90 



MU. W. t. PYCBAFT Olf THE 



[Feb. 16, 



more to the purpose is the fact that there are many points of resem- 

 blance between the pelvis of Fregata and that of certain Procel- 

 laricv, e. g. Bnlivena. That this indicates a relationship, though 

 remote, is not improbable. This being so, Fregata may be regarded 

 as a link connecting Fhaeihon, which is undoubtedly one of the 

 least specialized and most primitive of the Sieganojpodes, with the 

 Procellariifonnes. 



The humerus assumes two forms : — 



(1) The pectoral crest is triangular in form, and the crista 

 inferior more or less inflated — Fregata, Fhaeihon (fig. 5), PeJecanus 

 (fig. 6) ; aiid (2) with the pectoral crest represented by a slight 

 I'idge but little raised above the level of the shaft, and the crista 

 inferior hardly or not at all inflated — Sula, Phalacrocorax, Plotus 

 (fig. 7). 



Fig. 5. 



Fig. 6. 



Fig. 7. 



Anterior aspects of the proximal end of the humeri of (5) Phaethon flavirostris, 

 nat. size, (6) Pelecatius rufcscens, J nat. size, and (7) Plialacrocorax carbo, 

 I nat. size. 



CL, Crista inferior ; P.c, Pectoral crest ; C.g., Coraco-hameral groove. 



The forearm and manus offer no characters sufficiently marked 

 to be diagnostic: that of Phadthon, for instance, is not easily 

 distinguishable from that of many Limicolce, and that of Pelecanus 

 from that of many Ciconice. 



