428 THE ROLLER 



Africa it is quite unknown as a breeding species, though large numbers 

 pass through on passage to Palestine and Asia Minor. No doubt 

 many birds also diverge eastward on their way through Arabia to the 

 highlands of Persia, Afghanistan, and Transcaspia. 



There is some difference of opinion as to the way in which the 

 roller migrates. It evidently moves by day, and some observers 

 speak of large flocks consisting of some hundreds, while others de- 

 scribe it as passing in small flocks or pairs. Probably these apparent 

 discrepancies are due to local conditions. Thus in Egypt it is 

 generally noticed either singly or in small parties according to von 

 Heuglin and Shelley, but in Palestine it arrives in large flocks, which 

 very gradually disperse themselves over the country. Tristram gives 

 a graphic description of the behaviour of these newly arrived migrants. 

 " For several successive evenings great flocks of rollers mustered shortly 

 before sunset on some dom trees near the fountain, with all the noise, 

 but without the decorum, of rooks. After a volley of discordant 

 screams ... a few birds would start from their perch, and commence 

 a series of somersaults overhead, somewhat after the fashion of 

 tumbler pigeons. In a moment or two they would be followed by the 

 whole flock, and these gambols would be repeated for a dozen times 

 or more. In about a week the immigrants dispersed . . . after this 

 dispersal not a roller ever came back to the dom trees where they had 

 roosted at first." It is interesting to see how they adapt themselves 

 to their surroundings. The telegraph wire is at once used as a sub- 

 stitute for the dead branch : in the valley of the Nile the stumps of 

 the durra, which are left standing in the fields to the height of some 

 feet, are freely used as perching-places, while Captain Sperling saw 

 them seated on the slender reeds in the swamps on the treeless plain 

 of Sharon, and Tristram watched them perched on knobs of gravel or 

 marl in the barren Ghor by the Dead Sea, waiting for the emergence 

 of beetles from the sand. 



The distribution of this species during the breeding season in 

 Europe and Asia has already been indicated in the " Classified Notes," 



