892 NATURAL HISTORY OF BIRDS. 



was usuallj' perched on tlie upju'r liranelies of some tree, from which it made short 

 flights into the air in pursuit of insects. The first one I siiot was only winged, anil, 

 turning on its back, and uttering harsh screams, it fought most savagely with my dog. 

 It was a male; length eleven inches; irides dark brown ; legs, fi'et, and beak, scarlet ; 

 ]ihiniage greenish blue; head almost black; wings very i)rettily marked with blue an<l 

 black, each having on it a spot of very pale blue; patch on throat rich violet; beak 

 short, strong, and hooked at ti|>; gape ami eyes very large."' 



The following family, the Lei'Tosomatid-k, is not the least remarkable for the fact 

 that there is only one species living, inhabiting the wonderful il.idagascar, while 

 another is known as fossil from the tertiary deposits in France. As may be expected, 

 this last survival stands alone among his more modern contemi)oraries, and many 

 are, indeed, its peculiarities, though the relationship to the rollei-s is 

 unmistakable ; in fact no other living birds come nearer to it, and 

 several authors refer it unhesitatingly to the Coraciada;. There are 

 two external jieculiarities which may l>e very well seen in the illustra- 

 tion herewith presented, viz., the jiosition of the nostrils at the middle 

 of the beak, the curious development of the loral ]>luiiu's into a con- 

 spicuous tuft, and the ap])arently yoke-toed feet. I say a])])arently 

 yoke-toed, since the fourth toe is not ]«laced directly behind, as in 

 the cuckoos or wood-peckers, but laterally, rather more lieliinil than 

 in front. This disposition of the toes iniluced the older systematists 

 to j)l:ice the bird in (juestion among the Cuculiilie, l>ut I)r. P. L. Sclater 

 has shown that it differs considerably from the latter, and approaches 



1. -. . . ■' , , , . ,,11 Fig. !!>:;.— Pier\l..si» 



the rollers, uecent studies ot its mvologv show that it rcallv lielon<j5 of /..•/»/n«oiwi.....ii.r- 



, ,1 ,, ^ " " ' 1- ■ • c •. ' 'iM salviow;/). iKiwiler- 



here ; llie breastbone presents some jieeulianties ot its own. 1 lie down paubes ; «, 



pterylosis is remarkable. It possesses long after-shafts, and the dor- *" ^'"' 

 sal tract bifurcates between the shoulders, and there are two highly developed powder- 

 down patches, one on each side of the rump, as shown in the ajijiended figure. There 

 seems, consequently to be characters enough to warrant its recognition as type of a 

 distinct family. 



Our figure of Ijeptosomns discolor, or kiniinlio, as it is called by some of tlie 

 natives of Mad:igascar, only represents the male, which is glossy green, with co]ijierv 

 reflections on the back and crown, cinereous on sides of head, round the neck, and 

 below. The female, which by Keichenbach was made the type of a different genus, 

 and i>laced in a difTerent part of the system, is barred and spotted with blackish and 

 rufous brown. 



The French traveler, Alfred Gran<lidier, gives the following account of the habits 

 of the kirumbo. They " live in bands of ten or twelve individuals on the borders of 

 woods. As soon as one of these birds is knocked over with a shot, all the others 

 place themselves at a little distance off or hover round the hunter, so that sometimes 

 one may kill as many as ten in less than a quarter of im hour." About the roller-like 

 tumbling in the air, Messrs. Roch and Newton made the following observations: 

 " It has a peculiar habit of )>layiiig in the air for some time over the same jilace, 

 ascending almost iieriiendicularly, as it were by a jump, to a great height, and descend- 

 ing again in a curve nearly to the top of the frees, by almost closing its wings, at the 

 same time uttering a whistle so like an eagle's tli;it it was for a long time doubted by 

 us whether the bird that jierformed this wonderful freak was not a raptorial. How- 

 ever, after having several times watched it with our glasses, we satisfied ourselves that 



