248 Annals of the Carnegie Museum. 



characters of the skeleton of these parts are truly those of the ibis in 

 A/a/a, yet in the latter the tihio-tarsiis, the farso-nietatarsiis, and the 

 joints of pes are each and all proportionately very much shorter and 

 stouter than they are in Plegadis. In the case of the tibio-tarsus it has a* 

 length in the Spoonbill exceeding that bone in the ibis by but a trifle 

 more than two centimeters ; while in the case of the tarso-metatarsus 

 there is a difference in length of but about half a centimeter in favor 

 of Ajaja. 



The several toe -joints in the foot of the latter are also proportion- 

 ately stouter than the corresponding ones in the skeleton of pes in 

 Plegatiis, but they vary considerably in the matter of their several rel- 

 ative lengths. For example, the basal phalanx of the middle toe in 

 Ajaja has an extreme length of 3.5 cm.; in P/egadis it measures 3.2 

 cm. while in the case of the second joint of the inside toe, we find 

 they are precisely of the same length in the two species. 



On the Taxonomy of this Group 



From my studies of the osteology of the American and other forms 

 representing the Herodiones, and taking into consideration all else 

 that has ever come to my notice upon the rest of their morphology and 

 habits, I conceive the several groups of birds composing this suborder 

 to have a relation to each other something after what is proposed in 

 the subjoined scheme, viz. : 



SUPERFAMILIES. FAMILIES. 



r T, -J ., f Plataleidse. 



Ibidoidea < ,, .,-, 



Suborder. I I Ibididn;. 



• T I /^- J ( Ciconiidas. 



Herodiones J Ciconoidea ^ Scopiidre. 



1, , • . .^ ., f Bala-nicipidae. 



baktmcipitoidea -( ^ ^, 



1 ^ ( LancromidtB. 



Ardeidoe. 



Taken as a whole there is hardly any question now I think, but 

 that the Herodiones are linked with the Anseres, through the Plata- 

 leidcB of the present group, and the Flamingoes, of the group next to 

 be considered. Perhaps Ajaja ajaja is the species among the Amer- 

 ican Ibidoidea most nearly related to Pluvuieoptems rither. Many other 

 avian taxonomers of the first rank likewise claim that the Herodiones 

 are in some way linked vvith the Accipitres or e\'en with the Stet^ano- 

 podes, with the first by the Storks leading through the Secretary 

 Bird, and with the latter in some strange way with the Tropic Birds 

 or even the Frigate Birds. To me, the first-named affinity may be 



