PREFACE TO THE SECOND EDITION. v 



or dentate ; feet webbed," it is at once seen that this will not do ; 

 "c". Cutting-edges of bill not fringed nor dentate, or else feet not 

 webbed," will do, however. It cannot belong to " d}" because the 

 legs are not " inserted far behind the middle of the body," etc., but it 

 agrees all right with " d'." The anterior toes are not webbed ; there- 

 fore it cannot belong to " eV' and search must be continued under 

 " e'." "P" will not do, for the thighs are not naked ; therefore 

 "y>" "y">" " /i\" and " /r," which follow successively, must be passed 

 over, and "y-" tried, which, being done, is found to agree. " g^" will 

 not answer, because the bill is not hooked, etc. ; therefore it must 

 come under "y^." This being determined, it is found that, " Ji}" being 

 inapplicable, " A^" fits. Then, the bill being " without a swollen 

 cere," it must come under " v'." It is next found that the wings are 

 "not very long and gape not very wide . . . and tail-feathers 12 in 

 number ;" the toes are not " only 2 in front," etc. ; therefore it must 

 be beyond " ¥." " F" is then tried, and found to fit ; the bird is 

 identified as a member of the Order Passekes, and the reader referred 

 to page 321 in order to determine the family to which it belongs. 

 Turning to page 321, as directed, it is found that the tarsus is cylin- 

 drical, etc. {a}), and that the inner toe has the "basal phalanx united 

 to that of the middle toe" {b^) ; therefore the bird belongs to the 

 Family Cotixgid^, which is further treated on page 323, where the 

 genera are analyzed. Seeking here for the genus, we find that, as the 

 nostrils are not " wholly exposed," it must belong under " a^." The 

 tip of the bill being only "very slightly hooked," it must belong to 

 " b^," and the tail being "graduated for only about half the length 

 of the exposed culmen," etc., it must belong to the Genus Platypsakis, 

 treated further on page 324. Comparison of the bird with the con- 

 densed description of generic coloration under the heading of " common 

 characters" removes all possible doubt as to its being a member of this 

 genus ; therefore it remains only to determine the species, for which 

 the key proves quite adequate. The bird fits exactly the characters 

 ascribed to " adult females and young males" under " a}." An ap- 



