IS'.JO.] riio.M THE Loxuox fj..vi" or siijji'i'iir. 781 



a? one or two minute foramiua, nmrkiiig points where the closure 

 of the cleft is incomplete. Probably in the early ancestors of 

 Phaethon the nares were schizorliinal, and in Fropliaetlion they still 

 approximate to that condition. It must be pointed out that in the 

 Stegaoopodes generally there seems to be a strong tendency to the 

 reduction in size o£ the external nares, and in Sula this has been 

 carried so far that the opening is reduced to a very small foramen ; 

 and it is remarkable that in Odontopteriix'^, which was a con- 

 temporary oi Prophaethoa, this condition seems to have been already 

 attained, so that in this respect this Eocene type is more specialized 

 than the recent Phaethon. 



iSo far as the evidence of the skull goes, it may be concluded 

 that Pro p?iaethon approaches very nearly ro Phaethox,, of which it is 

 probably an ancestral form, exhibiting in a few points more primi- 

 tive characters. 



The dimensions of the skull and mandible are : — 



milliiu. 

 Totallength from oec. condyle (tip of beak wanting) . . 112 



Width at squamosal prominence (approx.) 40 



AVidth at temporal fossae 26 



Width at postorbital process 46 



Width between orbits 17 



Width at rostral hinge 16 



Width of beak opposite anterior angle of antorbital 



vacuity 16 



Diameter oi foramen magnum. 8 



Length from occipital condyle to rostral hinge in straight 



line 51 



The Pelvis. (See Plate LI. and text-figure 2, p. 7S2.) 



By the careful removal of the mati'ix a great part of the pelvis 

 is now exposed, but it is incomplete posteriorly and the right side 

 of the preacetabular portion is still concealed. 



In the preacetabular region the ilia are united along the middle 

 line A^dth the neural spines of the sacral vertebrae to form a broad 

 low iliac crest. Their lateral (gluteal) surfaces are very concave, 

 and they seem to have been widened out anteriorly as in the pelves 

 of Phalacrocorax and Ploius. 



Just in front of the acetabulum the dorsal edges of the ilia 

 diverge one from another and the whole pehis increases in width, 



^ This bird is rpgarded by most authors as an undoubted Steganopod, but 

 in his original description Owen pointed out some points of resemblance witli 

 the Anserine birds. I have lately cleared the matrix from the orbit, quadrate, 

 and ptei-ygoids of the type specimen, and the new characters thus revealed 

 point rather strongly to Anserine affinities: for instance, tiie ftrm of the 

 pterygoids is extremely duck-like, and tbey articulate by broad subcircular 

 surfaces, situated at their anterior ends, with corresponding facets near the 

 base of the rostrum ; many of the Steganopod-like characters, however, are of 

 considerable importance. I hope shortly to publish a note on Odonttyptcryx 

 with figures of the quadrate and pterygoids. 



