CORVINE PARADISIC. 75 



on the other parts of the spinal tract, the lumbar tracts, the apices of the arm-remiges, and perhaps 

 also those of the rectrices : I have, however, been unable to find them anywhere on the inferior 

 tract of very young nestlings. 



On examining the pterylographic differences, to the description of which I now pass, it will be 

 soon seen that they cannot here, any more than among the Rapacious Birds, be used in the forma- 

 tion of a natural classification, as the most nearly allied genera vary in the form of the saddle and 

 of the pectoral tract, the only parts of the pterylosis which furnish differential characters, although 

 this is not frequently the case in the species of the same genus. Here, therefore, I follow that 

 arrangement of the genera which I regard as the most natural, and describe under each those 

 peculiarities and variations which appear to me to be deserving of notice. 



]. CORVINE. 



All the members of this family examined by me have the saddle of the spinal tract broad and 

 laterally acute-angled, enclosing an elongated, fissure-like space. I find nineteen, and in rare 

 instances twenty remiges, of which ten are on the hand ; of these the first three are abbreviated, and 

 the fourth and fifth usually the longest. In Corvus fuliyinosus and C. azureus, however, which possess 

 twenty remiges, the fifth and sixth, and probably also the seventh, are longer than any of the 

 rest, and the first four are graduated. Moreover, the dilatation of the pectoral band is separated 

 from the main stem for a considerable distance at the end, and the space is removed somewhat 

 nearer to the posterior shorter part of the saddle, but enclosed only by the biserial arms of the 

 rump-tract, which is at first four, subsequently five, and at last six feathers in breadth. The 

 rectrices are iisuaiiy of equal length, but are graduated in Corvi pica, vagabundus,fuli(/inosus, and 

 azureKa. 



I find precisely the same structure of the tracts as in Corvus in Glaucopis leucoptera and 

 G. cinerea, which agree most closely in the form of the wings with C. fuliyinosus. 



In Glaucopis varians TEMM. (Phrenotrix temia HORSF.) I have not only found eighteen 

 remiges, of which only nine are on the hand, and the fifth and sixth the longest, but, what is much 

 more remarkable, only ten very much graduated rectrices, of which the two middle ones are 

 dilated at the apex. The most careful examination showed no gaps in the tail. Moreover, the 

 different feathering of the nasal grooves not only indicates that this species should be separated 

 from Glaucopis, but even that it should be united with the following group. 



2. PARADISIC. 

 A. Without an ephippial space. 



\. Paradisca apoda (PI. Ill, fig. 13). Saddle widely rhombic, acute-angled; inferior tract 

 with no separation of the dilated portion, but remarkable in the male from the approximated large 

 shafts of the .ornamental feathers, which are rooted only in this part of the inferior tract, but are 

 surrounded by no softer feathers. Twenty remiges, ten on the hand ; the first five graduated, 

 the sixth the longest. 



