WINTERING OF SWALLOW TRIBE IN S. HANTS. 269 



Last night's severe cold seems to have lowered the bird's vitality 

 almost to the lowest ebb, and it is more than likely that to-day 

 will be its last effort to survive ; the odds are very much 

 against its being able to stand a temperature as low as that of 

 last night. 



24th to 27th. — Visited the stream each day, but no signs 

 whatever of the Martin since 23rd ; it has without doubt suc- 

 cumbed, but it certainly deserved a better fate. 



Before finally dismissing these diary notes, there is one very 

 puzzling point which I am unable to settle, viz. where during the 

 time the Sand-Martins continued to remain in the district — i.e.. from 

 November to January — before finally perishing, did they roost or 

 conceal themselves at night ? 



As is well known, these birds live in holes or burrows during 

 their stay in this country, and sleep in the burrows at night, 

 never, as far as I have been able to discover, in any other situa- 

 tion. Immediately on arrival in spring the birds repair to their 

 nesting haunts, and use the old burrows for roosting purposes. 

 In the district around Havant I know of no place where Sand- 

 Martins breed within a distance of thirteen miles of the town, as 

 the geological formation is mainly chalk, which is quite unsuited 

 to the requirements of these birds for breeding purposes. Hence 

 it has puzzled me very much to know in what situations and 

 under what conditions the birds found shelter at night. Would 

 they be likely to roost in reed-beds, or take shelter in out- 

 buildings or stacks of hay or straw ? The observations and 

 criticisms of ornithologists on this point would be welcomed by 

 the writer. 



The following weather table serves to illustrate my remarks 

 and criticisms bearing on the subject of migration as dependent 

 on wind direction, which Mr. E. K. Robinson maintains so 

 strenuously is a vitally important factor in bird migration (see 

 table on next page). 



In the ' Country-Side ' for week ending Jan. 5th, 1907, 

 p. 109, the Editor, Mr. E. K. Robinson, to whom I submitted 

 some notes on the abnormally late departure of Swallows and 

 Martins, endeavours to explain the reason for the "prolonged 

 and untimely stay" of these birds in South Hants by saying 



