478 Br. A. L.T^homson: Mesults of a Studfj of [Ibis, 



accordance with the districts in which the birds were 

 marked, as is expLained more fnlly in Section II. : a dis- 

 tinction has also been maintained between birds marked 

 when young and birds obtained under other circumstances, 

 the latter class being further sub-divided under seasonal 

 headings. The object in view has been to keep together 

 records relating (o birds originally belonging to presumably 

 homogeneous groups, and in this way to eliminate errors 

 due to possible geographical differences. But where the 

 grouping has revealed no marked divergency, the separate 

 treatment of the groups is abandoned. 



The data thus grouped have been analysed, as a rule, in 

 three different ways. The principal analysis is a seasonal 

 one, the reappearance records of all birds of a particular 

 group, or series of similar groups, being classified according 

 to locality and calendar month. In this main analysis the 

 records of birds recovered in their first, second or subsequent 

 seasons, are treated alike : to exclude any error from this 

 source a second analysis is added in which the classification 

 is by localities and ages. The third analysis is similar to 

 the secondj but is based on the calendar year (reckoned 

 from summer to summer) instead of on the year of the 

 bird's life, and it thus affords a check on possible errors 

 due to meteorological differences between one winter and 

 another. The fulness with which the grouping and 

 analyses have been published will be found to vary with 

 the circumstances. 



A further section of the report contains a series of brief 

 summaries of the records relating to those species for which 

 the data are insufficient to warrant any more elaborate 

 treatment. IMany of these summaries, however, contain 

 more or less isolated records which are of some interest 

 despite the danger that lies in too much importance being 

 attached to occurrences which may possibly be exceptional. 

 The possibility of entirely exceptional individual movements 

 being recorded by the marking method is indeed a point 

 which must constantly be borne in mind by students of the 



