HOUSE-SPA.RROW. 91 



The whole of its outworks are composed of straw, hay or 

 dry grass, often intermingled with such shreds of manufac- 

 tured stuffs as may be in the way, while the interior is 

 profusely lined with feathers, and access thereto is gained by 

 a hole left in the side. When, from the locality selected, 

 the dome and outworks are not required, the amount of 

 vegetable matter used is much less, and when a very small 

 cavity is occupied perhaps only two or three straws may be 

 found among the feathers which seem indispensable. Indeed 

 so great is the bird's fondness for warmth that abundance of 

 feathers are used to line even a nest in the inner side of the 

 thick thatch of a barn, and it has been seen collecting them 

 in winter and carrying them to the hole in which it often 

 roosts in the company of its fellows. The first batch of 

 eggs usually consists of five or six, and two other sets are 

 frequently produced in the season. They are greenish- 

 white, blotched, spotted, streaked or suffused with ash-colour 

 and dusky brown, varying considerably in the quantity of 

 this secondary colouring matter : their length is from *95 to 

 •82, their breadth from '66 to *64, but an exceptionally small 

 egg will measure *75 by '55 in. 



Occasionally the Sparrow plays the invader's part and 

 seizes on the mud-built nest of the House-Martin, which 

 after vain show of resistance, has to yield possession to the 

 intruder, though cases are recorded in which the evicted owners 

 are said to have revenged themselves bywalling-up their enemy 

 alive, and leaving him to die — a prisoner in the domicile he 

 has violated.* This act of aggression is perhaps the only 

 charge against the Sparrow that can be maintained in an 



* The story is a very old one t but though instances of Sparrows turning out 

 Martins occur every year, evidence of the revenge said to be taken l>y the latin- 

 upon the former is most unsatisfactory. It is generally offered without even the 

 slight corroboration that would he afforded by information as to time, place or 

 observer — and the last, from the language used, would seldom seem to have been 

 a naturalist. Most of the instances, even in modern times, rest admittedly on 

 second-hand reports, as those given by Jesse (Gleanings, ser. '_', pp. 99, loo) and 

 Macgillivray (Br. B. iii. pp. 591, 592). The best authenticated perhaps is 

 that for which M. de Tarragon personally vouches (Rev. Zool. 1843, p. J524), 

 but this witness speaks of the aggressor being a "moiueau friquet " i.e. a Tree- 

 Sparrow, and this fact casts a shade of suspicion on his evidence. 



