

round, strongly muscular, lined internally with dark cuticle, and contained no insects 

 whatever, but only hard seeds and pulpy masses. This indicates that Phceornis lives 

 on stony fruit and soft berries, especially since the stones or hard kernels are 

 also found in the gut, and are consequently passed out through the vent, a habit 

 common among Thrushes. The whole gut is correspondingly wide, especially the 

 sacculated duodenum and the rectum, which again is characteristic of frugivorous 

 birds. The caeca are narrow non-functional tubes 0*8 centim. long. The total length 

 of the gut is only 21 centim., the relative length only 3'5. This shortness, again, 

 indicates soft and easily digestible food. The intestinal convolutions are very simple, 

 as in most Oscines with short guts. The right lobe of the liver is three times as 

 large as the left. 



Summary. — Phceornis has no resemblance either to the Muscicapidae or with the 

 Pachycephalinae, as these groups are limited in the Catalogue of Birds in the British 

 Museum ; its supposed affinity with Eopsaltria can be disregarded, because that genus 

 is intermediate between the two groups. There remains the question of its being- 

 related to the Prionopidae, as suggested by Mr. Sharpe. Of the genera which he (Cat. 

 Birds Brit. Mus. vols. iii. and iv.) has made to constitute the Prionopidae, only those of 

 Australian and Malayan origin can be taken into consideration for comparison. Of these 

 Grallina is out of the question ; there remain consequently Pedes, Collyriocincla, and 

 Pinarolestes. According to the key (vol. iii. pp. 270, 271), Phceornis would coincide 

 with Collyriocincla, a genus which Gray associated with the Pachycephalinae, and of 

 which the questionable species " sandwicliensis " (Hand-list of Birds, no. 5832) is pro- 

 bably our Phceornis ohscura. Unfortunately only skins and skeletons of Collyriocincla, 

 with none of the soft parts, could be examined owing to want of material. However, 

 Collyriocincla is not a Pachycephaline bird, and it also differs considerably from 

 Phceornis in its strong transverse scales on the front of the metatarsus, in shape of bill, 

 formation of the tip of the wing, and above all in the bones of the palate, which 

 bear no resemblance to those of that species. Moreover, the three Prionopine genera 

 in question agree with each other and differ from Phceornis in the much more basal, 

 round, and almost concealed nostrils. Lastly, according to Gould, Collyriocincla 

 " feeds upon insects of various kinds, caterpillars and their larvae," while Phceornis is 

 essentially a vegetable eater, a frugivorous bird. 



"Which of the still-existing birds form the link between Turdidae, Muscicapidae, and 

 the ill-defined Prionopidae we do not know, owing to the want of well-preserved spirit- 

 specimens. Phceornis agrees fully with none of these families, but it differs least from 

 the Turdidse. Its frugivorous habits are much in. favour of its affinity to the Thrushes, 

 although many of these birds vary their diet with snails and other soft-bodied inverte- 

 brates. There are many instances known in which birds that originally fed on vegetables 

 have changed into insect eaters ; but the reverse, the change from essentially animal to 

 vegetable food, implies much more serious changes of the alimentary system. The 

 apparent scarcity of Hawaiian insects, or rather the extremely hidden life which many of 

 them seem to lead, has forced the more indigenous insectivorous birds, the Drepanididee, 



g2 



