402 MR. W. p. PYCEAFT ON THE 



to bear uncinate processes. In JFidigiila, as in Anser, there are 

 six uncinate-bearing ribs, the first being lliat commonly regarded 

 as the last cervical, since it does not become attached to the 

 sternum. In Erismatura there are seven uncinates, the first 

 attached as in Biziura. 



There are altogether 10 thoracic ribs in Biziura, of which 8 

 are attached to the sternum, the 9th just fails to reach the 

 sternal border, while the 10th is vestigial. JErismatura possesses 

 9 thoracic ribs, of which 8 are attached. Only in Anser and 

 the Swans, it appears, is so high a number of ribs as 10 

 attained. In Cygnus hiiccinator there are 9 ribs attached to the 

 sternum, while the 10th is of great length. 



No less than four ribs are overlapped by the j)re-ilia in Biziura, 

 and the same is true of Anser and Branta among the Geese ; 

 but in no other Ducks are there more than three overlapped 

 ribs, so far as I have been able to discover. 



The great number of ribs in Biziura is to be taken as a 

 primitive character, since in the more highly specialized ducks,, 

 as in Querquedula for example, there are only 8 pairs, of 

 which but 6 pairs reach the sternum. 



The Pelvic G-iedle. 



The pelvic girdle of Biziura is remarkable for its extreme 

 narrowness, which gives it an unusually elongated appearance. 

 This narrowness is partly the result of a reduction of the width 

 of the postacetabular ilium, and partly of the transverse pro- 

 cesses of the synsacrum. The result of adaptation to diving 

 habits, this girdle has come to bear a really striking resemblance 

 to that of Podiceps or Colymhus, in which this narrowness has 

 reached its maximum. As in these birds, so in Biziura, the 

 preacetabular ilium has become shortened, while the postace- 

 tabular region has become lengthened. 



Eelatively longer than in any other Anatidse, the propor- 

 tions of the innominate bones are markedly diff"erent. Thus 

 in Biziura the length of the preacetabular ilium from its 

 anterior margin to the base of the antitrochanter is exactly 

 \ of the whole pelvis. Ifergus approaches nearest to these 

 measurements, the length of the preacetabular ilium, taken as 

 in Biziura, is |. CEdemia, Somateria, and Mrismatura give a 

 measurement of f. 



This forward shifting of the acetabulum is to be taken as a 

 sign of specialization — of adaptation to the requirements of 



