ANATOMY OF THE SHOE-BILL. 685 



precisely similar in position and relations to the rounded tendon 

 under the biceps in Balceniceps. But for the fact that the 

 tendon stops short at the head of the fibula and does not pass 

 through the knee capsule to a normal ambiens muscle, it can- 

 not be distinguished from the ambiens head of the perforated 

 flexors. GaiTod (17) first called attention to the interest of the 

 ambiens muscle and regarded it as a major key to the classi- 

 fication of birds. He divided the Class into two Subclasses, the 

 Anomalogonatse, containing the Piciformes, Passeriformes, and 

 Cypseliformes in which the ambiens is never present, and the 

 Homalogonataj, containing all the other groups of birds and 

 showing that in them the ambiens was normally present. Among 

 the AnomalogonattB there is no species in which the ambiens has 

 been found ; among the Homalogonatae there are families and 

 genera in which it is absent, and Garrod believed that in 

 such cases it had been secondarily lost. In a much later con- 

 tribution to the subject (24), I showed that in the Night Heron 

 and in Edectus, birds without an ambiens but belonging to 

 Garrod's Homalogonata?, there existed an ambiens head to the 

 perforated flexors, absent in the Anomalogonatse, and plainly 

 suggesting that it was a remnant of the ambiens muscle. In 

 a memoir on the anatomy of the Hoatzin (27) I was able to 

 desciibe from dissections of different examples of that bird, a case 

 of this possible degeneration in actual progress. Garrod had 

 dissected both legs in three examples of the bird and in all cases 

 found the ambiens small but normal above the knees, but in five 

 out of the six legs it was lost at the knee-joint. He does not 

 appear to have had his attention called to the impoi'tance of the 

 ambiens head of the perforated flexors. I examined each leg in 

 two examples, and found in every case an ambiens head to the 

 perforated flexors, but the ambiens muscle in some instances 

 absent above the knee, in others small and lost at the knee-joint. 

 It may therefore be taken as established that the ambiens head 

 of the perfoi'ated flexors represents a vestige of a complete 

 ambiens muscle, and its existence in Balceniceps is of morpho- 

 logical rather than systematic importance. The ambiens is 

 usually present in the Steganopods, present in the Spoonbills, 

 jDresent in some genera of Storks absent in others, absent in 

 /Scopus, absent in Herons and Balceniceps, but in the last two 

 cases its recent loss is shewn by the existence of the vestige to 

 which I have now called attention. 



Flexor profundus or perforans and Flexor longtis hcdlucis. — The 

 deep flexor as in Storks and Herons arises by fleshy digitations 

 (text-fig. 129, Fl.Pro.) from the side of the fibula and from down 

 the shaft of the tibia to form a strong i-ound tendon. The flexor 

 longus hallucis comes from the inner surface of the outer condyle 

 of the femur (text-fig. 129, Fl.Hal.) and similarly forms a round 

 tendon. The two tendons pass down to the flexor surface of the 

 foot in the usual way. The deep flexor (text-fig. 131, B) breaks 

 up into a branch for digits 2, 3, 4, and the hallucis tendon, 



