
and less obvious, thus indicating its special development, and 
independent development, in both. 

The second point concerns the syrinx, which as a special- 
ised organ is wanting in both storks and Cathartidse. In 
neither group are there intrinsic muscles. This point of 
resemblance rests, however, upon mere negativity. The 
details of the conformation of the lower part of the larynx 
are, as may be inferred from the preceding explanations and 
figures, extremely different in both ; it seems as if the syrinx 
has degenerated in both, but along quite different lines, the 
loss of the intrinsic muscles being about the only point in 
common. And we know from other groups that this muscle 
may be independently lost. There is, to my mind, as much 
to be said for a derivation of the Accipitres from the crane as 
from the pelargine stock, the fact being that we must 
probably seek for the origin of both from a low branch, 
perhaps common to all. The matter is further dealt with 
under the section devoted to the Grues. 

The skull of the flamingo is not duck-like. The back 
view of it, with the occipital fontanelles, might, it is true, be 
mistaken for that of an anserine bird ; but there are no 
salient likenesses elsewhere. There are but rudimentary 
basipterygoid facets, and the palatines have (which they 
have not in the duck tribe) a well-developed internal lamina, 
which, as is so often the case, is sharply bent downwards at 
its edge. In front of the fused maxillo-palatiiies there is no 
palatal vacuity, as in nearly all Anseres. The lacrymals, 
like those of the Anseres, are certainly large ; but their size is 
not so conspicuously marked in the length of their base of 
attachment to the margin of the orbit as in the length and 
great breadth of the descending process, not a feature of the 
anserine skull. The ectethmoids seem hardly ossified. The 
interorbital septum is largely deficient in front, as in Clianna, 


but is also more vacuolated behind, and in a different way from 
what is found among the few Anseres in which the inter- 
orbital septum is not solid. The length of the angular 
process of the mandible, though a duck character, is also 
found in the Plataleidte. There are so many characteristic 
duck-like characters wanting in the skull of Phcenicopterus 
that we cannot place it with that group. 

Of fossil Herodiones a considerable number of genera and 
species have been described. These range from the Cretaceous 
to cave deposits, and have been found in Europe, India, 
Mauritius, and Rodriguez. If Scaniornis, described lately 
by DAMES, 1 from the Cretaceous of Sweden, be really an 
ally of the flamingo-like bird Palcelodus, it is important to 
note that this family goes back further into the past than 
any other living family, so far as our information allows us 
to say. HUXLEY'S name of ' Amphimorphse ' for the group, 
and his remark to the effect that they are so thoroughly 
intermediate between the storks and ducks, will occur to the 
mind in this connection. This bird is known by a scapula, 
coracoid, and humerus. 

The other forms upon which new genera have been founded 
are not known by even so much as the scanty remains of Scanior- 
nis. Thus Palc&ociconia, Propclar/jns, Ibidopodia are only known 
by the tarso-metatarsus (incomplete), while the other genera are 
founded upon equally fragmentary remains. Tantalus Milne- 
Edwardsii 2 (nearly perfect tibio-tarsus), from middle Miocene, 
France. Palcelodus, however, is known by coracoid, scapula, and 
some of the ' long bones,' as well as the sternum, ' scarcely to be 
distinguished from the somewhat larger sternum of Phoenicopterus 
roseus,' furcula, one or two vertebras, and metacarpal bones. The 
bird seems not to have had such long legs as the modern flamingo, 
but longer toes. Elornis (Eocene and Miocene) was also a 
flamingo, but intermediate in the length of its legs between 
Phcenic<ij>t<'niN and Palcelodus. Agnopterus (Miocene) is reckoned 
a flamingo by FURBRINGER, placed ' inccrtcc scdis' by LYDEKKER. 

1 ' Uber Vogelreste aus dem Saltholmskalk von Limhamn bei Malmo,' Bill. K. 
SvensTt. Ak. Handl. xvi. 1891. 

- SHUFELDT. ' Fossil Bones of Birds,' &c., P. Acad. Nat. Sci. Pliilad. 1890, 
p. 507. 



Definition. Nostrils produced into tubes. Aftershaft present, aquinto- 
cubital. Oil gland tufted. Skull schizognathous, holorhinal. 
Two carotids present. Large supra-orbital glands. 

The peculiar form of the external nares, which has given 
to the group its name, characterises it. It is a character 
which is found in no other group, 1 but in all members of 
the present. The Tubinares are nearly the only large group 
of birds which can thus be diagnosed by a single character. 
The tube itself is, according to FOEBES, whose ' Challenger ' 
memoir upon the group forms the chief classic upon the 
subject, in Majaqueus, and probably in other genera, 
caused by a growth of the catilaginous walls of the nasal 
sacs. The degree of fusion between the two tubes varies in 
different genera. In Procellaria, for example, they quite 
coalesce and the external aperture is single. In Pelecanoi:l< ^ 
there is a distinct and not broad septum ; while in Bulwcria 
and others the septum is so broad that the tubes almost 
appear double. 

The petrels are web- footed birds with a small hallux, 
which in Pelecanoides is quite absent. The web does not 
take in the rudimentary hallux. The general number of the 
rectrices is twelve, but Ossifraga has as many as sixteen. 
The after shaft is always present. The pterylosis does not 
vary greatly ; the dorsal and ventral tracts are well separated 
upon the neck. The ventral pteryla is divided on the 
neck ; in the pectoral region each branch again divides. 
The dorsal tract divides at the middle of the scapulae into 
two, which unite later, thus enclosing a space. 

The tongue varies much in size and in the amount of its 
spiny bordering or covering. In Diomedea, for instance, it is 
much covered superiorly with spines. In (E 'sir 'data and 
others the tongue has a bordering of spines which are lateral 

1 It has been pointed out that the complex and somewhat protuberant 
nostrils of CMonis bear a little resemblance to those of the Tubinares, but are 
differently developed. 


as well as posterior ; in Daption capensis, and in many 
others, the spines are confined to the hind margin of the 
organ. The more usual condition is the intermediate 

There is a well-developed but small gizzard present. 
The cccca are absent in the Oceanitidae, but present, as a 
rule, in the Procellariidse ; they are small and nipple-like, 
and in Cymochorea appear to be reduced to a single csecum. 

The gall bladder is 
always present, and the 
lobes of the liver are 
equal or nearly so. The 
arrangement of the 
intestine is shown in 
fig. -210. The duodenal 
loop is double ; the 
greater part of the in- 
testine is drawn out into 
a considerable number 
of straightish loops. 

As to muscular 
auntomij, the great 
FIG. 210. INTESTINES OF Fulmar us glacia- pectoral is divisible into 

two layers by an inter- 
posed tendon, as in storks 
and Steganopodes. But in Lams, according to FURBRINGER 
(and in Podica also), the same division occurs, which tends 
to lessen the differences between the Tubinares and the 
Laridse, so insisted upon by GARROD, FORBES, and some 
others. The tensor es patagii in the Tubinares are com- 
plicated, but not in all the genera. In the Oceanitidte they 
are simplest. In these petrels the tensor brevis is a simple 
tendon. In Pelecanoides, and in some others, there is the 
additional complication that the tendon bifurcates near to 
its attachment on the extensor tendon, and gives off an 
anterior slip inserted more wristwards. In Prion affairs 
are still further complicated by the metamorphosis into 
tendon of the whole of the extensor inetacarpi radialix 

.r, short-circuiting vessel divided. 


superficialis. In (Extrcluta brevirostris we first meet with 
a recurrent slip going to the longus tendon and arising 
from the brevis in front of the termination of its anterior 
branch. The more typical tubinarian arrangement is seen 
in (Estrelata Lessoni. We find here the characteristic 
ossicle of the tendon of the brevis which is found in 
so many Tubinares, and which has been held by some to 
be a character of much systematic importance in differen- 
tiating the group a character which FURBRINGER thinks 
has been ' overvalued.' ! From this ossicle spring some 
of the fibres of the extensor muscle of the fore arm ; it is 
also the starting point of the recurrent tendinous fibres, 
which unite the brevis and longus tendons ; these tendons 
tire, moreover, in close apposition for nearly the whole of 
their course itself a characteristic feature of the Tubinares. 
In Ossifraga and in some other genera there are no 
ossicles, but the tendons are highly complicated. In some 
petrels for instance, in Diomedea exsulans the wing sesa- 
inoid is double, and in this bird also there are considerable 
complications of the various tendons. While, therefore, we 
cannot define the petrels by the arrangement of the tendons 
of the tensores, as is sometimes possible, it is evident that 
we have what might be expected in a large and important 
group a very considerable series of modifications of these 
organs. No petrel has, strictly speaking, a biceps slip, and 
the biceps itself, though perfectly normal in origin and 
insertion, has much more largely degenerated into tendon 
than is usual for this muscle. There is, however, a curious 
modification of this muscle in Pelecanoides and in a few 
others. Here the coracoidal head alone forms the muscle ; 
the humeral head goes entirely to the tensor patagii long us ; 
this slip is, therefore, as FORBES remarks, ' functionally a 
biceps slip.' Something apparently representing the true 
biceps slip is occasionally found in the Tubinares. In a 
few there is a tendon derivable from the humeral head of the 

1 Justly, as it is found not only among the Steganopodes, which may IK- 
fairly re^in''!'-'! as allies of the Tubinares, but also in so remote a type as 

Me i -nps ! 


biceps, which appears to end on the fascia of the wing. 1 
The expansor secundariorum, found only among the 
Oceanitidse, is peculiar in that it arises from the surface of 
the pectoral ; its muscular belly is, as usual, at the elbow, 
and the tendon is joined by a branch from the scapularies. 
The anconffus has a well-marked tendinous attachment to 
the humerus. 

The muscles of the hind limb, to which GAEEOD attached 
so much classificatory importance, vary much in the group. 
The ambiens is present in all except Fregetta and Pele- 
canoides. In Garrodia and some others the tendon does 
not cross the knee. 2 All have a femorocaudal, but the 
accessory is absent in Bulweria and Pelecanoides. The 
semitendinosus has an accessory in the Oceanitida?, but not 
in the others. It is inserted separately from the semi- 
mernbranosus. The deep flexors blend about halfway down 
the leg ; but when a hallux is present it receives no slip 
from the conjoined tendons. The syrinx of the Tubinares 
shows an interesting series of gradations, from a quite 
ordinary tracheo-bronchial type to what is very much like the 
bronchial syrinx of the Caprimulgidae, though FUEBEING-EE, 
while admitting the ' bronchophone tendency ' of the syrinx 
of the Strigidse, as a point of similarity between that group 
and the Caprimulgidse, considers that only ' artificially ' can 
the Tubinares and the Spheniscidse be brought into the same 
line. Nevertheless in the series of Tubinares the bronchial 
rings to which the intrinsic' muscles are attached seem to 
move further down. FOEBES, however, regards this as a 
splitting of the trachea, and holds that the intrinsic muscles 
are invariably fastened to the fifth semi-ring. ' It is in the 
genus Pelecanoides,' remarks FOEBES, ' that the typical 
construction of the syrinx of the Tubinares is seen in its 
simplest form.' In this bird all the bronchial rings are 
semi-rings, and there is a three-way piece of the usual 

1 In a specimen oi^Nycticorax griscus on one side of the body I found a 
tendon from the biceps running I to the tensor brevis tendon, which may be com- 
parable to the above-described slip. 

- Cf. as to this 0>isthocomus, CEdicnemits, and Ccisuarins. 



structure. In Garrodia the first three bronchial rings are 


complete. In Thalassoeca glacialoides the last four tracheal 
rings are incomplete behind, and are quite like the four 
succeeding bronchial semi-rings, being, moreover, like them, 


FIG. 211. SKULL OF Diomedea cxulans 

Vo, vomer ; ?,fxp, maxillo palatines ; 1't, pterygoid. 

FIG. 212. SKULL OK Procellaria 
girjantea (AFTER HUXLEY). 

PI, palatines ; *, basipterygoid process. 
Other letters as iu fig. 211. 

ossified. This modification is carried to its furthest extreme 
in Ossifraga. In this petrel the last nine or ten tracheal 
rings are incomplete in front, and the last twelve or so are 
incomplete behind. Thus the membrana tympaniformis 
does not commence for some distance away from the bifur- 
cation of the respiratory tube, the rings being double and 

G G- 


complete for the same way in front of it. The division of 
the trachea is carried upwards above the point where it is 
externally divided into two tubes by an internal septum. 
And in Adjust.'*, though there is no external division of 
the trachea, there is an incomplete internal septum. In 
Diomedea, according to SWINHOE, at any rate in two 
species, the bronchi are long and convoluted, as in certain 

The petrels are schizognathous and holorhinal birds, 
some with basipterygoid processes. The skull possesses a 
small peculiar bone, which has been termed the ' ossiculum 
lacrymo-palatinum/ l whose relations are sufficiently in- 
dicated b} T its name. This bone occurs in many genera ; 
but it is said to be of no great classificatory significance, 2 
since it also occurs in birds so remote from the petrels and 
from each other as Cuculidae, Cariama, and Laridae. The 
skull is markedly excavate above the orbits for the supra- 
orbital glands. The skull of an albatross (Diomedea melcuio- 
phryx) is in more than one particular like that of the 
steganopods. It is, it is true, schizognathous ; but the 
interval separating the maxillo-palatines and the palatine 
expansions of the maxillae in front of these is very slight ; 
and, as FOEBES has observed, the downwardly curved 
extremity of the vomer partly fills up this gap, though there 
is 110 actual fusion between it and the maxillo-palatines. 
The hooked bill is Fregata-like; and the closely approxi- 
mated and downwardly produced internal laminae of the 
palatines are highly suggestive of the pelican. There are 
always fifteen cervical vertebra. The sternum varies so 
much, from having a plain hinder contour to the presence of 
two notches on each side, that its characters need not be 
given in detail. 

The petrels can be divided into two families, the 
Oceaiiitidfe and the Procellariidae, which may be thus 
defined : 

1 By BRANDT. PARKER has termed it ' uncinate.' 
See, however, under Steganopodes, p. 41(3. 



Tarsi ocreate or covered by large transwrsely 
oblique rentes anteriorly. No cceca. Expansor 
secundariorum present. Accessory semitendinosus 

prese'nt. Ainbiens (ichcii present) docs not pass 
over knee. Wing shorter than leg. 


Tarsi covered iritJt Jir.ragunal scnteUa. Cceca 
present (except in Halocyptena). No expansor 
secundariorum or accessory semitendinosus. 
Ambiens (absent in Pclecanoides) crosses knee. 
Leg sJiorter than icing. 

Not many fossil Tubinares are known. If Hi/drornis natator 
of MILNE-EDWARDS, from the lower Miocene of France, be referable 
to this group, it has a considerable antiquity. Diouicdrd anglica 
has been lately described by LYDEKKEB. 1 

The nearest allies of the petrels appear to be the 
Steganopodes (q.v.) 


Definition. Aftershaft rudimentary. Aquincubital. Oil gland tufted. 
Muscle formula, A(B)XY + . No biceps slip. Expansor secun- 
dariorum present. Carotids, two. Caeca large. Skull desmo- 
gnathoxis, holorninal, with, basipterygoid processes. Ribs without 
uncinate processes. 

This well-marked group of birds has only two genera at 
most, which collectively contain but three species. It has 
been chiefly investigated by PARKER, 2 GARROD, 3 MITCHELL,' 
and myself/"' 

1 In Quart. Journ. Geol. Soc. xlii. p. 366. 

- ' On the Systematic Position of the Crested Screamer,' P. Z. S. 1863, 
p. 511. 

3 ' On the Anatomy of Cliauna derbiana,' &c., ibid. 1876, p. 189. 

1 On the Anatomy of Cliauna chavaria,' ibid. 1895, p. 350. 

3 ' On some Points in the Anatomy of Cliauna chavaria,' ibid. 1886, p. 178, 
and BEDDAKD and MITCHELL, 'On the Anatomy of Palon/fi/fii,' ibid. 1894, 
p. 536. 

G G 2 


The skin in these birds is excessively emphysematous, 
save only on the shoulder, and in Palamedea and Chauna 
derbiana on the tibia. The feathers, however, do not per- 
forate the cutaneous air cells, ' but cause the skin to be 
indented where they are inserted.' The aftershaft, though 
present, is confined to the feathers on the nape of the neck. 
The rectrices are twelve in Chauna and fourteen in Palame- 
dea. The oil gland in Ch. derbiana has a single orifice on 
each side ; it is encircled by feathers which constitute the 

In Gh. char aria the summit of the oil gland is covered 
by feathers, a line of which separates the two orifices of the 
gland. In Palamedea the same is the case, but the encircling- 
ring of feathers is not complete on the ventral side. The 
pterylosis is almost unique in the fact that there are no 
apteria, except, indeed, a space in the axillary cavities, and 
these are covered with down feathers. The strong horny 
spur borne on a bony core, an outgrowth of the first meta- 
carpal, is comparable to a thickened featherless patch of skin 
in a corresponding situation in Sarcidiornis. 1 

The patagialis muscle is not reinforced by a biceps slip ; 
the brevis tendon is single, but broad, and without a patagial 
fan. The expansor secundariorum is present and ' ciconiine.' 
The insertion of the deltoides posterior is extensive in Pala- 
medea for about three inches down the humerus. The 
anconaus has a well-marked humeral head. The division of 
the biceps commences in the fleshy belly of the muscle. 
In all the members of this family a very peculiar muscle 
exists, to which Mr. MITCHELL and I have given the name of 
costo-sternalis externus. It arises from the third, fourth, 
and fifth ribs by a tendinous head, and is inserted on to the 
costal edge of the sternum half an inch from the posterior 

The muscles of the leg are complete, as regards those 
upon which GAEROD laid stress in his classification, in 
Chauna. In Palamedea, however, the accessory femorocau- 
dal is absent. The biceps has not an anserine insertion, but 

1 Not, of course, to the carpal spur of Plectropterus. 



passes through a loop in the ordinary way. Both peromal 
muscles are present ; the tibialis anticus is split just at its 
insertion. The deep flexor tendons differ somewhat in the 

FIG. 213. C.T.CA OF Chauna chavaria (AFTEK BEDDARD). 


three species. In Palamedea the flexor longus hall nets is 
slender, and gives off a vinculum to the flexor commit n is 
tendon, before supplying the first digit ; in Cliauna derl'unM 
there is no branch to the first digit at all ; Cliauna char aria 

is like Palamedea, but there are two 
distinct vincula. 

The palate (of Ch. chavaria at any 
rate) is provided with three longitudinal 
rows of papillae ; the tongue is just over 
an inch long, and its base is edged with 
spines. There is no transverse constric- 
tion or oblique groove, such as is found 
in many anatiform birds. The proven- 
tricidus is peculiar ; there is a narrow 
zone of glands round the cesophageal 
aperture, from which a broad triangular 
patch extends down one side of the 
cavity. The gizzard is decidedly small. 
The lobes of the liver are more nearly 
equal in size in Cliauna than in Pala- 
medea ; in all there is a conspicuous gall 
bladder. The cfsca (fig. 213) are in 
some respects unique in structure ; they 
are in the first place large, measuring 
three inches or so in length ; they are 
sacculated by a single band. 

The windpipe agrees with that of 
some of the Anseres in having two pairs 
of extrinsic muscles : : the upper pair is 
inserted into the middle of the membrane, which runs 
between the coracoid and the corresponding limb of the 
furcula ; the lower pair close to the costal processes of the 
sternum. The intrinsic muscles cease some little way in 
front of the syrinx. 

The prebronchial and subbronchial air sacs are, in Ch. 
chavaria at least, much divided up, as in the storks. In 

1 Apparently first noted by CRISP, ' On the Visceral Anatomy of the Screamer,' 
P. Z. S. 18G4, p. 14. 

Palamedea (AFTER BED- 



Chauna the lower pair of extrinsic muscles fan out upon the 
aponeurosis of the lungs. 

There are eighteen cervical vertebra in Palamedeaaxidim 
Ch. chavaria, nineteen in Ch. derbia/ia. There are seven 
complete ribs in Palamedea, eight in Chauna. The sternum, 
which has one notch on either side, has neither external nor 
internal spina. There are in neither genus any traces of 
uncinate processes on the ribs, a 
character which is unique among 
living birds. 1 

The skull of the Palamedeidse 
lias many anserine characteristics, 
which have been emphasised by 
PARKER, and perhaps rather too 
lightly touched upon by GARROD. 
In their desmogiiathism (which is 
a complete fusion) they are, of 
course, anserine, the form of the 
maxillo-palatines most recalling 
those of Mergus. The form of the 
palatines is duck-like or gallina- 
ceous in the rudimentary character 
of the internal lamina, which is a 
mere ridge. The pterygoids are 
articulated to the large oval duck- 
like basipterygoid processes, nearer 
to their middle than is the case 
with the Anseres. The lacrymals 
are small, quite contrary to what 
is found among the Anseres, 

though the ectethmoids are not unlike those of the 
latter group. The anterior part of the face, with the clear- 
cut holorhinal nostrils and the hooked bill, is suggestive of 
Cariama or a gallinaceous bird ; it does not at least recall the 
duck or goose. The interorbital septum is deficient in front, 

FIG. 215. SKULL OF Chauna 
dcrbiana. VENTRAL ASPECT. 

1 Not even a rudiment of these characteristically avian structures has, so far 
as I am aware, been detected. 


but not fenestrate posteriorly. Occipital fontanelles are 

Though perhaps rightly placed in the neighbourhood of 
the geese, it is obvious, from what has been said, that the 
Palamedeida? are distinguished from them by many differ- 
ences, of which the most important are perhaps 

(1) Continuous feathering. 

(2) Absence of biceps slip to patagium. 

(3) Peculiar form of intestinal casca. 

(4) Normal character of biceps cruris. 

(5) Emphysematous character of skin and breaking up of 
cervical air sacs. 

On the other hand there is nothing in the skull which 
forbids an association with the Anseres, and the windpipe, 
with its two pairs of extrinsic muscles, is decidedly goose-like. 
But it must be remembered that this feature is also found 
among Galli, and in a few other forms. 


Definition. Oil gland tufted. Aftershaft small or absent. Aquin- 
cubital. Two carotids. Trachea with, two pairs of extrinsic 
muscles. Casca long. Gall bladder present. Biceps slip present. 
Glutseus maximus large. Muscle formula of leg, ABX + . Skull 
desniogiiathous, with, basipterygoid facets. 

The swans, geese, ducks, and mergansers, which make up 
this large assemblage of birds, are all aquatic, or semi-aquatic, 
in habit, and in correspondence have webbed feet, with the 
exception only of Anseranas. They are also for the most 
part strong flyers, excepting only the living Tacliyeres cinereus 
and the extinct Cuemiornis calcitrans. 

Tacliyeres cinereus, the ' steamer duck,' from the shores 
of Patagonia, has been investigated by E. 0. CUNNINGHAM. 2 
It does not appear from his memoir, which relates chiefly 
to osteology, but in the course of which he describes and 

1 H. SEEBOHM, ' An Attempt to diagnose the Sub-Orders of the Ancient 
Ardeo-Anserine Assemblage,' etc., Ibis, 1889, p. 92. 

2 ' On the Steamer Duck,' Tr. Z. S. vii. p. 493. 



figures the gizzard and windpipe, that there are noteworthy 
modifications of structure. The bird flies strongly when 
young, but swims only when adult. CUNNINGHAM suggests 
a greater density of the bones ; but when weighed and com- 
pared with the skeleton of some other Anseres no reliable 
differences were apparent. 

Cnemiornis calcitrans is a large anserine bird from the 
Pleistocene deposits of New Zealand, which, according to 
Sir JAMES HECTOR/ stood over two feet high, and was at 
least three feet in length. Cnemiornis is characterised by 
the great weight of its bones ; the following comparative 
table from HECTOR'S paper brings out this fact :- 



< 'iii'iniontis 



















It seems clear from these comparisons that Cnemiornis 
could not have been a flying bird. Moreover the sternum 
has a keel whose highest elevation is under three lines. 
This bone appears to have possessed no lateral notches, but 
only a slight median concavity. On the other hand the 
' rough tubercular surface ' of the keel is perhaps a little 
suggestive of a missing cartilaginous piece. 

The suggestion of FURBRINGER to separate Cnemiornis 
into a sub-family, Cnemiornithinae, seems perhaps, in the 
light of the above facts, to be justifiable. But the remainder 
of the members of the family cannot be divided with precision. 
It is above all impossible to divide swans, geese, and ducks 
from each other into three such groups. Nevertheless, as 
will be gathered from the following account of their struc- 
ture, the Anseres vary very much among themselves more 
especially in the structure of the windpipe. 

All the Anseres, as noted in the definition, have a tufted 
oil gland. 

1 ' On Cnemiornis calcitrans,' &c., P. Z. S. 1873, p. 7G8, and 1874, p. 307. 


The aftershaft is always small, and sometimes absent. 

In the 2)terylosis (according to NITZSCH) the pectoral tract 

divides about halfway down the 
neck ; it is broad, and gives off on 
each side a stronger outer band. 
The dorsal tract also divides 
about halfway down the neck, and 
encloses a long narrow space. 

The number of rectrices in 
various genera is shown in the 
annexed table. 


Bizinra lobata . . . .24 

Anas specularis . . . .14 

Aims boschas . . . . .14 

Aix sponsa ... .16 

Erismatitra rnbida . . . .18 

Cereopsis Novce Hollandics . . 14 ' 

Plectropterus gambcnsis . . .16 

Cygnus musicus . . . .24 

Fuligula rufina . . . .16 

Dafila acuta ... .16 

Bernicla canadensis . . 18 

Mergus merganser . . . .18 

,, albicillus . . 16 

Bizinra lobata (AFTEK 

p, pouch : /, tongne ;f.l, fra?iium 

The tongue is strong and 
bordered with spines. In Mergus 
only is it thin and pointed. 

The liver has a large gall 

Biziura lobata is unique among 
Anseres 2 in having a subgular pouch (fig. 216), formed by a 
duplicature of the fraenum linguae. 

The most noteworthy point that appears from the facts 
given on the next page is the gradual reduction of the caeca 
in the mergansers and smews, and the great range of variation 
in size which they exhibit in Cygnus iiit/r/coUi.s, a variation 

XIT/SCH says 16. 

- Cf. Glides for si similar structure. 



which does not, it will be observed, bear any relation to the 
variation of the total length of the intestine. 

The following are some intestinal measurements : 


L. I. 


Ft. In. 



Mcnjiis merc/ftnscr 

5 4 



,, castor .... 

4 8 



,, albicillus 

4 3 


2 ' 

Biziura lobata . 

8 6-5 
6 11-25 




Anas specularis ... 5 2-5 



Bhodonessa caryophyllacea . 4 

2-2, 1-75 

Aix sponsa .... 41-2 



Sarcidiornis melanota $ . . 47 


2-75, 3 

,, carunculata $ 36 



Bernicla canadcnsis ... 76 


jitbata 3 . 3 4 


2-5, 2-75 

Chcnalopex j ubata ... 4 



1 U 


Cygnus nigricollis . . 11 4 



9 6 


Dendrocygna aiit/imnalis . 

3 6 


3-2, 3-4 

A i n-K 



Mi-tojiitina pcposaca . 

n -L \J r^f 

4 3 

5-5, 6 

Plectropterus gainbcnsis 

6 9 


6-75, 7-25 

nificr $ 

6 1 



< 'iTi-ii/isis Novce Holland-ice 6 



Anser albifrons 3 ... 58 



,, indie us $ . . ' . 

5 1 



Cliloepliaya ina/jcllanica <$ 

6 8 



< '//i/itus olor c? . 

11 9 



ferns 3 

8 6 




8 4 



., buccinator $ 

O ^ 

8 6 





atratus $ . . . 92 



Tadorna tadornoides % . . 56 


3-5, 4-5 

Dafil/i ficutaQ .... 46 



Pcecilonetta bahamensis . 210 



bahamensis $ . 34 



Qucrqitedula circia ? 



Hymenolamus malacorhynclms % 



3-8, 4-25 

Spatula clypcata $ 

8 9 
6 6 



3-5, 3-75 

Fuligula ferina 3 

4 2 




4 5 




r cj 

4 2 


6-5, 7-5 

Nyroca leucophthalma $ . . 48 


1 i inly one 

The muscular anatouy of the Anseres is very uniform, 
which coincides with their uniformity of life habit. Even 
presumed aberrant types, such as Mergns, Biziura, and 



Tachyeres, are hardly if at all to be distinguished anatomically 
from the typical geese and ducks. Our knowledge of the 
muscular structure of this group of birds is mainly due to 
FUEBEINGEE, who has illustrated the fore half of the body 
by two double plates referring to Anser cinereus. 

An interesting duck character (also, however, found in 
Colymbus, Tinami, and some Galli) is the meeting of the two 
great pectoral muscles over the carina sterni. In Mcrgus 


mergaiixcr, for example, they blend for a space of half 
an inch. 

There is a biceps slip present, 1 and this has at least 
sometimes a peculiar arrangement, which is remarkably like 
that of the Colymbi. In Anser cinereus FURBRINGER 
figures the biceps slip as attached in the ordinary way to 
the tendon of the tensor patagii longus ; but before it is thus 
attached it gives off a slender tendon, which exactly as in 
the Colymbi runs over the patagium and is inserted on to 
the fore arm in front of the broad and diffuse tensor patagii 
brevis tendon. 

In Anas FURBRINGER figures this tendon as joining the 

1 Not in Cygnus Bciciclii. 



patagial fan, a state of affairs which is exactly paralleled in 
jgZchmophorus (see p. 387) . In ducks there is a peculiarity in 
the biceps to which attention appears to have been first 
called by FURBRING-EE. There is a tendinous sheath 
partly covering the patagial side of the muscle and derived 

Sarcidiornis mclanota $ 

FIG. 219. SAME OF 
S. melanota 9 


FIG. 220. SAME 
Rhodonessa caryo- 
phyllacea 9 (AFTER 

from the pectoralis primus. A similar structure occurs in 
Colymbi and storks, and in Casuarius its homologue is more 
independent of the biceps. The latter at any rate, if not 
the former, seems to me to correspond to a peculiar muscle 
found in the tinamous (cf. p. 489). 

The expansor secundariorum shows some variations. In 
Biziura lobata it is entirely absent. In Aix sponsa the 
tendons thin off and are lost on the interthoracic septa. In 
others, as, for instance, Bernicla canadensis, the expanded 
tendons end upon the oesophagus ; on the w r ay thither they 


blend with the sheath of the carotids, which when pulled 
upon they compress, as also the jugulars. This may con- 
ceivably be a provision for increasing the blood supply of the 
wings by interrupting that of the head. Finally in Cijgnus. 
Mergus, and Sarcidiornis the expansor secundariorum is 

The deltoid has a strong scapular slip. The an con (BUS 

Fiu. 221. SYKIXX OF R. caryopliyllacea $. EIGHT-HAND FIG., FRONT 

has a humeral slip. In the hind limb all Anseres have the 
formula ABX+. The femorocaudal is slender, the acces- 
sory very large. In Biziura lobata the ambiens has a pecu- 
liarity found also in Phalacrocorax and the extinct Hesper- 
ofnis that its tendon perforates the patella. The glutaus 
maximus is very large, and its origin descends below the aceta- 
bulum. A marked peculiarity of the Anseres, found else- 
where in Struthio, is that the biceps femoris gives off a slip 
to the gastrocnemius. In most Anseres the flexor lone/us 
liallucis gives off a slip to the hallux before fusing with the 



flexor profundus ; in Biziura lobata 
the same slip is given off, but it be- 
comes lost on one of the annular masses 
of fibrocartilage surrounding the other 

The windpipe in the Anseres is nearly 
always straight, the only exception 
among the ducks and geese as yet 
recorded being Anseranas melanoleuca, 
in the males of which the trachea forms 
a double loop, extending to quite the 
end of the pectoral muscles. A 
second peculiarity of the trachea is 
seen in the males of Metopiana pepo- 
saca, Strictonetta ncevosa, 1 Melanitta 
fusca, Nyroca, Mcrgus, Somatcri-a, and 
Clanc/i/hi. In these (see fig. 222) there 
is a bulbous enlargement of the trachea 
some little way in front of the syrinx. 
It is present, but very slightly developed, 
in FuUgula rufina. In nearly all the 
ducks the syrinx in the male has a re- 
markable asymmetrical enlargement, 
which is as a rule entirely bony, but is 
sometimes (M erg us, Clangula, Nyroca, 
FuUgula) mainly formed of membrane. 
The accompanying figures will give 
some idea of the form of this structure, 
which shows differences in different 
species. The figures are taken from 
a memoir upon the subject by Professor 
GAREOD. 2 YARRELL 3 and EYTON 4 have 

1 E. P. RAMSAY, ' Note on the Tracheas of certain 
Australian Ducks,' Proc. Linn. Soc. N. S. W. iii. 
1879, p. 154. 

2 ' On the Form of the Lower Larynx in certain 
Species of Ducks,' P. Z. 8. 1875, p. 151. 

3 British Birds. 

1 Monograph of the Anatidce or Duck 

FIG. 222. WINDPIPE or 
Metopiana peposaca $ 


also figured a good many of the syringes of the ducks in 
illustration of this matter. The only ducks in which there is 
certainly no modification of the syrinx of this kind are 
Biziiira lobata, (Edemia nigra, and Melanitta fusca ; in the 
former bird FORBES l has described a plain syrinx (fig. '2'Jo) 
with a box at the bifurcation of the bronchi formed of the 
last tracheal and of a few of the anterior bronchial semi-rings. 

He suspects, however, that the 
genus Erismatura 2 will be found 
to have a similar syrinx minus a 
lateral outgrowth, as MAcGiLLi- 
VEAY appears to have described 
something of the kind. Somateria 
molUssima has a very slight sym- 
metrical enlargement of the syrinx. 
A very marked characteristic of the 
Anseres is the possession of two 
FIG. 223.-SYRIXX OF Biziura $ g of extr i ns i c tracheal muscles. 


In this they agree with Pala- 

medeidas. The single pair of intrinsic muscles are as a rule 
attached to the third or fourth tracheal ring in front of 
the syrinx in the ducks. 

Among the Cijgnince (swans) there is frequently a looped 
trachea, the coils being intrasternal. This is so with both 
sexes of C. ferns, C. buccinator, C. americanus, and C, 
Bewicki ; there appears to be a trace of the looping in C. 
atratus. In C. olor, C. immutabilis, C. nigricollis, and C. 
coscoroba the windpipe is straight in both sexes. It is 
interesting to note that in C. buccinator, at any rate, the 
intrinsic muscles do not follow the coils that bridge across 
the loop. This species of swan is also remarkable for the 
extraordinary dilatation of the middle of each bronchus, which 
is, again, characteristic of both sexes. These dilatations would 


be almost spherical were it not for the irregular crumpling 

1 'A Note on some Points in the Anatomy of an Australian Duck (Biziitrn 
lobata),' P. Z. S. 1882, p. 455. 

2 ' There is no expansion or tympanum, as in other ducks ' (Orn. Bioyr. iv. 
1838, p. 331). 


here and there ; their diameter is about an inch. The 
bronchial rings which cover them have largely lost their 
individuality, and form an irregular network of partly 
cartilaginous and partly osseous bars. 

The males ('? as to females) of Cijgnus ferns and C. 
Bewicki at any rate show less marked traces of the same 

In the former there is a distinct fusiform dilatation, but 
further down the bronchi than in C. buccinator, between 
the rings of which there is some slight formation of 
anastomoses. In C. Bewicki both features are still less 

Cygnus olor, C. atratus, C. nigricollis, and C. coscoroba 
have no trace of this remarkable structure. I have examined 
males of all and females of two. 

As to the geese, Bernicla canadensis, Anser indicus, and 
Gereopsis Novce Hollandia (and doubtless many others) have 
a syrinx without the anatine bulbus. It is present in 
Plectropterus (gambensis, Ruppeli, niger) and Sarcidiornis, 
and present, though small and solid, in Chenalopex jubata. 
It is fenestrated in Plectropterus, not so in Sarcidiornis. 

Dendrocijgna appears to illustrate the commencement of 
the syringeal enlargement. In the male of D. arcuata the 
last twelve tracheal rings are widened and enclose a spacious 
chamber about twice the diameter of the rest of the trachea. 
The intrinsic muscles are attached to the beginning of this 
thin-walled box. In the female there is an indication of 
this in the fact that eleven of the tracheal rings in front of 
the last three are imperfect posteriorly, being closed by 
membrane. The intrinsic muscles also are attached opposite 
to the commencement of this modified region of the trachea. 

In D. autumnalis there is the same box, which is 
strengthened posteriorly by a strong bony bar. The wind- 
pipe of the female has no such modification as has been 
described above in D. arcuata. 

It is possible that this single enlargement in Dendrocijgna 
is the beginning of both the tracheal swelling and the 
syringeal bulbus in Mergus, &c. 

H H 


The number of cervical vertebra 1 among the Anseres varies 
considerably. The smallest number is found, for example, 
in Plcctropterns gambcnsis, Biziura lobata, and Taclujcrcs 
cinercns, where there are only sixteen. In (Edemia niyra 
there are seventeen ; among the swans, twenty to twenty- 
four. The number of true ribs also varies considerably. 
The smallest number is to be seen in Cereopsis, where there 
are but five. Tacliyeres, Plectropterus, and a number of 
other genera have seven; there are eight in Tadorna vulp- 
ii user, and as many as nine in certain swans and geese. 

The sternum^ which has a moderate spina externa, but 
no spina interna, is whole in Cncmiornis, but has two 
notches or foramina in other Anseres. The coracoids come 
into contact, but do not overlap at their sternal articulation. 
The procoracoid is small, and does not reach the clavicle, 
which, however, reaches the scapula. 

The skull has large oval sessile basipterygoid processes. 
It is holorhinal and desmognathous. There are frequently 
lateral occipital fontanelles, as in many ' pluvialine ' birds. 2 
The palatines are remarkable for the rudimentary character 
of their inner laminae, which brings about a resemblance to 
the gallinaceous birds, as has been pointed out by HUXLEY, 
and, it may be added, to the parrots. That part of the 
palatine is only indicated by a not well marked ridge which 
is totally absent in Bernicla leucopsis, Chen ccerulescens. 
The general direction of the bone, therefore, is oblique ; it is 
only near to the attachment with the pterygoids that it 
becomes feathered out in a horizontal direction. That, at 
least, is the more normal arrangement ; for in Mergus the 
greater part of the bone has its upper and lower surface co- 
incident with the horizontal axis. This, too, is the case with 
Biziura lobata. 

The oval basipterygoid facets for articulation with the 

1 For osteology see PAKKEH, ' On the Morphology of the Duck and the Auk 
Tribes,' C unnincjhdm Memoirs B. Irish Ac. No. 6, 1890, and SHUFELDT, ' On N. 
American Anseres,' P. U. ,S. Nat. Mus. xi. p. 215. 

- Absent in Cereopsis, Biziura, Cygnus, and Cncmiornis. See OWEN, Tr. Z. 
S. ix. pt iii. 



pterygoids are placed so far forwards that the anterior ends 
of those bones articulate with them. 




l-'iii. 2'24. SKULL OF Quo'^it^iliiln crecca. LATEKAL VIEW. 

/'.. frontal : Set, nasal : I'm.i; i>renia.\illa : I'n. voiuer : l'. palatine: I't. i>ter\ -^niii. 

The vomer is a thin deepish plate of bone which is more 
or less intimately connected with the 
median septum and maxillo-palatines in 
front. The maxillo-palatines are com- 
pletely fused across the middle line in 
many Anseres (e.g. Clicn, Hijin,'iin],nnus} ; 
in Biziura and Neryus they come into 
contact but are not fused. The latter 
has very un-ducklike palatine in that the 
somewhat delicate maxillo-palatines diverge 
from each other after their junction pos- 
teriorly as well as anteriorly, the palatine 
vacuity in front being (for a duck) un- 
usually extensive. In CJtcii <'a'ritl.<'sc<'iix, 
indeed, secondary bony growths have 
almost completely obliterated this vacuity, 
a kind of ' false ' palate having been 

The lacrymal bones are large, having 
a considerable length of line of union with 
the skull ; they are sometimes (e.g. Chloe- 
pliaya) and sometimes not (e.g. Mertjux} 
ankylosed with the orbital margin. Cere- 
opsis has (among the genera which I have examined) a skull 

H II 1? 

Fin. 225. VENTRAL 

M.i'fi. inaxilln-iiiiLitiiies. 
Other IrtUTsasiuti^. -'-'4. 



which is peculiar in that posteriorly the lacrymal is free 
from the orbital wall, but is fused with a process of the 
frontal at the anterior end of the supra-orbital impression, 
leaving (as in Cliionis and some other birds) a foramen. 1 In 
Cereopsis the descending process of the lacrymal curves back- 
wards and comes near to the zygoma ; 2 the junction is 
completely effected, and there is ankylosis, in Denclrocygna. 

The ectethmoids of the Anseres are often largely deficient 
as ossifications ; when present they are thin-walled bubble- 
like structures, coming into relation with the lacrymals. 
The interorbital septum is generally very complete ; but it 
is largely vacuolate in Mergus and Eiziura. 

The following table shows the number of cervical 
vertebra and the condition of the hsemapophyses and 
catapophyses in a series of anserine birds :- 

C. V. 



Cat. nearly 


Plectropterus gambensis 
Metopiana peposaca . 
Sarcidtoniix carunculata 



C12, 13 
C12, 13 
Cll, 12 


Hymenolamns . 



Oil, 12 


Tacliyeres cuicrcns 



Cll, 12' 


(Edeinid iin/ra . 



CIO, 11 


Biziura lobatu . 



Cll, 12 


Dendroci/t/ini antnmnalis 



Cll, 12 


Bernicla brvnta . 




CIS, 19 - 

1 Fused to form a solid bitid ' hffimapopliysis,' as iu some other birds. 
'- They are rudimentary and do not mount upon the hypapophysis. 

As a possible appendix to the Anseres must be mentioned 
three or four species of an extinct genus of birds, Gastornix, 
flightless and larger than an ostrich. It has been found 
only in Europe and from Eocene beds. It is placed among 
the ' ratites ' by LYDEKKEB and some others ; this is largely 
on account of the coracoid, which is imperfect above, and 
appeared to LEMOINE and others as probably ' platycora- 
coidal.' FURBBINGEE, however, considers that the tuberosity 

1 This is figured by OWEN in his paper on Cnemiornis, Tr. Z. S. ix. pi. 35, 
fig. 8. 

2 Sometimes joining. Cf. OWEN, loc. cit. pi. 35, fig. 6. 


described by the latter as ' tuberosite pregleno'idienne ' is 
really the broken end of the scapula, which would be thus, 
as in Didus, ankylosed with the coracoid, and would also 
form with it an angle approaching to a right angle. The 
supposed remains of the scapula, on the other hand, are for 
FtiRBRiNGER the acrocoracoid. On this interpretation the 
shoulder girdle of Gastornis would be a nearly typical 
carinate shoulder girdle. The length and slenderness 
of the coracoid too is not a ratite character, but it does 
ally Gastornis with Cnemiornis (and also for that matter 
with Pliororliacos). In spite of the freedom of the meta- 
carpals (a character only known elsewhere in Archceo- 
pteryx), the complete furcula and various points in the bones 
of the lower limb, pointed out by NEWTON, together with 
the facts already referred to, seem to point to a greater like- 
ness to Cnemiornis than to any other known type. The 
skull had basipterygoid processes and seemingly teeth in 
sockets. The pygostyle appears to have been at most very 
small and probably absent. 


Definition. Small toothed birds with, carinate shoulder girdle and 
sternum. Bones of pelvis not united. Quadrate single-headed. 
Vertebrae amphicoelous. 

This group of birds, from the Cretaceous of North 
America, has been investigated by MARSH, who, in his great 
work upon the toothed birds, placed it in a group Odonto- 
tormse, as opposed to Odontolcae (Hesperornis), on account of 
the fact that the teeth are implanted each in a distinct 
socket. Its relationship to other groups is doubtful ; but it 
is probably not greatly misplaced if we consider it in the 
neighbourhood of the stork and plover tribe, as has been done 
by FURBRINGER. The group contains two genera Iclitliy- 
ornis and Apatornis. Of the former MARSH refers to several 
species, viz. I. dispar, I. victor, I. validus, I. tener, I. agilis, 
and I. anceps. 

Of Apatornis there is but one species known, A. celer. 


These two genera comprise a number of ' small birds, 
scarcely larger than a pigeon. In their powerful wings and 
small legs and feet they remind one of the terns, and accord- 
ing to present evidence they were aquatic birds of similar 
life and habits.' 

The restoration of Ichthyornis given by MAESH has 
been extensively copied in various works, in some of which 
it would appear as if our knowledge of the osteology of the 
species selected were greater than is really the case. It has 
been made, for example, to show schizorhinal nostrils and a 
pelvis constructed after the carinate type, with the ischia and 
ilia fused. It is not known w T hether the skull was schizo- 
rhinal, as only the calvarium and the lower jaw and a 
fragment of the upper jaw have been discovered. The skull 
has well-marked grooves for the supra-orbital glands ; the 
quadrate, as stated in the definition, is single-headed, as in 
Hespcroriiis and many Struthiones. The brain, like that of 
Hesperoniis, is small, and the cerebellum is remarkably 
large as compared with the hemispheres. The teeth of Ichthy- 
ornis are implanted in distinct sockets. 

MARSH has remarked upon the close resemblance between 
the lower jaw, with its teeth, and that of the smaller mosa- 
sauroid reptiles. In Ichthyornis dis'par there are twenty- 
one distinct sockets in each ramus of the jaw. I. victor had 
the same number of teeth ; in I. anceps the teeth were more 
numerous, and at the same time more slender. The jaws 
were united, as in Hesperornis, by cartilage or ligament. 

The vertebra, as already mentioned, are amphiccelous ; 
but an approach to the typical saddle-shaped vertebrae is 
seen in some of them. The atlas is notched for the odontoid 
process of the axis. None of the dorsal vertebra? appear to 
have coalesced, and there is a pygostyle quite typical in form, 
but rather small. The shoulder girdle of both Ichthijornis 
and A'patornis is constructed upon the carinate plan. 1 There 
is the same angle between the scapula and the coracoid, and 
the clavicles are well developed. There are, however, differ- 

1 SHOFELDT (' Notes on the Extinct Bird Iclitliyornis,'' J. Anat. Phys. xxvii. 
p. 336) especially compares Iclitliyornis with Ehyncliops and Sterna. 


ences in detail. In Apatornis there is a very long acromial 
process. The coracoids overlap at their articulation with the 
sternum, more so in Iclitliijornis than in Apatornis. The 
clavicles are generally figured as typically carinate ; hut the 
only part of this bone known is ' a fragment from the upper 
end of that bone in Apatornis.' The sternum is deeply 

The bones of the fore limb are well developed ; the 
humerus has a very large crest, surpassing in comparative 
size that of any recent bird ; this clearly indicates a powerful 
flyer, and the rest of the bones of the limb bear out this view. 
Though nothing is known of the structure of the feathers, 
there are upon the ulna impressions for the quill feathers. 

In the pelvis all the bones are free posteriorly, as in Hes- 
perornis, Apteryx, &c. The acetabulum is perforate, as in 
most recent birds ; but the perforation is of moderate size, as 
in the tinamous. 

In the neighbourhood of Ichthyornis are possibly to be 
placed MARSH'S genera Baptornis, Telmatornis, and SEELEY'S 
Enaliorms. These birds, however, Cretaceous, like Hesper- 
or nix and Ichthyornis, are known by such limited material 
that their position is absolutely uncertain. 

The affinities of Ichthyornis to Hesperornis have been 
dwelt upon by some ; but it appears that LYDEKKEE'S 
remark, that ' the Odontornithes are a series of birds ances- 
tral to the modern series of toothless carinates,' expresses the 
truth. He has furthermore added that this series ' differs 
from the Euornithes (STEJNEGER'S name for carinates) by the 
absence of union between the rami of the mandible and 
between the distal ends of the ischium and ilium,' likenesses 
which do not mean a near relationship, but express the 
degree of development of bird structure at that period. 



Definition. Aquincubital. Oil gland present. Two carotids. Skull 
desmognathous and holorhinal. Caeca rudimentary or absent. 
Ambiens present. Biceps slip absent. 

This large group of birds admits of but a scanty defini- 
tion, if we are to include in it, as is here done, the secretary 
bird and the American vultures ; for it then shows a con- 
siderable amount of structural variation. The oil gland, 


t.p.l, tensor patagii longus ;, tensor brevis ; Anc, ancoujeus ; D, deltoid. 

invariably present, is generally feathered, but nude in the 
Cathartidse. In Serpentarius this gland varies in size ; in 
one specimen it was found to be very small and to have a 
very minute tuft. 

The after shaft is absent in the Cathartidse and in Pandion, 
present in other Accipitres. Twelve rectrices is the usual 


number, but fourteen occur in Neophron percnopterus and 
Rhinogryphus 'calif ornianus. 

There are powder-down patches in Eton us, Circus, and 
Gypaetus. The ptcryloxis is described for a variety of types 

The ventral tract broadens out on the breast, where 
it is even sometimes (Gyps fulvus) divided into an outer 
and inner branch. The dorsal tract forks upon the 
shoulders ; in Gi/paetus barbatus each limb of the fork is 
connected by a single row of feathers with the long single 
median posterior portion of the tract. 

In Periiis apivorus these latter slender forks are figured as 
being much longer, and in Falco peregrinus they dilate into 
four or five rows of feathers before uniting. In Falco 
brachypterus there is the usual dorsal fork, but between its 
extremities lies the beginning of the very broad posterior 
part of the tract. In all these birds there are lateral neck 
spaces. The lumbar tract is but little marked, or is entirely 
deficient. A large amount of detail is given in NITZSCH'S 
account of this family, which is treated more fully than many 

The tensor patag ii brevis is simple in all accipitrines, and 
there is never a biceps slip. There is, however, a certain 
amount of variation in the tendon. The simplest form of 
the tendon is seen in 

Vultur monachus 

Falco melanogenys 

Gyps fulvus ,, cesalon 

Tinnunculus alaudarius Thrasaetus harpyia 

Microhierax ccerulescens 

where it is a simple tendon without branches, as in many 
picarian birds. On the other hand in 

Buteo vulgaris tipilornis bacha 

Circus nut it r us 
,, Gouldi 
Helotarsus ecaudatus 

,, cheela 
Neophron percnopterus 

Nil Dago chima-ck i ma 


Milvago cli inut ngo Melierax polijzon us 

Gypactus barbatns Pohjborus brasiliensis 

DryotriorcJiis xpcctabilis Polyboroides typicus 

Aquila impcrialis Haliaetus albicilht 

Loplwae tics occipitalis Milvus ictinus 

Melierax monograminicus Astur approximans 

a wristward branch (as shown in fig. 226) is given off, the 
tendon, in fact, bifurcating. FUBBRINGER, however, while 
figuring the two types, distinguishes in the apparently single 
tendon of Tinnunculus two separate tendons in close contact 
for their entire length. 

A further complication is seen in Gijpaetus and Gypo- 
hierax, 1 where a small recurrent tendon (patagial fan) joins 
the anterior branch of the bre.vis with the long us, a state of 
affairs found to characterise Serpentarius, as will be pointed 
out later. 

The expansor secundariorum - is present in Milvago chi- 
mango, Harpijhaliaetus coronatus, Falco, Pohjborus, Tinnun- 
culus, and Microhierax coerulescens ; in others it is absent. 

The anconccus arises in Polyboroides and in some other 
types by a single head from the scapula, which is partly 
fleshy and partly tendinous. In Vultur aurieularis, on the 
other hand, the muscle arises by two completely tendinous 
heads, so that the muscle has not that value in the classi- 
fication of the Accipitres that I at one time thought. 3 In all 

1 In GiTiinaiii'titx melanoleucus, which has, as have all the last-mentioned 
genera, a bifurcate tendon of the brcvis, a small muscular belly ending in a 
tendon which becomes lost upon the patagium arose on the right side, in a 
specimen which I dissected, from the extensor muscle near to the end of the 
anterior branch of the brevis tendon. I am uncertain of the exact homology of 
this structure in Geranoaetns. 

- For muscular anatomy, &c., of Accipitres see GIEBEL, ' Bemerkungen iiber 
Cathartcs aura,' etc., Zeitschr. /. d. fjcs. Natww. ix. (1857), p. 420 ; ' Zur 
Anatomie von Yultiir fulvus? ibid. xxi. (1863), p. 131; 'Zur Anatomie des 
Lammergeiers,' ibid, xxviii. (1866), p. 149 ; S. HAUGHTON, ' On the Comparative 
Myology of certain Birds,' P. R. Irish Ac. ix. (1867), p. 524 (crane and goose 
as well as hawks). He deals with weight only. E. NEAXDEK, Undersukningar 
af Muxlt-iihititrcn hos sliigtet Buteo, Lund, 1875 ; MILNE-EDWABDS in Rccliercfas 
Anatomiques, d-c., des Oiscaux Fossilcs dc la France, Paris, 1867. 

'' ' On certain Points in the Anatomy of the Accipitres,' I'. Z. S. 1889,. 
p. 81. 


there appears to be an accessory tendinous origin from the 
humerus. The pectoral/* />rinix is commonly divisible 
into two layers, but not in Milvago and Dryotriorchis. 

The deep flexor tendons of the foot belong in the 
majority of species to type described above on p. 101. But 
there area few variations of the typical arrangement. In Astnr 
tibialis the slip to digit II. is present, but it is very small. 
In Baza, on the other hand, the vinculum is alone present, 
there being no special slip to the tendon supplying digit II. 

In Dryotriorchis spectabilis, Vidtur auricular is, and 
Milvus ictimis, the fibres of the vinculum are perfectly 
continuous with the slip to digit II., and form with it one 
single band of connection. 

Both peroneal muscles appear to be present in the 

All genera have the ambiens and the femorocaudal. In 
Falco and Circus maurus there is also a slender semi- 
tendinosus. 1 Glutceiis I. is generally absent, glutens V. 
commonly but not always present. 

The syrinx of the Accipitres is of the ordinary tracheo- 
bronchial form. 

In Falco peregrinus the intrinsic muscles are inserted on 
to a transversely elongate fibro-cartilaginous bar which runs 
across the interannular membrane of bronchial semi-rings 
1 and 2. This membrane is very wide, owing to the fact 
that the first bronchial semi-ring is much arched, the con- 
cavity being downwards, while the second semi-ring is equally 
arched, but the concavity is upwards. None of the tracheal 
rings are fused, and the last gives rise to a pessulus. F. 
caiidicans,F. lanarius, F.biarmicus, F. Feldeggi, F.asalon, 
F. sacer are perfectly similar, and the bronchidesmus (in those 
specimens in which it had been preserved) is complete. 

Much like the syrinx of Falco is that of Hieracidea 
berigora ; I can, indeed, detect no differences. So too 
Tinnunculiis alaudarius and Erythropus vespertinus. In 

1 FORBES in a MS. note records what I call ' semitendinosus' in 
ccerulesccns as a product of the division of the semimembranosus on account 
of its origin from ischium and pubis. 


Milvago cliimanr/o and M. cliima-cliima the syrinx is at first 
sight perfectly similar, but the intrinsic muscles only just 
get beyond the first bronchial semi-ring. In Hcrjjctotheres 
cachinnans this divergence from the normal falconine 
syrinx is carried still further, the intrinsic muscle being 
attached to the first semi-ring. 

The syrinx of Polybonis brasiliensis is an exaggeration of 
the falconine type. The first and second bronchial semi-rings 
are very prominent and wide apart, thus leaving a very 
spacious interannular membrane, to which the intrinsic 
muscles are attached. The last few tracheal rings are fused 
mesially in front and behind. The remaining forms, so far 
as I have studied them (comprising the genera Melierax, 
Nisaetus, Gypaetus, Thrasaetus, Buteo, Milvus, Spizaetux, 
Urubitinga, Haliaetus, Vnltur, Spilornis, Morphnus, Helo- 
tarsus, Leucopternis, Circus, Aquila, Circaetus, Gyps, Archi- 
buteo, Geranoaetus, and Asturina), differ from each other 
in details such as the completeness or incompleteness of 
the bronchidesmus, the degree of ossification of the rings 
and semi-rings, the number of the last tracheal rings which 
are fused, and the attachment of the intrinsic muscles 
(semi-rings 1, 2, or 3) but they agree to differ from the 
falcons in the absence of a pronounced oval gap between 
the first and second bronchial semi-rings, which gives to 
the syrinx of the falcons so characteristic an appearance. 

The lobes of the liver are subequal, and a gall bladder 
is present. The cseca of the Falconidse are minute. 

Haliaetus albicilla is a fish-eating bird, and for some 
reason birds with such habits are furnished with a long in- 
testine, as will be seen from the measurements in the table 
on p. 477. The duodenal loop in this bird, exceptionally, 
is thrown into a series of subsidiary loops, a state of affairs 
which, as it occurs in the remote penguin, may have some 
relation to habits and may not be a character upon which 
stress is to be laid. The greater part of the intestine pre- 
serves the simple archaic form of a number of irregular 
coils ; but near to the caeca are two spirally twisted, elongated 
loops. In other Accipitres it is more usual for the upper 



loops to be long and twisted, a circumstance which recalls 

the structure of the loops in the stork (see fig. 208, p. 437). 

The following are a few intestinal measurements : 

Ft. In. 



Scrpi'iituriufi n'{ifilironis 3 

6 6 



)> it 6 

7 9 




C 8-5 




7 (i 



(rii/iagiis papa $ 



)j ji 



Catliartca atmtits 



Poli/borus brasilicnsis $ 

5 3 



Spizaetus coronatus Q 




,, caligatus $ 

2 (3 



Fulco biannicus 1 ^ 

2 3 



M ileus ictinus 9 

3 11 



Circus Gouldi 9 

4 7-75 



ccruginosus . 

3 8-5 



Haliactus albicillu , 




vocifer $ . 

8 7 



Aqniln n/Tvioides 9 

4 4 



The sJi'itll of the Falconidse is described and figured by 
HUXLEY, J PARKER, 2 and SHUFELDT. S The palate is described 
as desmognathous ; but it is always the case that a large' 
portion of the maxillo-palatines the posterior region are 
not in contact. In two skulls of Lophoaetus occipitalis the 
palatal surfaces of the bones were nowhere in contact, and 
were only in contact for a minute space in a skull of Vultur 
calms. Neither is Elanus desmognathous, according to 
SHUFELDT. The maxillo-palatines are large and swollen. 
The vonier is long and knifeblade-shaped ; 4 there is often a 
medio-palatine, for instance in Haliaetus albicilla, where it is 
embraced by the bifurcate posterior extremity of the vomer. 
The lacrymal is large and has a separate ossification, the so- 
called infraorbital, attached to its posterior extremity in many 

1 In P. Z. S. 1867. - Linn. Trans. (2) i. 

3 ' Some Comparative Osteological Notes on the N. American Kites,' Ibis, 
1891, p. 228: 'Osteology of Circus hudsonianus,' J. Comp. Mcd. 1889. 

1 It has been found to be bifid in front, after the charadriiform plan, in youn<>- 
of Tinnunculus (cf. SUSCHKIN, ' Zur Anat. u. Entwicklungsgesch. d. Schaclel d. 
Eaubvogeln,' Anat. Anz. xi. p. 767). 


hawks, for example in HaUactus alb i cilia, LopJioaetus occipi- 
talis, Circus Goitldi, Asturina Natteri, Astur Nova Hol- 
Idinliit", in others, such as Herpetotheres cacJtinnans, Vultiir 
calvux, the large size of the lacrymal suggests that such a 
bone is present, but ankylosed with the lacrymal. The bony 
nostrils of the Falconidae are holorhinal, sometimes (e.g. 
Herpetotheres each in /nuts) reduced in extent by alinasal 
ossifications ; the long septum between them is more or less 
perfect. The number of cervical vertebrae, ribs, hsemapophyses, 
and uncinate processes of a few types is shown in the fol- 
lowing table : 




T 7 nninate 




HerjH-fi if lieres cacliiunaus 


(s) r + r' + 6 


On 3-7 

Dri/utriorchis spectabilis 


(>S) r + r'+5 + r 



Circus Gonldi . 


(9) v + v'+7 



Asturina Natteri 


(9) r + v' + 6 + r 



Airl [>iter nisiis . 


(10) r + v' + 7 + v' 



Astur Novic HollandicB 


(9) r + 8 



LoplicxH'tiifi occi/'ifn/ix 


(9) v + V + 6 + r 



Melicra.i- monogrammicus 


(9) v + r' + 6 + v 



The sternum is whole or with one pair of foramina, some- 
times notches, and often only developed on one side. The 
coracoids slightly overlap in Dryotriorchis, Herpetotheres, 
and Melierax ; they do not quite meet in Accipiter, Lopho- 

Pandion is undoubtedly an aberrant genus, which is by 
several (c.r/. GADOW) made the type of a separate family, and 
is thought by some to lead towards the owls. It differs from 
other falcons in having no aftershaft, in its somewhat 
peculiar tensores patagii and deep plantar tendons. 

The tensor patagii, brevis has the additional ' aquiline ' 
wrist ward slip, from the middle of which rises a short 
recurrent slip which joins the insertion of the main tendon. 
The tendon of the biceps muscle is split for nearly its whole 

The deep plantar tendons are not accipitrine ; they blend 
completely, as in owls, hornbills, &c., the area of fusion being 


The syrinx is not remarkable in form. Anteriorly the 
last three tracheals are fused medianly ; posteriorly the fusion 
is more extensive, and includes the first bronchial semi-ring. 
The second bronchial semi-ring is in front close to the first ; 
behind it is united with the third, upon which latter are 
inserted the intrinsic muscles. 

The skull is accipitrine and not strigine. The descending 
process of the lacrymals, however, is firmly and entirely 
blended with the ectethmoid, but the former bone has no 
backwardly projecting frontal portion, let alone a separate 
ossification at the end of it, such as is met with in some 
Accipitres. The vomer is long and ends in front in an olive- 
shaped swelling which fits in between, but is not attached 
to the diverging limbs of the anteriorly fused maxillo- 

The ring of the atlas is incomplete in the middle line above ; 
there are fifteen cervical vertebne. The haemapophyses are 
very feeble on the earlier cervical vertebrae ; they commence 
on CIO, where they are double ; they are strong over the last 
cervical and the first three dorsals, where they end. Six 
ribs reach the sternum, of which the first four have uncinate 
processes. Both the tibio-tarsus and the tarso-metatarsus 
have a bony bridge for tendons ; the latter has one behind 
as well as in front. 

This bird possesses a scapula accessoria in the glenoid 
capsule, the significance of which as a point of affinity with 
the owls is marred by its occurrence in toucans, &c. (seep. 192). 
The coracoids slightly overlap, as in some Accipitres. 

Whatever may be thought about Pandion, it is clear 
that the separation of the secretary bird to form a distinct 
family, Serpentariidse, is perfectly justifiable. 1 

Serpentarius has basipterygoid processes, and its muscle 
formula is BXY + . 

The tensor patag ii l> rev is is more stork or crane like than 
accipitrine, and indeed resembles Cathartes in the presence of 

1 The claims of Poltjboroides to be a member of this family have been dis- 
missed by MiLNE-Ei>\vABDS (Hist. Xnt. Madagascar) and myself (loc. cit. on 
p. 474). 



a slip (see fig. 227) uniting the brevis and longus tendon. It 
must be remembered, however, that this also exists in some 

In Serpentarius there is a longer attachment of the 
deltoid to the humerus than in other birds of prey ; and there 
is an accessory biceps muscle (see fig. 227). : 



t.p.l, tensor longus ;, tensor brevis ; Bi, biceps ; Hi', accessory bicc-i's : 
Anc, anconseus ; />, deltoid. 

The anconceus has a very broad tendon of origin from 

In the syrinx a strong box is formed by the last tracheal 
ring, and the intrinsic muscles are attached to bronchial 
semi-ring 2. 

There is a powerful expansor secundariontui. 

1 Cf. Rhinochctus, p. 371. 




The skull has strong basipterygoid processes. The 
lacrymals are large and extend backwards in close connec- 
tion with the skull wall ; they are not ankylosed to it. The 
descending process is thin and articulates with the slight 
ectethmoid. There is a small knife-shaped vomer. 

The family Cathartidae consists of the genera Sarco- 
rluimplius (condor), Gyparchus (or Gypagus, king vul- 
ture), Cathartes (turkey 
vulture), and Eliinognj- 
plius. They all have, so 
far as is known, the oil 
gland nude, twelve rec- 
trices, no after shaft, and 
are aquincubital. 

The tongue is large 
and fleshy, with denticul- 
ations of its upturned 
lateral margins. 

The stomach is not 
a gizzard. There are no 
intestinal cceca. The in- 
testines are 61-inch in 
Gyparchus, 49-inch in 
Cathartes atratus. Of 
the heart of the condor 
some observations will be 
found above (p. 50) ; both 
carotids are present. The 
liver is equilobed, with a 
gall bladder. There are in 
Gyparchus traces of a crop. 

The most distinctive feature of the Cathartidse, however, 
is the windpipe, from which a proper syrinx may be really 
said to be absent. The only muscles upon the trachea are 
the sterno-tracheales, which (in G. papa) are very short and 
broad, and arise from the sternum in the middle line, close 
together between the inner ends of the coracoids. Intrinsic 
syringeal muscles are entirely absent in the Cathartida?, 

i i 

FIG. 228. SKULL ov Serpentarius 

1'jciii, premaxilla ; J/.r/i, inaxillo-palatim s ; I'l. 
palatine ; Pt, pterygoid ; x, basipterygoid process. 


unless, indeed, their homologues exist l in the form of a 
muscular covering to the terminal purely membranous 
section of the bronchus (fig. 229) which, dividing into three 
slips, runs from thence to the parietes. There are in any 
case no muscles at the actual bifurcation. Nor is there any 
change in the character of the rings themselves such as to 
suggest even the rudiment of a syrinx. In Cathartes 
the rings at the bifurcation are extraordinarily thin, leav- 
ing wide membranous intervals, which are occasionally 

O * 

traversed by bridges putting successive rings into communi- 
cation. In Sarcorhamphus and GyparcJms, which also agree 
(see below) in their 'muscle formula,' the rings are thicker 
and closer together (see fig. 229). And in these two genera 
the bronchi are incomplete internally, giving rise to what 
may be termed a membrana tympaniforrnis. 

In GyparcJms papa the tendons of the patagium are 
somewhat complicated. The brevis consists of a separate 
anterior and posterior section, of which the latter is thinner 
and more diffuse. The anterior tendon divides into two, 
of which the foremost gives off a slip to the longus. There 
is no biceps slip. The tendons, in fact, are thoroughly 
stork-like, as are those of the condor (Sarcorhamphus) 
and Cathartes. In this character the family is very uni- 

The expansor secundariorum is present in all. 

The pectoralis primus is well divided into two parts, of 
which the lower (in G. pa23a) is inserted by a thin round 
tendon altogether below insertion of superficial layer. The 
head of the anconceus is distinctly bifid and entirely tendinous, 
arising from scapula and from supinator muscle. There is 
a humeral slip of moderate size. (This muscle is described 
from Cathartes.) 

The anibiens is present in all Cathartidae ; so too the 
semitendinosus and its accessory. Cathartes has in addition 
the femorocaudal, which is absent in the other genera. In 
GyparcJms ( ? as to the others) the semitendinosus and semi- 

1 BEDDAKD, ' Notes on the Anatomy of the Condor,' P. Z. S. 1890, p. 146. 



membranosus are inserted in common. There are two 
peroneah (at any rate in Cathartes). The deep flexors are 
fused before origin of four slips to four toes. 

The cjlutccus I. covers over biceps, and gluttmis V. is 

The sTiidl of the Cathartidae has basipterygoid processes. 
It is desmognathous, but the desmognathism is totally 



FIG. 230. SKULL OF Cathartcs 

Tr, trachea ; ce, oesophagus ; 0, ostia of lungs ; s, sp, Pmx, premaxilla : ^f.^p, ruaxillo-pala- 
septa between air sacs ; in, muscles ensheathing end tines ; PI, palatines ; Pf, pterygoid ; 
of bronchi. ,^ ossified septum ; .;, basipterygoid 


different from that of other vultures and hawks. The 
maxillo-palatines proper (see fig. 230) are very far indeed 
from meeting in the middle line ; indeed, they only just get 
beyond the shelter of the palatines. But a flat dorsal pro- 
cess of each of these bones (S) meets and is co-ossified with 

i i 2 


the nasal septum in the middle line. In Gyparclms papa, at 
any rate, there is a small medio-palatine. The lacrymal (at 
least in Cathartes atratus) is a smallish bone completely 
filling a notch in the frontal margin ; its descending process 
ankyloses with the ectethmoid, forming the usual ring. In 
Gyparchus papa the orbital portion of the lacrymal is greatly 
reduced ; the nostrils are not so elongated as in Cathartes ; 
the palatal bridge is more plainly an alinasal fold. The 
bony nostrils are holorhinal, but much more elongated than 
in the Falconidae ; there is no trace of an ossified internarial 

Of fossil Accipitres the remains of a number of different species 
have been found. The most interesting of these, on account of its 
age, is the Litliornis vultiirinus of OwEN, 1 from the London clay. 
It had been beld to come nearest to Cathartes, an interesting fact 
in view of its occurrence in this country ; but LYDEKKER regards 
it as clearly accipitrine and allied to Accipiter and Circus. Har- 
pagornis,- from the Pleistocene of New Zealand, was a large bird, 
one and a half time the bulk of a golden eagle, also belonging to 
the same division of Accipitres. Teracus and Palceohierax are 
extinct genera from the lower Miocene of France, known only by 
femur and tarso-metatarsus respectively. They are also probably 
true falcons. Serpentarius is known by an extinct form, S. robustus, 
from the lower Miocene of the same country. 

The following table shows the main differences between 
the several families of the Accipitres : 




Oil gland 
Muscle form a In 
Accessory scniimcmb. 
Cceca .... 
Syrinx .... 
Basijit. j->rc. . 

+ (exc.Panclion) 
A + 
+ (rud.) 

Of maxillo-pal. 


BXY + 

+ (rud.) 

Of max.-pal. 


(A)XY + 

Of alinasals 

1 'Description of the Fossil Eemains of ... a Bird (Litliornis vultiirinus) 
from the London Clay,' Trans. Geol. Soc. (2), vi. 1841, p. 206. 

2 Cf. HAAST in Trans. N. Zealand Inst. iv. 1871, p. 192, and ibid. vi. 1874, 
p. 64, and OWEN in Extinct Birds of New Zealand. 


It is clear from the few characters the principal ones, 
however given in the above list that the Cathartidae are 
more aberrant (considering the Falconidee to be the typical 
birds of prey) than are the Serpentariidaa ; for the 
Cathartidse diverge in all eight characters from the Fal- 
conidae, while the secretary vulture only diverges in three. 
What reason is there, it might be asked, to retain the 
American vultures within this order at all, particularly 
if the owls are to be as I think they should excluded ? 
The only group which has the distinctive characters of the 
Cathartidae (besides, of course, the present group) is that of 
Herodiones. There only do we find birds with ambiens and 
expansor secundariorum, without biceps slip, holorhinal, and 
with rudimentary or absent cseca. The Steganopodes also 
are not far off. It really conies to the beak and claws, the 
ceroma, and to the presence of various structures (e.g. the 
peculiar palate, the basipterygoid processes) which forbid 
their association with the Herodiones. The several groups 
are not far off, but on the whole the American vultures are 
more like the remaining birds of prey than like the stork 
tribe (see also under the discussion of the affinities of the 
Grues, p. 882). 


Definition. Oil gland tufted. Q,uintocubital. Muscle formula of 
thigh, ABXY"+. Expansor secundariorum present. Biceps 
slip absent. Both carotids present. Large caeca and crop. Skull 
dromseognathous. Tail short without ploughshare bones. Bones 
of pelvis free distally. 

The tinamous are purely South American birds, of which 
in his recent catalogue Count SALVADOEI allows nine 

The tinamous have a tufted oil gland, but the tuft is often 
very minute, and in Calodromas elegans consists of only four 
feathers, two larger and two smaller, the larger ones being 

I take my account of the pterylosis of the tinamous from 


PYECRAFT'S careful description 1 of Calodromas (which I can 
confirm) and Rhynchotus rufcscens. In the former bird the 
body is fairly covered with feathers, the apteria being narrow. 
There is no down save on the wings. The spinal tract soon 
divides into two ; but they rejoin near the base of the neck. 
These tracts again divide and reunite some way in front of 
the oil gland, enclosing thus a dorsal apterion. The ventral 
tracts also divide early upon the neck, and each of them 
again divides on the pectoral region into a stronger, outer, 
and a somewhat weaker, inner, tract. Until about halfway 
down the neck the dorsal and ventral tracts are in contact. 
In Rhyndiotiis rufescens there is no spinal apterion. The 
after shaft is much more rudimentary than in Calodromas, 
where it is well developed. Both birds have ten feebly 
developed rectrices. Rh. perdicarius has eight. 

The aftershaft is apparently in the process of disappear- 
ance among the tinamous. In Nothocercus, writes Mr. PYE- 
CRAFT, ' it is evidently degenerating, inasmuch as the shaft 
is almost, if not quite, obsolete, only the rami remaining.' 
In Tinamus solitarius the aftershaft is absent. Powder- 
down pat dies exist in a few tinamous. They occur, for ex- 
ample, in Tinamus major. In Cryptu rus ta ta upa the powder- 
down patches extend down 011 each side of dorsal tract from 
a little in front of humerus nearly to oil gland. After the 
end of the scapula they thicken and spread outwards as far 
as the head of the femur, and are in contact for nearly two 
inches along mid-line ; they then narrow again and terminate 
half an inch in front, and slightly to the side, of the oil 

Rhynchotus perdicarius has apparently no powder- 

The tongue of the tinamous is small and triangular in 
form. The crop is present and large. The provcntricitlus 
is zonary ; the liver subequilobed, with a gall bladder. 

The following are measurements of the alimentary 
canal : 

1 Ibis, 1895, p. 1. 



S. I. 

L. I. 


Crypt urus tataupa 




,, obsoletus 
sallai . 



| 2-5 



RJiyiicliotiis rufcscens 
Nothura maculosa . 


i ' \ 
1 3 

8-5, 9-5 

Tinamus solitarius 



The ccBca, it will be observed (fig. 231, p. 488), are well 
developed, particularly in Rhynchotus rufescens ; they are 
also large and very peculiar in form in Calodromas. 1 The caeca 
of this bird are not merely much wider than is customary, 
but they are beset with numerous small diverticula, which 
diminish in size towards the apex of the caecum. These 
peculiar cseca are absolutely unique among birds, and nothing 
at all like them has been described in any other tinamou. 

A curious feature of at any rate some tinamous (shared, 
however, by the Anseres, PalamedeidEe, some gallinaceous 
birds, and perhaps Toccus) is the existence of two pairs of 
extrinsic muscles upon the trachea. In Cnjptiinis tataupa 
one of these pairs is stouter than the other, and they are both 
lost on the fascia covering lungs. This genus has no intrinsic 

In other tinamous intrinsic muscles are present. 

In Calodromas elegans the anterior face of the lower part 
of the trachea (about an inch in length) is covered with a 
sheet of muscle, which is the extrinsic muscle, and probably 
(judging from the conditions which obtain in the female) is 
attached to the long fascia. The very broad intrinsic mus- 
cles underlie this, and are inserted a long way down the 
bronchus to the four or five rings following the third. When 
viewed laterally the curvature of the last four tracheal rings 
is seen to gradually increase ; there is thus a considerable 
membranous interval left between the last tracheal and the 
first bronchial, which is straight. The membrana tympani- 
formis is narrow. In the hen bird the extrinsic muscles are 

1 BEDDAKD, ' On the Cseca of Calodromas,'' Ibis, 1890, p. 61. 


small, and do not form a sheet of muscle covering the end of 
the trachea. 


FIG. 231. C^ECA OF Calodromas elcgans (AFTER 

S.I, small intestine ; L.I, large intestine ; C, cseca. 

FIG. 232. C^CA OF No- 
tliura maculosa (AFTER 
IN FIG. 231. 


In Tinamus solitarius the intrinsic muscles are large and 
are inserted upon the fifth or sixth bronchial semi-ring. In 
Mhynchotus rufescens they are also large, but attached 
higher up. 

The muscles * of the tinamous .are remarkable on account 
of their soft texture and pale colour. The pectoralis I. is 
very large and meets its fellow of the opposite side and for 
a considerable portion of its extent. The actual junction of 
the fibres is prevented by a fibrous septum, which is a con- 
tinuation of the carina sterni. The second pectoral is also 
large. The tensor patagii tendon is abroad diffuse band, as 
in gallinaceous birds ; it has no biceps slip unless a muscle 
that will be referred to in connection with the biceps 
immediately really represents this. The biceps itself pre- 
sents no remarkable features, but a kind of accessory biceps 
runs along the front of the humerus, which is quite distinct 
from the biceps proper ; I have found this both in Calo- 
dromas and Bhynchotus. This is regarded by FURBRINGEE 
as part of the coracobrachialis externus. 

In Cr upturns (as also in gallinaceous birds and sand 
grouse a noteworthy fact, perhaps) the pectoralis abdomi- 
nalis has a remarkable ending. 

Instead of being inserted directly upon the humeral crest 
it ends upon a tendinous bridge, to which are also attached the 
pectoralis, the latissimus dorsi posterior, and the expansor 

The tinamous have the gallinaceous muscle in the fore 
arm and the expansor secundariorum. The anconceus has a 
humeral slip. 

In Cryptnrus tataupa and Notliura maculosa&i any rate 
the expansor secundariorum ends in a tendon which is 
inserted on to the scapula on the one hand and the manu- 
brium sterni on the other. 

The tinamous have the complete muscular formula of 
the leg, i.e. ABXY+. 

1 ALIX, ' Sur la Myologie du Rliyncliotus ntfescens,' Juitru. de Zool. v. (1870), 
p. 411. ' Memoire sur I'Ostiologie et Myologie du Notliura major,' 1 Journ. 
Zool. iii. 1874, pp. 167, 252. 





The glut cei I.-V. are well developed. A very interesting 
feature of the thigh muscles (referred to by GARROD ! ) is the 
existence of a small ' suprasciatic ' slip of muscle arising 
behind the acetabulum, which reinforces the accessory 
femorocaudal. The interest of this small muscle lies in the 
fact that it has its precise counterpart in the struthious birds 
(q.v.) This muscle was found by FORBES to be absent in a 

male Crypt tints tattnqja; it was present 
in a female of the same species. 

The two deep flexor tendons fuse and 
then supply digits II. -IV. ; before uniting 
the flexor hallucis gives off a slender slip 
to hallux, which is wanting in Crypt it nix 

The skull of the tinamous, as was 
first pointed out by PARKER, 2 is com- 
pletely ' struthious ' so far as concerns the 
palate. As will be seen from the annexed 
cut (fig. 233) the vomer is broad and unites 
in front with the maxillo-palatines, as in 
Dromceus. Its ends receive behind the 
pterygoid and palatines, which are thus 
prevented from articulation with the 
basisphenoidal rostrum. There are large 
basipterygoid processes and the head of 

FIG. '233. SKULL OF J . . , ,, . 

Tinamus robustus the quadrate is single, as in struthious 
(AFTER HUXLEY). birds. The supraorbital chain of bones 
figured by PARKER in Tinamus robustus 
palatiues ; is another archaic skull character of these 
birds. 3 The nasals, lacrymals, and 
adjoining bones are very much like those of Eliea and not at 
all like those of gallinaceous birds. Between the nasals 
posteriorly is a considerable tract of ethmoid, 4 appearing 

1 ' On certain Muscles of the Thigh of Birds,' etc., P. Z. S. 1873, p. 642. 

- ' On the Osteology of Gallinaceous Birds and Tinamous,' Zool. Trans, v. 

:! Absent, according to LUCAS (' Notes on the Osteology of the Spotted 
Tinamou,' Proc. U. S. Nat. Mus. x. 1887, p. 157), in Nothum inaculosa. 

* Prof. PAEKER wrote, in 1862 (loc. cit. p. 213) : ' I suppose that in the tina- 
mou, as in other ostriches, the broad top of the ethmoid is separately developed 

T1NAMI 491 

upon the surface of the skull. The outer descending part of 
the nasal reaches the maxilla, and with the upper part of the 
bone encircles the holorhinal nasal foramen ; it is not 
ankylosed with the lacrymal. The latter descends and 
articulates with the jugal by a very distinct facet, especially 
distinct in PJujnclwtux ntfrscens. The bone also becomes 
fused with the lateral wing-like process of the ethmoid, 
forming a complete ring of bone round a relatively very 
wide foramen. A special point of resemblance to Ehea and 
Droin&us is the perforation of the descending process of the 
lacrymal itself. This is best seen in Eliijnchotux rufescens ; 
in Calodromas elegans, Notliura maculosa, and Grypturus 
tataupa, there is merely a notch which in the fresh skull may 
possibly be converted into a foramen by a ligament. 1 

FURBRINGEE gives 16-18 as the number of cervical 
vertebra. The cup of the atlas is perforated for the odontoid 
process in Crypturus, Rhynchotus, and Nothura. As in 
the gallinaceous birds, &c., some of the dorsal vertebrae are 
fused together. In Tinamus solitariusthisw&s the case with 
the first three and to a less extent with the last cervical. In 
Nothura maculosa five vertebrae were thus fused, and a 
strongish longitudinal piece of bone, formed of ossified liga- 
ment, connected their transverse processes. In Cr upturns 
tataupa there were four vertebrae fused and one in front 
partially so. Four ribs reach the sternum in Tinamus 
solitarius. The sternum of the tinamous is very remarkable 
in form. The manubrium is slightly bifurcate ; the middle 
portion of the sternum, which bears the keel, is exceedingly 
narrow, and a wide space is left on each side between it 
and the lateral processes, which are thin and as long as 
the middle piece. The anterolateral processes are well 

The pelvis is so far on the struthious pattern that the 

by a long piece growing from above downwards between the anterior ends 
of the frontals. No suture remains to tell me that ; but if it be so all is 
perfectly struthious, for those birds differ in this from all others examined 
by me.' 

1 This, however, is not a unique feature of the birds in question. It occurs, 
for example, in Tantalus and XenorTvynchus. 


three bones do not fuse posteriorly ; the pectineal process is 

There are at most faint traces of a ploughshare bone. 

In Cryptunis the clavicles come into contact with the 
acrocoracoid and the scapula, but not with the small pro- 

The two coracoids are not nearly in contact at their 
articulation with the sternum. 

The only birds with which the tinamous have been com- 
pared are the ostrich tribe, the gallinaceous birds, rails, 
bustards, and some of the Limicolse. PAEKEE saw in the 
tinamou a ' cock ostrich mule ; ' and perhaps the prevalent 
opinion is that they lie on the confines of these tw T o groups- 
It is unquestionably to the Struthiones that they show the 
greatest numbsr of important likenesses, 1 so much so, indeed, 
that their inclusion in one great group with them would be 
by no means an unreasonable way of disposing of them. 
The salient points of resemblance are by no means confined 
to the skeleton, but the most numerous resemblances are in 
that part of the body of the birds. The skull, with its 
' dromseognathous ' palate, is strikingly like. The appearance 
of the ethmoid as a median ossification of the skull roof is 
struthious, but it also occurs, though not so markedly, in 
Galhts and (according to SELENKA) in the CaprimulgidEe. It 
is very conspicuous in the tinamou and the Struthiones. The 
open pelvis is especially like that of Aptenjx. The single- 
headed quadrate is struthious ; but, as already mentioned, 
the struthious birds are not uniform in this character. 

The sternum, with its antero-lateral and postero-lateral 
processes, recalls that of Aptenjx, in spite of the enormous 
length of these parts and a consequent superficial dissimi- 
larity. The absent, or rudimentary, ploughshare bone may 
perhaps be passed over as correlated with the imperfect 
flight. As to the soft parts, the peculiar additional accessory 

1 A singular if less important likeness than some mentioned above has been 
referred to by Mr. BARTLETT (' Notes on the Breeding of several Species of Birds,' 
&c., P. Z. S. 1868, p. 114). The male Rliyncliotus rnfescens incubates, as in 
the Struthiones, while the chick ' much resembles the young of a Rhea.'' 



femorocaudal muscle is a striking resemblance to the Stru- 
thiones ; the peculiar accessory biceps muscle of the arm 
may have its degenerate counterpart in a sheet of strong 
tendinous tissue which runs along the humerus in certain 

The following table shows some of the more striking 
likenesses of the Tinami to the Struthiones and Galli : 






Bones free 


Bones free or 

Ischia and ilia 

Oil gland . 


but little united 

fused posteriorly 
Tufted, or nude, 

Rein ex V. . 


+ ? + 

Leg muscles 
Swprasciatic muscle . 

Trachea . 


Sometimes with 

(A)BXY( + ) 
In Apteryx 
Only one pair 

(A)BXY + 

Sometimes two 

two pairs ex- 
trins. muscles 




Definition. Flightless birds without stiff contour feathers. Oil gland 
absent. "Wing small. Expansor secundariorum and biceps slip 
absent. Semitendinosus and its accessory always present. An 
additional slip to accessory femorocaudal present. Skull 
dromaeognathous with basipterygoid processes ; holorhinal. 
Sternum without a well-developed carina. Coraco-scapular 
angle wide. Coracoid fused with scapula. Casca large. 

As will be seen from the above definition, the characters 
of this group are to a considerable extent negative characters. 
They are for the most part such characters as are correlated 
with the loss of the power of flight. AVe need not, therefore, 
lay too much stress upon them as indicative of the natural- 
ness of the group. But even when these characters (as, for 
instance, the absence of the carina sterni, the open angle 
between the coracoid and scapula, the absence of a plough- 
share bone, which, moreover, is occasionally and exception- 
ally present. There is a skeleton of an old Stnithio in the 
Cambridge Museum in which several of the last vertebrae 


are fused) are set aside as comparatively valueless as marks of 
near relationship, there remain enough anatomical resem- 
blances to justify the older view that all these birds are but 
members of one and the same group. FUEBKINGEK has 
denied this in his ' Untersuchungen,' and places Apteryx, 
together with the Dinornithidse, apart from the other stru- 
thious birds, and has again separated Struthiiformes from 
Eheiformes and Casuariiformes, deriving all from different 
levels of the ornithic tree. There is no doubt that the 
various types of struthious birds do require separating into 
at least six families ; but the likenesses among them appear 
to me to forbid any wider separation. The close resemblance 
of the palate throughout the group, so far as we know it 
(JEpyornis is not known), is a strong reason for associating 
them together ; perhaps even the osteological and other 
characters, which, as already suggested, are but evidence of 
the loss of the flight power, may be of more importance as 
an argument for affinity than is generally admitted ; it may 
show that they are allied, because the degeneration has pro- 
ceeded along the same lines. There is, it is true, not a great 
deal of evidence in favour of this view ; but we have the 
penguins also with a degenerate wing, in which the modifi- 
cations of structure ' have progressed along different paths. 
They have, for example, lost the biceps, which is present in 
all Struthiones, while the feathers of the wing are equally 
inefficient as aids to flight with those of the Struthiones, but 
are quite unlike them. The peculiar muscle of the thigh, 
which will be found described as an adjunct of the accessory 
femorocaudal, is one of those apparently small facts of struc- 
ture which, on account of their very minuteness, seem of 
importance as a mark of true relationship. 

The fact that all of the struthioris birds have large or 
moderately developed caeca is further evidence of affinity. 
It might be thought that the usual absence of the oil gland 
was one of those characters affording clear evidence of degene- 
ration ; but its capricious appearance and disappearance in 

1 See, however, the qualifying remarks with regard to the wings of Apteryx 
on p. 499. 


other birds forbid us to assume this without any further- 

While the Struthiones present collectively and individually 
a larger number of important differences from other birds, 
their organisation is essentially on the plan of that of the 
remaining members of the class. To take only one at the 
same time one of the most striking of these correspondences 
in anatomical structure, the respiratory organs may be con- 
sidered. It is hardly too much to say that there are not even 
differences of detail in the arrangement of the lungs and air 
sacs among the struthious birds. Professor HUXLEY ex- 
ploded some years ago the idea that the oblique septa of the 
Apteryx were more like the mammalian diaphragm than the 
corresponding structures of other carinate or ratite birds. It 
is inconceivable that there should be this minute correspond- 
ence of detail with detail, if we are to assume with some 
that the struthious birds have arisen from a totally different 
stock from that which produced the carinates. They would 
derive the former from the dinosaurs and the latter from the 

The existing struthious birds are the genera Strutliio, 
Afro-Arabian in range; Rhea, South American ; Droma'iis, 
Australia; Apteryx, New Zealand ; and Casuarius, Australian 
region. The structure of these living members of the group 
will be considered first, after which some account will be 
given of the Dinornithidae and other extinct and undoubted 
members of the group, as well as of a few dubious forms 
which have been placed here rather because they do not 
definitely fit in anywhere in particular than from their 
obvious affinities with the Struthiones. 

The genus Strutliio appears to contain two species, the 
more common Struthio camelus and the Somaliland S. 
molijbdophanes. The ostrich has two toes, Nos. III. and 
IV. There is no oil gland. The pterylosis, continuous in 
the adult bird, show's two distinct apteria in the embryo, as 
has been shown by Miss LINDSAY.' In the young chick 

1 ' On the Avian Sternum,' P. Z. S. 1885, p. 684. See also W. MARSHALL, 
' Beobachtungen iiber das Verhiiltniss der Federn,' &c., Zool. Gart. xvi. (1875), 


there is a ventral apterion in the sternal region and a lateral 
apterion outside each half of the ventral tract. The adult 
ostrich has a claw on each of digits I. and II. The arrange- 
ment of the wing feathers has been carefully worked out by the 
late Mr. WE AY. 1 He finds the remiges to be quite distinct, 
as well as the tectrices majores ; the tectrices mediae are but 
scantily represented, and there is an incomplete row of 
tectrices minores. The number of remiges upon the hand, 
including one upon the carpus, is sixteen. There are four 
to the ala spuria. The number of cubitals is about twenty. 
It has, therefore, more primaries than any bird except the 

The genus Aptenjx, entirely confined to New Zealand, 
consists of three or four species, viz. A. australis, A. Mantelli, 
A. Oweni, A. Haasti, and A.Bulleri. 

It has been described as possessing a continuous, un- 
interrupted plumage ; but this, according to T. J. PAEKEE,' 2 
is far from the truth. ' In a fresh specimen of A. Bitlleri,' 
he remarks, ' I find the lateral apterium to be fully 2 
cm. wide, and to extend about 5 cm. cephalad and 9 cm. 
caudad from the axilla, its total length being, therefore, 
about 14 cm. In the same specimen the ventral or inferior 
space was of about equal width (2 cm.), and extended about 
11 or 12 cm. caudad from between the origins of the wings. 
Moreover the inner (ventral) surface of the wing is always 
nearly devoid of feathers and so constitutes a well-marked 
lower wing-space.' 

The oil gland is present and the feathers have no after- 

The relatively minute wing of the Apteryx has a true 
alar membrane, which, as PAEKEE has justly pointed out, is 
further evidence for regarding this bird as the derivative of 
a flying form. 

p. 121, and ZANDER, ' Uber das Gefieder des afrikanischen Strausses,' Schr. 
phijs.-ok. Ges. Konigsb. xxix. 1889, SB. p. 31. 

1 ' On some Points in the Morphology of the Wings of Birds,' P. Z. S. 1887, 
p. 343. 

2 'Observations on the Anatomy and Development of Aiitcry.?,' Pltil. Trans. 


Though no rectrices can be distinguished, there are 
recognisable remiges. PAEKER counted nine or ten cubitals 
and two or three metacarpals and a single mid-digital ; there 
are also tectrices majores. An extraordinary peculiarity of 
Aptenjx is the situation of the nostrils near the very end of 
the beak. 

Of Casiiarius there are some ten species which are found 
in several of the islands lying to the north of the continent 
of Australia, such as New Britain, Ceram, &c., as well as- 
one species, Casuarius australis in the north of Australia 
itself. They are remarkable externally for their black 
coloration, brown in the young, and for the horny casque 
upon the head. The neck is naked and adorned with 
bright colours, in which blue is especially prominent, and 
there are often dependent folds of bright-coloured skin in 
this region. The feathers have an aftershaft as large as 
the feather itself; the rectrices are unrecognisable, but the 
remiges are present in the shape of long spines which corre- 
spond to the stems of the feathers. The claw of the inner 
of three toes is very elongate. 

The emu, Dronmus, 1 is entirely Australian in range, and 
contains two species. This genus, agreeing with the casso- 
wary in laying a green egg, has no helmet or wattle, or stiff 
spines upon the wing. It has, however, like the cassowary, 
a large aftershaft. 

The fourth genus of Struthiones is the South American 
Rhca, of which three species are recognised. These have 
been carefully compared by GADOW. 2 The genus is cha- 
racterised, so far as external characters are concerned, by 
the want of an aftershaft and by the feathered neck not 
naked, as in the ostrich ; it lays a yellowish white egg. The 
Rlica is three-toed. There is a distinct ventral apterion 
running from sternal callosity to vent. 3 

1 G. DUCH.UIP, ' Observations sur 1'Anatomie clu Dromons? Ann. Sci. Nat. 
(5), xvii. 1873. 

- ' On the Anatomical Differences in the Three Species of Rhca,' P. Z. ,S. 

->, p. 308. 

3 ' A. BOECKING, De Rhea Americana, Diss. Inaug. Bonn, 1863 ; J. F. 

K K 


Strnthio has been found fossil in the Siwalik Hills, in South 
Eussia and Samos. Rlica is found fossil in America (South). 
LYDEKKER considers Hypselornis sivalensis, whose place of inter- 
ment is indicated by the name, to be an emu. It is only known 
from the second phalanx of the third digit of the pes. 

( it'// i/i'>niis Newtoni, from Australia, 1 with a skull a foot long, 
seems to have been a gigantic emu. But it has not as yet been 
fully described. 

Dasornis londincnsis (from the Eocene clay of Sheppey) is 
placed by FURBRINGER among the Ratites, rather in deference to 
the opinion of Sir R. OWEN - than from conviction. GADOW, on 
the other hand, places it among Stereornithes. It is only known 
by a water-worn skull fragment, indicating a skull as large as that 
of the Dinornithidse. It seems useless to speculate upon the 
affinities of this fragment. 

Macrornis of SEELEY must remain for the present a name. 

In surveying the muscular system of the Struthiones 3 it is 
clear that, so far as concerns the muscles of the manus, 
Apteryx is, in accordance with other reductions in the bones 
of that limb, the most degenerate type. On the other hand 
(assuming, of course, the derivation of the Struthiones from 
some carinate form) the shoulder girdle of Apteryx has 
retained more of the primitive musculature than the other 

In all the genera the following muscles have disappeared : 
the pectoral-is propatagialis, biceps propatagiaMs, deltoid cs 
'propatagialis,* deltoides minor, scapulo-Jiumerdlis anterior, 
expansor secundariorum. 5 

The pectoral-is major is in all very reduced. 

All the struthious birds except Apteryx have also lost the 

VAN BEMMELEX, ' Onderzoek van een Rhca-'Emloijo,' Tijd. Ned. Dierk. Ver. 
1888, p. ccv. 

1 STIRLING and ZIF.TZ, ' Preliminary Notes on Geni/oriiis.' Ac.. Tr. Roy. Soc. 
S. Australia, xx. 

- On Diiiurnis (part xiv.), Tr. Zool. Soc. vii. p. 145, pi. xvi. 

3 GADOW, Zur vergleichenden Anatomic tier Musk/ilatur dcs Beckcns mid 
der liinteren Gliedmcisse der Eatiten. Jena, 1880. 

4 In Apteryx some elastic tissue in the patagium possibly represents this. 

s Traces have been asserted to exist in Aptcnj.r and Dromceus, but require 


serratus metapatagialis, the latissimus dor si metapatagialis, 
and the pectoralis abdominalis. 

On the other hand Apteryx has lost what the other 
struthious birds have retained, the latissimus dor si anterior 
and the rhomboideus profundus ; the latter muscle, however, 
is not distinguishable in the cassowary. 

It must be admitted, therefore, that Apteryx, so far as 
concerns the anterior extremity, has diverged from the 
hypothetical ancestral condition in slightly different lines from 
other Struthiones. 

In the cassowary l both rhomboidei are present, but they 
originate from ribs and not from the cervical vertebrae. 
The rhomboideus profundus is parallel with and hardly dis- 
tinguishable from a portion of the serratus profundus ; hence 
FUEBRING-BE is indisposed to admit the existence of a sepa- 
rate rhomboideus profundus. 

The serratus superficialis consists of two separate fan- 
shaped bands of muscle. The coraco-brachialis interims is 
entirely converted into tendon. The biceps originates only 
from the coracoid, and ends without being definitely split 
into two tendons upon both radius and ulna. There is only 
one scapulo-humeralis muscle, which is, however, of fail- 

The subscapularis is a single-headed muscle arising from 
the scapula only. 

The anconceus has a single origin from the scapula, and 
has no attachment to the humerus. 

In the hind limb all five ylutcci are present ; they are all 
large, especially gl. I. and gl. V. The ambiens is absent as 
a rule ; it is occasionally present, but is then imperfect, 
reaching only as far as the knee. The semitendinosus and 
its accessory are well developed. The femorocaudal is a 
small slender muscle ; it is inserted in common with the 
accessory, which is enormous in size. In Casuarius Bcnncttii 
at any rate there is an additional adductor of peculiar origin ; 
the muscle is two-headed, one head being a tendon which 

1 J. F. MECKEL, ' Beitrage zur Anatomie des indischen Casuars,' Arch. f. 
Anat. u. Pliys. 1830, p. 200, 1832, p. 273. 

K K 2 


springs from the muscular fibres of the accessory femoro- 
caudal, the other fleshy and springing from the pelvis 
just behind the acetabulum. It is inserted along the femur 
below the vastus interims and over the conjoined fernoro- 

In Strutliio the rhomboide-us superficial is arises, as in 
carinate birds, from the spinous processes of the vertebrae 
(1-3 cervicals) ; it is inserted only on to the scapula. The 
rliomboideus profundus arises from the spinous processes of 
the last cervical and first dorsal vertebrae ; it is inserted on to 
the end of the scapula. The serratus superficialis of Stru- 
tliio is a single muscle arising as two or three bands, either 
from the last cervical and first dorsal rib or, in addition, 
from the second dorsal rib. It is attached to the ventral 
border of the scapula. The serratus profundus is divisible 
into a more superficial and a deeper layer ; the former is 
the less extensive, and arises either by a slip from the rib 
of cervical vertebra 19, and by two slips from the last 
cervical rib, or by two larger slips and one very small one 
from between the last cervical and the first dorsal rib, and 
from the latter ; they are inserted on to the inner border of 
the scapula. The deeper layer also varies, but arises in 
several slips from the last two cervical ribs. It is also 
inserted on to the inner border of the scapula. The coraco- 
brachialis externus is very large as compared with the same 
muscle in the carinate birds ; it is not quite so large as in 
Eliea. The coraco-brachialis interims is larger in Strutliio 
than in any other ratite. The biceps arises from the spina 
coracoidea ; its muscular belly is not well developed ; it is 
inserted on to the radius and ulna, and on to the membrane 
between them. As with Ehea the deltoid arises from the 
scapula and neighbouring region of coracoid. The teres 
major, again, as in Rliea, is a comparatively large muscle. 

On p. 87 et seq. will be found an account of the muscles of 
the hand in Palamedea, which I have taken to illustrate that 
of the carinate birds in general. 

The differences which are to be noticed in Strutliio are, 
apart from minor divergences, the following : 


The extensor metacarpi radialis is single. 

The ectepicondylo-ulnaris is absent or fused with the 
extensor metacarpi ulnaris. 

The extensor digitorum communis supplies only the 

The two pronators form only one muscle. 

The flexor digitorum subliinis and the^. dig. profundus 
arise by a single head from the flexor condyle of the humerus. 
The two muscles immediately divide ; the upper part ( = 
sublimis) ends in two tendons, of which one is inserted on to 
radiale, the other fuses with the upper tendon of profundus, 
and also gives off two slips which surround that tendon and, 
reuniting, fuse with the lower tendon of the profundus. The 
lower part of the muscle (= profundus) gives off two ten- 
dons, of which the upper ends on the first metacarpal, while 
the lower runs to the base of the last phalanx of the 

The flexor metacarpi ulnaris ends fleshily on ulnare, but 
is prolonged beyond this bone, receiving also some fibres 
from it, to the metacarpal. 

The radio-metacarpalis ventralis or at least a muscle 
which, if it be not this, is not found in Palamedea arises 
from the ulna and not from the radius. 

The total number of muscles in the hand of the ostrich 
is twenty-three, allowing for the absent ectepicondylo-ulnaris. 
The additional muscle is a small pronator quadratus, running 
from the ulna to the radius. 

It appears, therefore, that, in spite of the small size of the 
manus of the ostrich relatively to that of flying birds, there 
is but little if any evidence of degeneration in its musculature. 
On the contrary, indeed, for it might be said that the wing 
muscles of Struthio are less degenerate, or at any rate less 
modified, than those of carinates in that amount of muscle 
as compared with tendon is greater. The complication of 
the conjoined flexores digitorum is highly suggestive of a 
walking or climbing animal. It seems to be conceivable 
that the ostrich branched off from the avian stem before the 
power of flight was perfectly established. 


The ostrich 1 has the complete leg muscle formula 

ABXY + . 

The femorocaudal is fleshy, but not large, and has no 
distinct tendon of its own. It blends above with the acces- 
sory. The accessory femorocaudal is an enormous muscle 
ending in a broad thin tendon which distally is lost in a 
fibrous expansion round the great vessels and nerves of the 
thigh. The accessory semitendinosus is small. The tendons 
of the semimembranosus and the semitendinosus become 
united just after the attachment to the latter of the acces- 
sory ; they soon, however, diverge, the semimembranosus 
being continued as a long thin tendon down the leg to join 
the tendon of the gastrocnemius. The obturator externus 
and the adductors are small ; the obturator interims is very 
large. The ambiens does not arise from the pectineal pro- 
cess, or even from the pubis, but from the ilium. An 
additional adductor muscle which has been referred to in 
the cassowary also exists in the ostrich ; it has, however, 
but one (tendinous) head, arising from the femorocaudal 

The rJiomboideus super jicialis of Rliea 2 springs, like that 
of Strutliio, from the spinous processes of the cervical verte- 
brae, but from a larger number (four). It is inserted on to 
the coracoid as well as the scapula. The rhomboideus pro- 
fundus arises from the spinous processes of the first three 
dorsals. As in carinate birds the serratus superficialis is 
composed of an anterior and a posterior section ; the former 
arises as a single band from the last cervical rib, and is 
attached to the front part of the scapula ; the latter is large 
and consists of three broad slips springing from the first 
three dorsal ribs and their uncinate processes ; it is attached 
to the hinder end of the scapula. It may be, FUEBEINGEE 
thinks, that a portion of this is really the pars superficialis 

1 S. HAUGHTON, ' On the Muscular Mechanism of the Leg of the Ostrich,' 
P. R. Irish Ac. ix. (lSf>G), p. ;">0 ; A. MACALISTEK, ' On the Anatomy of the 
Ostrich,' ibid. 1867, p. 1 ; Ai.ix, ' Sur 1'Appareil Locom. de 1'Autruche cle 
1'Afrique,' Bull. Soc. Philom. 1868. 

- S. HAUGHTON, ' Muscular Anatomy of the Khea,' P. E. Irish Ac. ix. (LSI)?), 
p. 497. 


of the serratus profundus ; otherwise that muscle only con- 
sists of the deeper portion which arises as two slips from the 
last two cervical ribs and runs directly backwards to be 
inserted on to the lower border of the scapula. 

The coraco-brachialis interims is largely tendinous ; its 
origin, contrary to what is found in other Struthiones, just 
extends on to the sternum. The origin of the biccjt.s is 
peculiar ; it arises not only from the coracoid spine by a 
rounded tendon, but also by a sheet of tendon edged with 
muscle from the whole' of the coracoid and from just an 
adjacent bit of the sternum. It is inserted on to both radius 
and ulna. 

In the manus of Rhea, on the other hand, we have more 
evidence of degeneration than in Stnithio. There are, in the. 
first place, only twenty-one muscles at most, and some o! 
these are much simplified. 

The muscles that appear to be totally wanting are (1) thf 
extensor digitorutn coinmunis, (2) the pronator profiindiu. 

The extensor indicis is only represented by the belly 
arising from the wrist. The flexor sublimis may possibly be 
represented by a slip of muscle arising from the tendinous 
edge of the flexor metacarpi ulnaris, which goes to be inserted, 
partly by tendon, partly by fleshy fibres, on to the ulnare and 
base of metacarpals 2 and 3. 

As in the ostrich the radio-metacarpalis ventralis arises 
from the ulna. In Rhea there is a special peculiarity in the 
presence of a muscular slip running from the tendon of the 
extensor metacarpi ulnaris near to its insertion to the extensor 
indicis. Finally the ectepicondylo-ulnaris is distinct. 

In the leg there is no femorocauda <l , the formula being 
BXY + . The accessory femorocan <lal is enormous, and there 
is a good struthious accessory adductor. Glut mix /iri/nus is 
very large and overlaps biceps ; glut(cus V. is present and 

In Dromaus 1 the rhomboideus superficialis and pro- 

' S. HAUGHTOX, 'Muscular Anatomy of the Emu,' P. E. Irish Ac. ix. 
p. 487 ; Cr. ROLLESTON, ' On the Homologies of certain Muscles connected with 
the Shoulder Joint,' Trans. Linn. Soc. xxvi. 1870, p. (ii)'.i. 


fnndus arise from ribs, the latter from only one, the former 
from three. 

The biceps apparently arises like that of Eliea. 

In the leg the ambiens and the f emor oca u declare wanting^ 

the formula being, therefore, BXY-. All the gluteals are 

present, and the first covers the biceps. The accessory 

femorocaudal is very large, but it does not appear to> 

possess the struthious accessory muscle. 

The muscles of the wing of Apteryx are, of course, de- 
scribed by OWEX in his account of the anatomy of the bird. 
But a fuller and later description, with illustrations, is to 
be found in T. J. PAEKEE'S paper upon Apteryx, and in 


There is no rhomboideus profundus. The serratus super- 
ficialis is one muscle arising from the first two cervical ribs ; 
it has the pars metapatagialis wanting in the other ratites. 
There is also &pectoralis abdominalis wanting elsewhere, but 
no latiss. dor si anterior. The latiss. dor si metapatagialis is 
well developed. 

The muscles running from the shoulder girdle to the 
humerus are reduced to six ; these are the two pectorals, the 
deltoides (single), the teres major, the coraco-brachialis longus,, 
and the c. br. brevis. 

The biceps is single-headed, arising from the coracoid ; it 
has long tendons at each end and a small belly in the middle ; 
it is inserted only on to the radius. The anconceus longus 
fuses early with the single-headed triceps. 

As might be expected from the presence of but a single 
finger, the muscles of the hand are much reduced. Perhaps 
the most noteworthy peculiarity is the presence of the gallina- 
ceous and tinamine muscle, the entepicondylo-ulnaris. 
There is a peculiar accessory bracJiialis anticus, seemingly 
only met with in Apteryx. The remaining ten muscles are 
bracJiialis <n/ticus, ectepicondylo-ulnaris, ectepicondtjlo- 
radialis, pronator ('? sublimis or profundus}, extensor meta- 
carpi ulnaris, extensor indicis longus, with one head from 
contiguous surfaces of radius and ulna inserted in A. australis 
on to carpo-metacarpus, in A. Bulleri on to base of distal 



phalanx, extensor longus pollicis ' inserted on to the thumb 
side of the carpo-metacarpus, flexor digitorum profundus, 
ulni-metacarpalis vent rails? ' In one specimen (A. an sir alls') 
a minute tendon was seen preaxial of that of the deep flexor 
and passing to the preaxial side of the carpo-metacarpus,' a 
rudimentary Interosseous dor sails in 
one specimen of A. Bi/Ueri. 

In the leg the muscle formula is 

The tongue in all these birds is 
aflat triangular organ, relatively small 
in size. The accompanying cut (fig. 
234) shows its characters in Rhea ; 
it does not show any differences of 
importance, in the other genera. 

The course of the intestine in 
Strutliio camelus is shown in fig. 
12, p. 28. It will be observed that 
it is in many respects exceedingly 
simple : thus the greater part of 
the small intestine is thrown into a 
series of short folds, with none of 
the longer and more specialised folds 
found in many other birds. The 
duodenal loop is the only part of the 
small intestine which shows a special 
fold, and this is a loop with a short 
lateral diverticulum Y-shaped, in fact. 

The next most characteristic feature of the intestines of this 
bird is the enormous large intestine, which is for the greater 
part of its extent thrown into folds like those of the small intes- 
tine. In Casuariiis (see fig. 11, p. 28) the small intestine is 
quite as simple as that of the ostrich ; the large intestine is 
short and straight. The emu is practically identical with the 

1 PAEKEE terms it extensor metacarpi radialis brevis. I identify it as 

' J Flexor ca>'2>i rmlialis of PARKER. 

wini (AFTER GADOW). 

.V, liyil(l'jln,-,;il Ili-vve. 


cassowary, but the duodenal loop was strong, longer, and 
narrower. It is interesting to find from MITCHELL'S paper 
that the intestine of Ehea is somewhat intermediate between 
that of Struthio and that of Casuarius. ' The anterior portion 
resembled Casuarius ; the rectum had an expansion recalling 
that in the ostrich, but much less strongly marked.' 

The following table contains measurements of the ali- 
mentary tract 1 in the Struthiones : 

Small Int. 

Large Int. 


Ft. In. 

Ft. In. 

Ft. In. Ft. In. 

Rlwa macrorhyncha 9 

4 2 


1 4i,l 9 

11 O 

4 5f 

1 11 

2 9 

M )) 


1 4 


Rhca amcricfiita 

15 3 

1 8 

2 4, 2 9 

9 . 

9 8 

2 2 

4 8 

" .. (young) 

. 5 10 

1 4 

2 6, 2 8 

Struthio camelus 9 . 

23 1 

32 9 

2 8, 2 11 

9 . 

23 4 

30 8 

2 10A 

9 . 


29 8 

2 7" 

<? . 

24 6 

31 8 

2 8, 3 1 

<? . 


24 9 

1 10 

9 . 

28 6 

33 2 

2 7 

Casuarius uniappendiculatus 9 

3 8 


4i, 5 

,, Beccarii 9 

4 8 



,, picticollis c? 





3 10i 


3 i. 3f 

Apteryx australis 

4 4 


t * *t 


Oiveni $ 

3 4i 

4 I 


Dromceus NOVCB Hollandice 

10 6 



Casuarius bicarunculatus 





All the Struthiones have cseca, which are especially deve- 
loped in the ostrich, where they have been described by Sir 
EVEEARD HOME as well as by GADOW. Apart from their 
length and structure the most remarkable fact about them 
is that, in contradistinction to what we find in other birds, 
they are inserted by a common orifice. 

These caeca dwindle gradually in diameter towards the 
tip, and are provided internally with a spiral valve of about 
twenty turns. I have attempted to compare these caeca with 
those of the Martineta tinamou (seep. 488). In the latter 
bird the caeca are furnished with numerous short diverticula, 

1 E. EEMOUCHAMPS, ' Sur la Glande Gastrique clu Nandou,' Bull. Ac. Belg. 1. 
(1880), p. 114. 



which are the outward expression of a reticulate internal 
structure, like a ruminant's stomach. Towards the extremity 
of the caecum the folds cease to be so definitely arranged in 
a network, and some to present an indication of a spiral dis- 
position. Rhea has also very large caeca, and there are 
traces of the spiral valve of the ostrich. The caeca of the 
cassowaries and the emu, as will be seen from the table of 
measurement, are very much smaller than those of the 
ostrich and Rhea. GADOW mentions an obscure formation of 


Mantelli. FRONT VIEW. 

o.oo, ill this and following 
figs., tracheal rings. 

an internal network by the presence of folds which may be 
compared with the structure of the caeca of Calodromas, 
already referred to. The caeca of Apteryx are long and nar- 
row, like those of the tinamous (excluding Calodromas) . 

Struthio has the most remarkable liver of all the Stru- 
thiones. The two lobes are intimately fused into one heart- 
shaped lobe. There is an indication of a spigelian lobe, as 
with other Struthiones, and the single bile duct (there is no 
gall bladder) opens, exceptionally, only 4 cm. from the 
pylorus. Ehea has, as a rule, no gall bladder, but GADOW 
found traces of one in a specimen dissected. This occurred 
(see figs. 19, 20, p. 34) both in Eli.Darwini and Eh. americu-n. 
Casuarius and Dromceus have a well-developed gall bladder, 


and MITCHELL noted in the former genus that the gall 
duct and the pancreatic duct opened into a distinct diverticu- 
lum of the duodenum. 

FIG. 237. SYEINX OF Rhea amcricana. FROXT VIEW. (AFTEK FOEBES.) 

In Apteryx there is a gall bladder. 

The chief source of information about the windpipe and 


syrinx of the Struthionidae is contained in a memoir by 
FOEBES. J These birds, as a rule, possess no specially modified 

1 ' On the .... Trachea in the Eatite Birds,' P. Z. S. 1881, p. 778. 



syrinx, a statement, however, which does not, as was first 
shown by ALix, 1 apply to RJiea. In that bird (see woodcut, 

FIG. 239. SYRINX OF Casuarius galcatus. FRONT VIEW.! (AFTER FORBES.) 


figs. 237, 238) there is not merely a pair of intrinsic muscles, 2 

1 Bull. Soc. Phil. 1874, p. 38. 

- First noted, apparently, by PARKER in TV. Z. S. vol. v. p. 238, foot note. 


but some considerable modification of the last tracheal and 
early bronchial rings. The last four tracheal rings are 
soldered together to form a cartilaginous box, which behind 
and in front shows no lines of demarcation between the 
several rings of which it is composed. There is a membrana 


Tp, Tp 1 , the pouch ; ap, opening into trachea ; c.rf.s, prolongation of upper end of pouch ; 

/&, /&', fibrous glands. 

tympaniformis and a pessulus ; the first three or four 
bronchial semi-rings are different from those which follow. 
The remaining Batitse have no distinct syrinx. In Struthio, 
for example (see figs. 43, 44, p. 64), although there is a mem- 
brana tympaniformis completing the bronchi internally, there 
is neither pessulus nor intrinsic muscles. The syrinx of 
Apteryx is about on the same level. Casuarius is rather 


different. The last few tracheal rings are incomplete 
posteriorly ; the space left between them is continuous with 
a membrana tympaniformis. There is no pessulus or 
intrinsic musculature ; in the division of the last tracheal 
rings there is a suggestion, faint perhaps, of the tracheal 
syrinx. Dromceus, as might be imagined, closely resembles 
Casuarius. It has, however, a peculiarity which has been 
fully gone into by MuETE, 1 who quotes the pre-existing 
literature upon the matter. In front of the trachea some 
way down the neck a certain number of the tracheal rings 
are deficient in front ; and the lining membrane of the tube 
here projects as a sac, which can be inflated, and has, no 
doubt, something to do with the drumming voice of the 
bird. The accompanying illustration shows this peculiarity, 
which is not met with in the cassowary. 2 

As to osteology? in RJtea (fig. 77, p. 140) the vomer tends 
to be bifurcate posteriorly where it is much widened out, and 
articulates both with the palatines and the pterygoids. The 
palatines are posteriorly flat and fenestrated. The maxillo- 
palatines are thin plates which meet the anterior bifurcation 
of the vomer. The descending process of each lacrymal 
has a large foramen, the presence of which led B. 0. CUNNING- 
HAM to distinguish Eh. Darwini from Eh. americana, where 
the foramen is simply a notch. GADOW, however, showed 
that the question of notch or foramen is simply individual 
variability, and I am in a position to assert the same of Eh. 
macrorhyncha. The existence of a complete descending 
process of the nasal has been denied. PAEKEB, however, 
has figured an ascending pillar of bone from the maxillary, 
and in a specimen of Rh. macrorhyncha this was joined by a 
suture to the anterior margin of the lacrymal. It has, it is 
true, no connection with the premaxillary part of the nasal ; 
but this can scarcely interfere with a comparison of the 

1 ' On the Tracheal Pouch of the Emu,' P. 7. S. 1.867, p. 40o. 

- For the lungs and air sacs of Struthiones see ante, p. 495. Those of 
Rhea have been described by W. N. PARKER, ' Note on the Respiratory Organs 
of Rhea,' P. 7. S. 1883, p. 141 ; of Dromceus by MALM, 'Om Luftrur-s-ickrn,' 
&c., Off. K. TV/. Ak. Forh. 1880, p. 33. 

3 PANDER and D'ALTON, Die Skclcte rfcr Ktraussartic;cn VUgcL Bonn, 1827. 



bone to the outer part of a nasal and to the naso-maxillary 
of the Dinornithidse (see below). 

There is a well-developed, though thin and curved, 
ectethmoid lamina, which joins the maxillo-palatine below 
and the descending process of the lacrymal above. This has 
been also stated to be absent. 

Rhea has seventeen cervical vertebra. The atlas is 
notched, as in Struthio, but not so widely. In the shoulder 
girdle the procoracoid is short, but is continued down to the 
articulations of the coracoid by the menibrana coracoidea, 
of which, in a specimen of Rhea macrorhynclia before me, a 

cc, coracoid grooves ; ca, anterior lateral process ; /, keel i ? ) : It; posterior lateral process. 

portion is ossified as a thin spicule of bone shutting in the 
foramen coracoideum. The sternum (see fig. 242) has a 
median ventral prominence and two lateral thin rings of the 
bone, which may be indications of foramina. Three (some- 
times four) pairs of ribs reach the sternum. The pelvis 
(fig. 243) has a small pectineal process. The pubes join 
the ischia posteriorly, and anteriorly an interobturator pro- 
cess, of which there are faint indications in Struthio, unite 
the two bones. Posteriorly the ilia are attached to the 



The structure of the skull of the emu : is not widely 
different from that of the skull of Ehea, The vomer is 
widely bifurcate behind, where it articulates both with 
pterygoids and palatines. The basipterygoids articulate 
with the pterygoids at the extreme posterior end of the 



il, ilium ; Ip, peetiueal process ; at, antitrochauteric process ; st, supratrochanteric process ; 
jjs, interobturator process ; ;', ischiuin ; p, pubis. 

latter, instead of nearly halfway along them, as in other 
birds. The maxillo-palatines are hollow swollen plates, 
which unite with the vomer and premaxillaries, but come 
apart in the dried skull. The descending process of the 
lacrymal has a foramen, as in Ehea. It joins the thin 
lamina of the ectethmoid. The descending process of the 

1 W. K. PARKER, ' On the Structure and Development of the Skull in the 
Ostrich Tribe,' Phil. Trans.. 1808, p. 113. 

L L 



nasal is only represented in the specimen before me by a 
minute pointed bit of bone attached above to a point cor- 
responding to that whence the ' naso-maxillary ' arises in 

Dromceus has twenty cervical vertebra. The atlas is 


Pmx, premaxilla ; Mxp, maxillo-palatine ; T", vonier ; PI, palatine ; I't, pterygoid : 

*, basipterygoid process. 

notched, very nearly perforated (fig. 66, p. 118). Thejprocora- 
coid is not quite so well developed as in Rkea, but there 
are a pair of rudimentary clavicles. 

The sternum (fig. 74, p. 129) ' much resembles that of 
Rhea ; ' it is not notched and is rather pointed at its 
extremity. Three or four ribs reach it. 



The pelvis (fig. 245) has the pubes and ischia quite free 
posteriorly in the dry skeleton ; but they are united by 
cartilage, as in the latter with the ilium. The interobturator 
process is present, and shuts off an anterior portion of the 
obturator foramen. 

The skull of Casuarius : is very like that of Drom&us. 

The number of cervical verterbrce in Casuarius varies 
from eighteen to nineteen. In the atlas the indication of the 




closing of the notch for the odontoid process may be com- 
pleted, as is shown in the cut (fig. 68, p. 118). The shoulder 
girdle is very like that of the emu, possessing also rudimen- 
tary clavicles. The membrana coracoidea may, however, 
be ossified, and there are two foramina. One of these lies 
between the membrana coracoidea and the coracoid, and 
is therefore apparently the homologue of the foramen in 

1 W. H. FLOWER, ' On the Skeleton of the Australian Cassowary,' P. Z. S. 
1871, p. 32. 

L L "2 



Dromceus ; the other is smaller and further back in the 
substance of the coracoid. 

The sternum (fig. 246) is an exaggeration of that of 
Dromceus, being longer and more pointed posteriorly. Four 
or five ribs articulate with it. 

The pelvis too, though very like that of the emu, is (fig. 
247) an advance upon it in structure. There may be (C. 



c, coracoid groove ; mx, posterior eud. Other letters as in fig. 242. 

galeatus) or may not be an osseous union between pubis 
and ischium and between ischium and ilium. 

The ostrich skull is rather unlike that of the other two. 

The vomer is very short ' and does extend back as far as 
the articulation of the palatines and pterygoids. The latter 
bones articulate not only with the basipterygoid processes 
but with the basisphenoid ; they bear off the palatines, which 

1 ' W. GRUBER, ' Ueber das Thranenbein cler straussartigen Vogel,' &c., Bull. 
Ac. Sci. St.-Pttersb. 1855, p. 161. It varies somewhat in length according to 
FURBRINGER, and was found in one case to be not unlike the vomer of an 
segithognathous bird. 



run forward in a straight course. The maxillo-palatines 
articulate with the vomer. 

The axial skeleton has been described in the greatest 
detail by MiVAKT. 1 

There are twenty cervical vertebra;., and then five which 
have ribs articulating with the sternum. The atlas is more 
simply ring-like than in other birds ; it has a very wide notch 



pl, pelvic rib. Other letters as in fig. 243. 

for the odontoid process of the axis. The catapophyses of 
the seventeenth cervical unite. 

The sternum has a raised and flattened tract posteriorly, 
which may be the equivalent of the keel ; it has two posterior 
lateral processes, which extend beyond the median portion of 
the bone. 

1 ' On the Axial Skeleton of the Ostrich,' TV. Z. S. viii. p. 385. 


The pelvis is remarkable for the symphysis of the pubes, 
which is shown in the accompanying figure (fig. 250). The 
ischia also unite each with its corresponding pubis. There 
is a well-developed pectineal process. GARROD ' and F. DAR- 
WIN described a small ossification attached to the front 


A", rostrum. Other letters as in fig. 244. 

margin of the pubis which may conceivably be the homologue 
of the marsupial bone of the marsupialia. 

The skull of Apteryx has been described by OWEN 2 as 

1 ' Notes on an Ostrich lately living in the Society's Gardens,' P. Z. S. 1872, 
p. 356. 

'' ' On the Apteryx australis,' Trans. Z. S. ii. p. 57, and iii. p. 277. 



regards its adult structure, and the development has been 
lately treated of by T. J. PARKER. 1 

Apart from the very elongated anterior part of the skull 




the characteristics are those of the struthious birds generally. 
The Y-shaped posterior end of the vomer bears off from 


sy, sympbysis pubis. Otber letters as in fig. 243. 

articulation with the rostrum the palatines a^id pterygoids. 
The basipterygoid processes are large. As in all other 

1 ' Observations on the Anatomy and Development of A /it, )//.>:,' Phil. Tran 
clxxxii. (1891), p. 25. 


struthious birds, with the exception of the adult cassowary 
and the Dinornithidae, the ossified ethmoid appears on the 
dorsal surface of the skull between the nasals. In the adult 
the sutures disappear, and the bones are so firmly united 
that the quadrate and the columella are the only movable 
bones in the skull. The quadrate has a two-headed otic 
process, differing so far from other struthious birds and 
agreeing with the Carinatae. 

PAEKEK'S statements as to this matter are opposed to 
those of FUEBEINGEE in the table of differential characters 
which he gives in his great work. The vertebral column is 
described by the authors already quoted as well as by 
MIVAET.' There are sixteen cervical vertebra, and four ribs 
articulate with the sternum. The atlas is either perforated 
or only notched by the odontoid process, and it is imperfectly 
joined above, not always, but in many cases, at the summit 
of the neural arch. The tenth and eleventh vertebra have 
sometimes a ventral hypapophysial canal, as in Herodiones. 
MIVAET found this to be the case with A. Oweni. I found 
the catapophyses to approach each other very closely in 
that species and in A. australis, but not to fuse. 

The sternum is somewhat variable in form. Occasionally 
the posterior lateral processes exceed the middle process in 
length ; sometimes they are less or subequal. As a rule the 
sternum appears to be broader than long, but this is not 
invariably the case. The varying proportions of the sternum 
and the lengths of its several processes seem to offer cha- 
racters diagnostic of the species. In two specimens of A. 
Bulleri PAEKEE found a ' low ridge nearly as well marked as 
the vestigial keel of Stringops.'' The shoulder girdle, like the 
sternum, is subject to great individual variation. The relative 
lengths of the scapula and coracoid vary ; the curve of the 
scapula varies, but it is in the coracoid that the most inte- 
resting variations occur. The coracoid notch, converted by a 
ligament into a foramen, but being in the embryo a distinct 
foramen in the cartilage, is sometimes absent, its place being 

1 ' On the Axial Skeleton of the Struthionidse,' Trans. Z. S. vol. x. p. 1. 
See also ALLIS, ' On the Skeleton of the Apteryx,' J. Linn. Soc. 1873, p. 523. 


indicated by a thinning of the bone. The rudiments of 
tuberosities for the attachment of the missing furcula are 
often fairly evident. The coraco-scapular angle oscillated 
between 150 and 122. There is a supracoracoid foramen. 

In the pelvis there is no fusion between ilium, ischium, 
and pubis. The pectineal process is long and appears to be 
ossified equally by pubis and ilium. In the skeleton of the 
adult foot two of the tarsals are present as free bones not 
fused with either the tibia or the metatarsus ; these are, 
according to T. J. PAEKEE, two centralia. 

The bones of the wing in the struthious birds are espe- 
cially reduced in the emu, cassowary, and Apteryx. 

The wing of the adult emu (Droiium* ater) has been 
figured and described by PAEKEE. There is no trace of a 
separate carpus either in young or adult. In a six-weeks-old 
chick the first rnetacarpal is half the length of the second, 
but in the adult it is reduced to a small prominence not a 
third of its length. There is no trace of a third metacarpal. 
The single finger (the index) has three phalanges and a long 
strong claw r . ' The wings of an adult are about the size of 
those of a jay or a bower bird ; in the young chick, with legs 
the size of those of a turkey, the wings are no longer than a 

In Apteryx the wing is in some respects further reduced 
than that of the emu ; in others less so. In the adult A. 
australis (T. J. PAEKEE) there are no distinct carpals, but a 
broad flattened carpo-metacarpus, with traces of being com- 
posed of three metacarpals. There are sometimes two and 
sometimes three phalanges the last clawed to the single 
finger (index) ; where one is atrophied it is the second. In 
A. Oweni there appears to be invariably a distinct radiale ; 
the third metacarpal is more distinct than in the last species, 
and in one case w r as entirely free. The clawed index has 
two or three phalanges. The single example of A. Haa^ti 
which PAEKEE examined had an ulnare as well as a radiale 
in the carpus, a fairly distinct metacarpale III., and three 
phalanges to the index. 

In A. Bullcri the manus shows much greater variations ; 


in one specimen a radiale is present in the carpus, in another 
a bone which appears to represent radiale and distal carpals ; 
this specimen had a free third metacarpal. In two other 
instances there is a carpo-nietacarpus, as in A. cms trails. 
There are two or three phalanges and, as always, a claw to 
the index. 

The development of the manus of Apteryx shows plainly 
what is also apparent from its adult structure, that it is in a 
condition of degeneration. Traces of three distal carpals, as 
well as of radiale and ulnare, are visible ; all the metacarpals 
are distinct, the third being as long as the second and having 
a rudimentary phalanx. 

The -ZEpyornithidse, containing the type genus sEpyornis 
and a recently established new genus, Mullerornis, 1 was for 
some time only known by the subfossil egg and by the 
bones of the hind limb. More recently Messrs. MILNE- 
ED WARDS and GEANDIDIEE, and more recently again Mr. 
C. W. ANDREWS, have described other parts of the skeleton, 
so that now, though there are still many lacunae, we have a 
fair knowledge of several important parts of the skeleton. 
This family is limited to Madagascar, where its remains have 
been found chiefly in marshes. 

The skull is only incompletely known the palate, for 
instance, so important in determining its affinities, is quite 
unknown being only represented by what is little more 
than a calvaria, and by an imperfect mandible. The occipital 
condyle is pedunculate, as in the moas. The frontal region 
of the skull is covered by many pits, which are arranged in 
a fairly regular fashion ; it is suggested that these may be 
the marks of the inplantation of feathers, of which, therefore, 
the Mpyornis may have possessed a frontal crest a feature 
which has also been observed in certain moas. There are 
also, as in the moas, a prominent basi-temporal platform, 

1 ' Observations sur les JEpyornis de Madagascar,' Comptcs Bend, cxviii. 
1894, p. 122 ; ' Sur les Ossements d'Oiseaux,' &c., Bull. Mus. Nat. Hist. 1895, 
p. 9 ; ' On the Skull, Sternum, and Shoulder Girdle of JEpyprnis,'' Ibis (7), ii. 
p. 376 ; ' On some Remains of JEpyornis in the British Museum,' P. Z. S. 1894, 
p. 108. 



an open Eustachian groove, and a similar structure of the 
articular facet for the quadrate. 

The sternum is singular by its extraordinary breadth 
and great shortness ; the length in the middle line is only 
one-fifth of the greatest breadth. The hinder border is not 
notched, but forms a ' gently concave curve.' The antero- 
lateral processes are stout. There is, of course, no keel. 
The coraco-scapula is typically ratite, the angle between 

sc, scapula ; pc, proooraooid ; f.spc, foraiin/n suiiracoracoideum ; ;/7, gleuoid cavity. 

the two being very slight. As will be seen from the figure, 
it most resembles that of C<t*n<irins. The bird had a 
rudimentary humerns. 

Dinornithidae. This family consists of a number of 
genera, all New Zealand in habitat ; their remains are so 
abundant in various parts of the country that they must 
have existed in countless numbers. That there should have 


been within so limited an area at least twenty-five distinct 
species is explained by Captain Hutton by the view that at 
one time the two islands of New Zealand were divided up into 
a greater number an archipelago, in fact the result of this 
being what we now see among the cassowaries, where each 
of the islands inhabited by them has its own peculiar species, 
isolation, indeed, permitting of the specialisation. All the 
moas, however, became extinct at a period not less than 
three or four hundred years ago. T. J. PAEKEE,* whose 
work on the cranial osteology of the group is the most 
recent, allows the genera Diuorui*, Pacliyoniis, Mesopteryx, 
Anomalopteryx, Emeus, and probably Megalapteryx, dis- 
tributed among three subfamilies. The moas a general 
term applied to all these genera were birds of fair, often 
large, size. The smaller species ranged from 2^- to 4 feet in 
height ; the largest were at least thirteen feet high. 

The skull of the moas had a short and wide beak. The 
occipital condyle is remarkable on account of its ' more or 
less pedunculate character,' a circumstance which is of 
importance in considering the relationship to the moas of 
the Madagascar ^Epyornis (see p. 522). 

The orbit is smaller than in other struthious birds. The 
nasals are peculiar in that they meet behind above the 
ethmoid, so that no part of the latter bone appears on the 
upper surface of the skull. It is only in the adult cassowary 
among recent struthious birds that the ethmoid is entirely 
hidden on a superficial view, a state of affairs which is 
brought about by the development of the crest, and does 
not exist in the young bird. The palate is like that of the 
emu and cassowary, but is most like that of Apteryx. 

The nasal bone is furnished with a slender maxillary 
process, or, as in emus, there is a corresponding bone 
separately ossified. The lacrymal is firmly ankylosed to 
frontal ; its descending process joins ectethmoid. T. J. 
PAEKEE has figured and described a peculiar thin scroll -like 
bone which appears on a lateral view of the skull and pro- 

1 ' On the Cranial Osteology, Classification, and Phylogeny of the Dinorni- 
thiclaV Tr. Z. S. xiii. p. 373. 


jects beyond the anterior margin of the maxillo-nasal ; this 
he has termed the alinasal. 

The number of cervical vertebrae is large, at any rate in 
Anomaloptcryx parva, the only species in which they are 
all without doubt preserved. There are in this bird twenty- 
one. The sternum is longish and rather narrow 7 in Atiomalo- 
pteryx casuariiia ; it is short and broad in Dinornis maximus. 
In all it has a pair of lateral notches strongly marked ; the 
lateral processes are strongly divergent. There is also a 
median posterior notch. 

The pectoral girdle is but little known, and appears 
sometimes to have been completely absent. 

In the pelvis the bones are separate and the pectineal 
process but little marked. 

That the feathers have large aftershafts, like the emu, 
&c., was first discovered by the late Mr. DALLAS.' Sir B. 
OWEN has figured the ossified rings of the trachea ; but they 
present no special features of interest. 

As to their relationships with other ratites, T. J. PARKER 
is of opinion that they form, together with the Apteryx and 
cassowaries, a definite branch of the struthious tree, as in 
the annexed diagram, which is from his paper. FURBRINGER 
comes to conclusions which are not greatly different. The 
relations of the Dinornithida3 to Strutliio and Eliea are 
' ganz entfernt,' to Dromceus and Casuarius ' fern,' but 
to Apteryx ' nahe.' 

There is no doubt that Struthio is removed far from the 
Dinornithidae, as well as from other ratites, by the structure 
of its palate, which diverges much. But it not clear that 
Rhea is so remote ; the existence of an apparent homologue 
of the maxillo-nasal bone, to which I have referred in the 
description of the skull of Rhea, is a point of somewhat 
striking likeness to Emeus, while the conformation of the 
skull generally in Rhea does not seem to divide it very 
deeply from Caxinirins, kc. Though no doubt T. J. PARKER 
is right in directing attention to the special resemblances in 
the skulls of Apteryx and the Dinornithidee, it must not be 

1 ' On the Feathers of Dinornis robnstus,' OWEN, P. Z. S. 1865, p. 205. 


forgotten that Dinoniis, like other ratites, except Apteryx, 
has a single head to the quadrate. In the characters of the 
pelvis Diiiornis is near to Apteryx and and 
remote from BJiea (as well as from Struthio). The large 
aftershaft allies it to the Casuariidse. NATHUSIUS has 
commented upon the practical identity in egg-shell structure 
which Rhea shows to Dinornis, a likeness which impressed 
him so greatly that he proposed to place them in the same 
genus. A considerable number of the special relations 
between Apteryx and the Dinornithidae, upon which FTB- 
BBINGEE writes, such as failing pneumaticity, absence of 
clavicle, mutual distance between coracoids, and even the 
form of the sternum, may largely depend upon the loss of 
flight, which is more complete in these birds than in the 
ostrich, for example. In no less than three footnotes 
FUBBEINGEE comments upon the supposed absence of 
uncinate processes to the ribs of the Dinornithidae ; but 
this difference from other ratites does not exist, as T. J. 
PAEKEE has definitely asserted their presence. It is signi- 
ficant that in his tables of differential characters FUBBEINGEE 
refers little to those of the fore limb girdle (including 
sternum) as distinguishing Dinornithidae from Rkea. A 
detailed account of the pros and cons will be found in the 
systematic part of FUEBEINGEE' s work, and as regards the 
skull in PAEKEE 's paper already referred to. 

The Struthiones have been often held to be more primitive 
than any of the existing groups of birds. 

There are really, however, not a large series of characters 
in which they may be fairly said to be more primitive 
than some other groups, and most of these are shared by 
some others. 

The form of the palate and the single-headed quadrate 
appears to be a low character ; but the former is shared 
with the tinamous, the latter with some other groups. The 
incompleteness of the fusion of the cranial bones may be 
looked at in the same way ; but the penguin is on the same 
level as the Struthiones. The absence of any fusion distally 
between the bones of the pelvis in Apteryx and Dinornis is 










dinosaurian ; but the tinamous are like Aptcryx in this. 
The complete procoracoid of Stnithio seems to be an archaic 
character, as do the two free centraliain the foot of Apteryx. 
As to negative characters, the most important of those that 
are possibly, but not certainly, to be regarded as primitive 
appears to be the usual absence of the oil gland. 

The long rectum of Struthio is probably an ancient 
character ; but whether the absence of a bird-like syrinx in 
all ratites except Rliea is a similar feature seems to be 
doubtful. The large size of the blood corpuscles in the 
ratites is noteworthy in this connection. 

The following table gives the principal characters of the 
existing genera. From it may be inferred the somewhat 
less modified condition of Aptcryx and the very isolated 
position of Stnithio among the members of the group :- 






Aftershaft . 



/Ifiiimboideiis prnfiiiuli i 





L/it. dni-fi (inferior 






Lat. dorsi metapat. 


fterritfHt: niii/tpat. 


Pectoralis abdom. 





Mindf for in ula of ley 


BXY + 











Syrinx . 






Lanje intestine 

Very long 

Not long 

Not long 

Not long 








Manus . 

3 digits 

3 digits 

1 digit 

1 digit 

1 digit 

Cervical vertebrce 






Atlas . 

Notched by 



Notched or 





A symphysis 

Ischia and 

Bones free 

Bones free 

Bones free 

pubis ; ischia 

ilia join 

join pubes 




S.\n. v ni!NITIIKS 



As there is but a single genus, and in all probability but 
a single species, 1 in this group, it is useless to attempt any 
formal definitions of family or other characters. I shall 
merely give the more important facts in its structure, as I 
have with the foregoing groups. As to extenutl cli'iracterx, 
the ArchcBOpteryx has an anisodactyle foot, like that of the 
Passeres. The feet and the digits of the maims have been 
stated to have been covered with scales. That scales may have 
been present, at least on the foot, is very probable, but there 
is not the faintest evidence of their having been there. Of 
feathers the remiges and rectrices are plain, while of the 
general body feathering there is not so much evidence. 
With the exception of a circle of feathers upon the neck, 
suggestive of those of the condor, and similar rings of 
feathers upon the ankle, it is thought by some that the 
Archceopteryx was naked. Most of the restorations, how- 
ever, admit a general feathering. The chief criticism to be 
offered is the extreme perfection of the remains of such 
feathers as are visible in the slab of stone in which the dead 
bird was originally imbedded. This being the case, the 
apparent absence of feathers over the general body surface 
gains more weight. That they may have been present and 
of the nature of down feathers is believed by reason of 
certain faint indications of something to round the contours 
of the body ; the group of contour feathers upon the leg are 
plainly visible even in photographs of the Berlin example. 
This example is much better than the specimen in London, 
which is the only other skeleton in existence. The rectrices 
are quite obvious, a pair to each of the separate vertebras of 
the tail. There appear to have been not fewer than thirty 

' It has been argued that specific and even generic differences exist between 
the London and Berlin examples. 

31 M 


of these ; FTIIBRING-ER places the number between that 
figure and forty. Most of the restorations allow thirty- 
two or thirty-four. This number is important ; it is in 
excess of that generally found in living birds, although the 
tail itself is not composed of actually more vertebrae. 
Among recent birds it is perhaps a significant fact that 
llie penguins alone have this number. Of remigcs seven- 
teen appears to have been the number, six or seven 
primaries and ten secondaries. No existing bird has 
so few primaries, the nearest approach being nearly all the 
Anomalogonatae (and some other birds too), which have ten. 
There is some diff