262 MAY AGNES HOPKINS. 



Type III. (Fig. 3), Ardea, Plmlacrocorax, Ara, Anatidae. 



Type IV. (Fig. 4), Ciconia, Rallidae, Grits, Eudytcs, Psittacus. 



Intermediate between III. and IV. are the genera Pelecanus, 

 and Meleagris. 



Type V. (Fig. 5), Falco, Charadriidae, Colaptes, Trochihis, and 

 the Passeres. 



In the above, as in the results from the comparison of the 

 measurements, the types of canals bear no relation to locomotion ; 

 for different genera having very different modes of locomotion 

 have the same type of canal ; for example in type II. occurs 

 CatJiarista, one of the best flyers, and in the same group Gcococ- 

 cyx which seldom flies but is one of the ablest runners. 



4. EUSTACHIAN TUBE APERTURES. 



The Eustachian tubes have a common oesophageal opening in 

 all cases, except in the genera Stnithio, Endytes and Tinainns, 

 where there is a pair of apertures. 



The Eustachian tube is a perfect bony tube in the following : 

 Sterna, Squatarola, Ceryle, Psittaciis, Geococcyx, Tetraonidae, 

 Phasianidae, Cat/iarista, Falco, Endytes, and Tinamns. In all 

 other cases it is imperfectly ossified. 



CONCLUSIONS. 



From a careful consideration of all the preceding data, the fol- 

 lowing conclusions may be drawn : 



1. The anterior canal is always the largest. 



2. The exterior canal is larger than the posterior in the greater 

 number of cases. 



3. The relative measurements of the canals to each other, as 

 well as their sums, bear no direct relation to modes of locomotion. 



4. Relative dimensions certainly stand in no relation to broader 

 racial affinities and such dimensions can be used as taxonomic 

 characters only to limited extent. 



5. What occasions differences in these dimensions remains to 

 be determined. 



6. The anterior vertical canal is the one most subject to vari- 

 ation of position. 



