SEJilEMA. 119 



never even seen an efjfj^ whicli some say is blue, others lirown and speckled; and none 

 can answer the question as to wlietlier the young birds are, or arc not, able to run 

 from the moment of their being excluded. We have kept these birds for some time 

 in confinement, feeding them on the large BuUmi (which can be jiurchased in the 

 market, whither they arc brought as an article of diet for our French colons), raw 

 meat, etc. They are noisy at night, uttering a guttural rattling note, and their antics 

 of an evening have sometimes reminded us of the African Scopus umhretta." 



The kagu has been kept in captivity. The superintendent of the London Zoologi- 

 cal Garden thus describes some of its peculiarities: "With its crest erect and wings 

 spread out, the kagu runs or skips about, sometimes pursuing and driving before him 

 all the birds that are confined with him in the same aviary, and evidently enjoying the 

 fun of seeing them friglitened. At other times he will seize the end of his wing or 

 tail, and run round, holding it in his bill. From a ])iece of paper or dry leaf lie de- 

 rives much amusement, by tossing it about and running after it. During his frolic he 

 will thrust his bill into the ground and spread out his wings, kick his legs into the air, 

 and then tumVile about as if in a fit." 



From Madagascar, the wonder-land where once the Epiornis roamed abont, hails 

 the third member of this puzzling group, not the least puzzle of the three types com- 

 posing it. In addition to its strange structure, the rarity of the alleged tw'o species 

 in the museums, the types being unicjue for many years, made it the more ditticult for 

 ornithologists to find out the truth about it. Jfesites, the only constituent of the 

 family Mesitid.e, was therefore by some referred to the Gallinaceous birds, by others 

 to the pigeons, and still in 1^72 Snnilevall insisted upon the bird being oscininc, only 

 with the larger wing-coverts abnormal, and the tibiie naked at the lower end, " like 

 the waders." The anatomical investigations of A. Milne-Edwards finally enabled the 

 systematists to settle the question as to its relationships. He himself referred it to 

 the rails, but Garrod and Forbes have pointed out the schizorhinal character of J/i:;<i(es 

 as compared with the holorhinal rails, and the position with Eurypyga and lihynoche- 

 tos was confirmed when Mr. E. Bartlctt, son of the gentleman mentioned above, dis- 

 covered that it has powder-down ])atclK's. He first found two very distinct ])atches on 

 the back of the neck. This naturally led him to search for others, which he also 

 found, viz., one on e.ach side of the lower part of the rum]), close to the tail, one on 

 each side of the upper part of the pectoral muscles, a third pair, one on each side, 

 running across the ribs, on to the pectoral muscle, and a fourth pair, one on each side 

 of the abdominal region, running jiarallel with the vent, making in all five jiairs. He 

 also found that the feathers of the back and wings show the same peculiarity as those 

 of Kiinjii;/yn described above. To complete the characterization of Mesiti'.i, we may 

 aild that the bill is long and slender, like that of the sun-bittern, the nostrils are long 

 and linear, overhung by an elongated cutaneous operculum. The tibiai are naked 

 below, and the tarsus scutellated in front and behind, like the cotresponding parts in 

 the other two families. A noteworthy character is also the blue nakeil space round 

 the eye. The general coloration is cinnamon brown, head and under side more or 

 less marked with black in ^^. varicfjata. The whole external aspect is a<lmitted by 

 Bartlett to be " very thrush-like." 



The su])er-family CAKI.V.MOIDE/E we sh.all introduce with a remark of Profcs- 

 sor Newton, to the effect that he " is inclined to think that those who have urged ita 

 affinity to the Accipitres, and among them taxonomers starting from bases so oppo- 

 site ns Sundevall and Professor Parker have more nearly hit the mark, and accord- 



