ON THE STRUCTURE OF PHILEPITTA. 147 



The trachea is slightly laterally compressed below ; the rings, which 

 are complete, are somewhat irregular in shape, owing to the greater or 

 less development of the notchings on their borders. In front, of the 

 last few rings preceding the terminal one, two or more are joined 

 together by vertically directed bars, which makes it difficult to count 

 their number with exactitude. Behind, however, they are all free. The 

 terminal tracheal ring is narrow laterally, and closely opposed to the 

 first bronchial semirings ; in front and behind it is produced downwards 

 triangularly, and behind bears a well-developed forwardly directed narrow 

 pessulus. As seen from behind, therefore, the terminal tracheal ring 

 has somewhat the shape of an arrow-head. The narrow sterno- 

 tracheales are inserted on about the last ring but six. The first 

 bronchial semirings are thickened, and very much arched, being concave 

 downwards. As seen from the side (fig. 4) they are more strongly convex 

 anteriorly than posteriorly. The second and third semirings are very 

 slender indeed, closely approximate, much shorter and much less concave 

 downward, so that a large membranous fenestra is left between them and 

 the first semirings. The fourth and fifth semirings are also slender, but 

 less so than the last two : they are slightly concave upwards, so that a 

 second, though shallower, fenestra is formed between them and the P- z < S. 1880, 

 second and third pairs. The fifth semirings are slightly dilated at their p ' 

 extremities, where they are in close proximity, before and behind, with 

 the first four pairs. The sixth and succeeding bronchial rings take on the 

 ordinary character, being deeper than those that precede them, and 

 gradually becoming more complete, till the fifteenth pair are nearly 

 perfect. On one side, the left, in this specimen, the eighth and ninth 

 semirings are partially fused externally. 



The lateral muscles of the trachea, after the insertion of the sterno- 

 tracheales, become excessively thin, so that it is difficult to make out 

 accurately their exact extent. They apparently fan out, so as to be 

 nearly in contact with each other before and behind, and are inserted 

 on to the first bronchial semirings for the greater part (as far as I can 

 make out) of their lateral surface, the tips, however, being quite free 

 from muscular fibres. 



Philepitta is therefore perfectly Mesomyodian, as Prof. Garrod pre- 

 dicted would probably be the case *. 



There are thus three families of Mesomyodian Passeres in the Old 

 World the Pittidse, the Philepittidae, and the Eurylaemidse. All agree 

 in the possession of a broncho-tracheal syrinx, in that respect agreeing 

 with the Cotingidae (including Rupicola), Pipridse, and Tyrannidas of the 

 New "World, and differing from the Tracheophone families, which are 

 all, as is well known, American. Philepitta differs in the details of its 



* Voice-organs of Passeres, p. 68. 



L2 



