STARLING. 237 



some thorn- trees in the Zoological Garden at Dublin, and 

 that this enormous estimate was the result of many observa- 

 tions. When the birds were first noticed their number was 

 put at from 15,000 to 20,000, but within three years it 

 seemed to have increased tenfold.* Accounts of two other 

 very large Irish Starling-roosts, at Lough Fern in Donegal, 

 and at Doohyle Lough, co. Limerick, were communicated to 

 the Dublin Natural-History Society (Jan. 8th, and Feb. 5th, 

 1858) by Mr. Montgomery and Mr. G. H. Kinahan. 



Though the ordinary flight of Starlings is very swift and 

 well- sustained,! it would not need special remark, were it 

 not for the wonderful performances of the multitudes, whose 

 congregations have just been noticed, prior to their going to 

 roost for the night, and occasionally during the day by smaller 

 bodies, especially when disturbed by a Hawk, but sometimes, 

 particularly in early spring, apparently for mere joy. Pos- 

 sessing very considerable powers of wing these are turned 

 to account in an extraordinary manner by the birds com- 

 posing the flock. They wheel, close, open out, rise and 

 descend, as if each were obeying a commander, and all this 

 is done with the most marvellous precision while the flock 

 is proceeding at a rapid pace high in air. At times it may 

 extend in a long and nearly straight thread ; suddenly an 

 undulation is visible along the line, and, in a moment, it 

 takes the form of a thin and smoke-like cloud : another 

 moment, and it is a dense and almost perfect globe 1 ; then, 

 possibly having preserved this appearance for a perceptibly 

 longer time, it becomes pear-shaped and, in another 

 instant, assumes a spiral figure ; an instant after, it has 

 spread out like a sheet, and its members are seen streaming 

 softly along the ground, perhaps to alight or perhaps once 



* Mr. Ball's description, originally published in 'Saunders' Newsletter' (March 

 25th, 1845), has been reprinted in full by Thompson. Mr. More (1877) informs 

 the Editor that for the last four years the Starlings have not resorted to this 

 roost, and that for some time before they used to come only at uncertain inter- 

 vals, absenting themselves for perhaps two or three years and then returning. 



t Gilbert White says " Starlings as it were swim along." 



J Sturnorum generi proprium catervatim volare, et quodam pilae orbe cir- 

 cumagi, omnibus in medium aginen tendentibus. C. Plinii Nat. Hist. x. 24. 



