210 THE VOYAGE OF H.M.S. CHALLENGER, 



cuius of contractores which do not reach the syrinx, although undoubtedly on other 

 grounds that species must be grouped along with others as members of the genus Spheniscus. 

 With regard to the second point — the number of rings forming the syrinx — it is 

 necessary to state that I experienced considerable difficulty in determining the exact 

 number of these in different species. This difficulty arose from the fact that in 

 some genera, for instance Eudyptes and Spheniscus, the lower tracheal and upper 

 bronchial rings are so closely welded together as to render it almost impossible to esti- 

 mate the number of rings forming the syrinx, whde in others, such as Pyyosceles and 

 Aptenodytes, the rings forming the sjrrinx are scarcely more closely approximated than 

 those which enter into the formation of the trachea and bronchi, and are scarcely less 

 moveable upon one another than these. Hence in the last-named genera it is difficult 

 to determine where the syringeal rings terminate and the tracheal or bronchial rings 

 proper begin. I have therefore been compelled to estimate the number of syringeal rings 

 in these genera solely in accordance with the fact that certain of them in the region of 

 the lower larynx are relatively less moveable upon one another than in the region either 

 above or below, but do not profess to have estimated their number with absolute cer- 

 tainty. The succeeding table (pp. 211-213) gives the results of my observations, what- 

 ever their value may be. 



In respect of the position of the interbronchial segment or pessidus of the last tracheal 

 ring, I have noticed what I consider to be a generic characteristic. In the genus 

 Eudyptes the position of the interbronchial segment of that ring nearly corresponds to, 

 or in other words is parallel with, the long axis of the trachea, while in Spheniscus it 

 forms an angle with the axis of that tube. In Pygosceles, again, the pessulus maintains 

 an intermediate position between that of EudypAes and of Spheniscus, whUe in Apteno- 

 dytes the interbronchial segment of the last ring of the trachea more closely resembles that 

 of Spheniscus than of Eudyp>tes. It may farther be observed, that while in Eudyp>tes and 

 Spheniscus all the tracheal rings are to a greater or less extent ossified, and form rigid 

 structures, in A'ptenodytes and Pygosceles, on the other hand, none of the tracheal or 

 bronchial rings ever undergo ossification, and hence in these genera the windpipe is much 

 more flattened and compressible than in the others above referred to. 



With reference to the third point, the extent of the septum tracheae, the table on pp. 

 211-213 sho'ws firstly, that in the genus Aj^tenodytes the septum tracheae relatively to 

 the trachea is of larger size than in the other genera, with the exception of Spheniscus ; 

 secondly, that in Pygosceles the septum relatively to the trachea is of smaller size than in 

 Aptenodytes ; and thirdly, that Eudyptes as a genus possesses the smallest septum. The 

 genus Spheniscus differs from the others, in as much as the species composing it vary 

 to a greater extent, in so far as the tracheal septum is concerned, than do the species of 

 other genera. In Spheniscus demersus the tracheal septum relatively to the trachea is 

 of nearly the same magnitude as in Aptenodytes, in Spheniscus mendiculus the tracheal 



