378 PROF. A. H. GARROD ON THE [Apr. 1, 



behind, supported by the membranous walls of the windpipe. This 

 is evidently the atrophied fourth ring, counting from below. Above 

 this an abrupt change occurs ; the rings attain their ordinary depth, 

 with only linear intervals between them. The fifth ring, again 

 counting upwards, differs from those above it in being slightly in- 

 complete behind, with downturned ends. The interval between it 

 and the fourth is about equal to its own depth. It in front, and its 



Fig. 32. 



Fig. 33. 



Front view. 



G alius banMva. 



Back view. 



superior two or three neighbours behind as well, is slightly Y-shaped 

 in the middle line 1 . 



The Cracidae are particularly uniform in the manner in which the 

 trachea bifurcates. In Mitua tuberosa there are no tracheal inter- 

 annular intervals of any kind. The pessulus is united with the 

 penultimate ring posteriorly and with the last ring in front, the 

 latter ring being therefore incomplete behind, as in all the birds 

 above described. Mid-anterior and posterior ossifications extend 

 upwards from the attachments of the pessulus, generally sufficiently 

 high to involve the four lowermost rings, which are therefore conso- 

 lidated together in the median lines. The lower lateral borders of 

 the last tracheal ring are slightly concave downwards ; the medio- 

 anterior descending process being small, whilst by its slightly trun- 

 cated triangular apex it forms a small portion of the actual margin 

 of the bifurcation. On account of the considerable length of the 



1 By Temminck (Joe. cit. pi. ii. fig. 4) a different figure of the windpipe of 

 G. bankiva is given. 



