480 Dr. A. L. Thomson : Results of a Study of [Ibis, 



records for a given species come from Ireland and others 

 from Portugal the former country must lie iu the route of 

 the birds which travel to the latter. 



The most important unknown factor, from a statistical 

 point of view, is the actual mortality rate and its geo- 

 graphical and seasonal incidence. The student of the 

 method; unfortunately, deals only with what may be called 

 the " recorded mortality," and he can only attempt to guess 

 what fraction of the whole it represents. Furthermore, the 

 proportion between true and recorded mortality must vary 

 in different circumstances according to the chances that 

 exist of a dead bird being reported. It may be siifely 

 assumed, for instance, that a smaller proportion of actual 

 deaths is reported from foreign countries than from the 

 British Isles, while the fact that birds tend to meet death 

 in different forms at different times of year, especially in 

 the case of species shot for sport, may materially influence 

 the matter. Even were it possible to estimate the true 

 mortality, this would not afford a perfect index of the com- 

 parative total numbers of birds present in particular areas 

 at particular times. 



The foregoing considerations apply mainly to records 

 which show actual movements, for it is true that iu these 

 cases a record usually originates through the death of the 

 bird. But a similar and even more uncertain factor exists 

 in the cases of more or less sedentary birds which are 

 re-caught, often many times, at the places where they were 

 marked. Ilecords of this latter kind depend largely on the 

 activity of the marker himself, who continues to trap birds 

 for furtlier marking and in the process constantly recovers 

 birds he had previously released. Not only may the 

 marker's efforts be erratic for personal reasons, but he will 

 be largely governed by the greater facility with which birds 

 are caught at some seasons as compared with others. An 

 uncertain factor is therefore introduced into the marker's 

 o.vn records of birds recovered, while a much greater one 

 must be allowed for if these records are compared with those 



