366 Brewster on Robin Roosts. [October 



the fluctuations under consideration is the Robin population of 

 the country along the boundaries between two contiguous roost- 

 ing areas. Why may not this constitute a neutral ground, its 

 inhabitants sometimes visiting one roost, sometimes the other, 

 as fancy or the conditions of wind and weather dictate? This 

 suggestion, however, is pure speculation unsupported by any 

 data.* 



After the middle of September the roosting flights diminish 

 rapidly, and by the end of the first week in October the roosts are 

 practically deserted. The latest date in my possession at which 

 any Robins have been actually found in a roost is Oct. 20, 1SS9. 

 when Mr. Faxon noticed a few still lingering at Beaver Brook, 

 but my notes record that on Nov. 6, iSSS, I saw a succession of 

 Hocks flying, at sunset, into these Beaver Brook woods which, 

 at the time, were "leafless"! About 200 Robins were seen on 

 this occasion. They were in unusually large flocks, one, which 

 passed me closely, containing fully 100 birds. If, as seems prob- 

 able, they were migrants from further north it is interesting that 

 they should have found their way to this roost; but perhaps 

 enough local birds were with or near them to serve as guides. 

 Mr. Faxon believes that our roosts receive some accessions from 

 the north as early as September. 



I had supposed that the old birds which accompany the young 

 in the earlier flights to the roosts are individuals of both sexes 

 that for one reason or another have not attempted to rear second 

 broods. But Mr. Faxon informs me that as far as he has seen 

 they are invariably males, and in disproof of the rest of my as- 

 sumption furnishes the following interesting bit of evidence. 

 The observations quoted were made in July, 1890, a little over 

 a mile from the Beaver Brook roost. 



* Since the above was written, I have definitely ascertained that over a belt of 

 country a mile or more in width lying about midway between the Beaver Brook and 

 Norton roosts the flight is directed sometimes to the former, sometimes to the latter 

 roost. On the evening of Sept. 6, 1890, the Beaver Brook roost drew not only the 

 entire Robin population from this neutral ground but also all the birds of an adjoin- 

 ing area which extended to within half a mile of, and had been before tributary to, 

 the Norton roost. A profusion of ripe rum cherries had attracted to this area an 

 unusual number of Robins. Probably the greater part of these belonged to the 

 Beaver Brook colony and the less numerous local birds followed their lead on the 

 return flight that evening. At least I can think of no other explanation of the 

 phenomenon — which was repeated on theeveningofSept.il. These observations 

 throw much light on the questions above raised and in the main appear to support 

 the theories which I have just advanced. 



