BIRDS. 



167 



from the south in a fairly distinct " wave." But they seem to 

 have a preference for a westerly trend in their dispersal. 



" Two summers ago," says the Eev. Mr. M'Connochie, " we had 

 two pairs of Kingfishers on two neighbouring streams, but this year 

 they have not reappeared. One pair nested for several years near 

 the mill here, and others have been seen, but are too often shot." 



It is reckoned rare in the north and east of Fife, but Mr. W. 

 Berwick has met with it on the banks of the Kenley Burn and 

 also on the Eden and Orr. 



Family CORACIID-a:. 



Coracias garrulus, L. Roller. 



Rare. Occasional visitant in autumn. Solitary. 



Col. Drummond Hay places on record two older occurrences — one 

 at Megginch, Carse of Gowrie, in 1823 or 1824, and one at Dunkeld 

 in 1882 {Proc. East of Scotland Scientific Societies, 1886, p. 34). Col. 

 Drummond Hay also speaks of having seen one many years previously 

 to 1874, also at Megginch, and it was called by the people "The 

 German Parrot," and it remained for some time in the neighbour- 

 hood. This may or may not refer to the one mentioned above. 

 Mr. Robert Gray notes the occurrence of one at Dunkeld {Birds of the 

 West of Scotland, p. 202). 



Next comes one at Ballinluig, obtained on 13th October 1903, as 

 first recorded in the Field (17th Oct. 1903, vol. cxi. p. 676).^ 



This Ballinluig specimen was shot by Lieut. B. C. A. Stewart, but 

 I do not know its destination. 



Family MEROPID^. 



[Merops apiaster, L. Common Bee-eater. 



, Rare, and very occasional visitant. 



One is entered in the Montrose Museum list as obtained at Laurieston, 



^ I give this particularly, because I find the record ver}' indefinitely quoted later in our 

 own Annals for October 190-1: (p. 210), which makes it difficult to be sure at first glance 

 whether the year of the capture was 1902 or 1903, the heading of the page giving 1903, 

 but the title of the article erroneously giviug 1902, while the date of issue of the number 

 was October 1904. In this case, no doubt, a mere technical overlook in revising proof, 

 but none the less puzzling, and necessitating reference to the cumbrous file of the new s- 

 paper. I think when quoting newspapers every facility should be given in similar cases 

 for direct reference to the original leading record, and so avoid confusion. All scientitic 

 records should give the dates of the occurrences in full. 



