10 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 1 39 



found them feeding in shallow waters, often in riffles where they 

 scoop at their living prey, rather than spear at it as is the custom with 

 the typical herons. The eyes, wood brown by day, at night reflect the 

 jacklight with a faint orange sheen, which I have not observed in 

 other herons. The eggs are pale, nearly white, and often are lightly 

 speckled with brown, so that they resemble those of the tiger bittern, 

 Tigrisoma lineatum, rather than those of the night herons, which are 

 deep blue. 



While there is no fossil record for the boatbill, I regard it as an 

 ancient sideline from the typical herons that, judged from its present 

 restricted range in the American Tropics, has not been too successful. 

 It may seem attractive to unite Cochlearius with the true herons, 

 but from long acquaintance I regard their characters, briefly outlined 

 above, sufficient to maintain a separate family status. 



In view of the fact that the structural characters of the Balaenicipit- 

 idae have been summarized clearly by Stresemann (1934, p. 809), it 

 seems strange that the status of this family has been a matter of ques- 

 tion. The single species shows affinity both with storks and with 

 herons, in addition to outstanding peculiarities of its own. Miss 

 Cottam (1957, pp. 51-71) has made a careful summary of the osteol- 

 ogy from which she deduces a pelecaniform relationship, but this 

 appears to be due to convergence rather than to actual relationship. 

 The great enlargement of the skull has occasioned superficial re- 

 semblances to pelicans, but these, and others seen elsewhere in the 

 skeleton, are subordinate to the general sum of all characters, which is 

 ciconiiform. 



Phoenicopteri. — The position of the modern flamingos, which show 

 characters that point on one hand to the Ciconii formes and on the 

 other to the Anseri formes, has been a matter of some variance in al- 

 location. Mayr and Amadon (1951, pp. 7, 33), with only brief 

 discussion, have set them up as a distinct order, but general opinion 

 has carried them as a suborder allied to the herons, storks, and their 

 relatives. The latter course remains justified when the fossil genera 

 Palaelodus and Elornis of the upper Eocene to Miocene of western 

 Europe are considered ( Wetmore, 1956, p. 3). This group of flamingo 

 relatives was identified in North America when Alden Miller (1944, 

 p. 86) described Megapaloelodus connectens from the lower Miocene 

 of South Dakota, a species to which remains from the upper Miocene 

 of California also are referred (Loye Miller, 1950, pp. 69-73; 1952; 

 pp. 296-298). The group may be recognized as the family Palaelodidae, 

 on the generic name Palaelodus Milne-Edwards (1863, pp. 157, 158). 

 (There has been confusion relative to the proper spelling, since Milne- 



