42 THE ZOOLOGIST. 



the bodily displacements in space experienced on the outward 

 journey. 



An extensive literature has accumulated around these two 

 theories. Among birds, most of the experimental work has been 

 performed on Carrier Pigeons. The results, too numerous to be 

 detailed here, show that the experimenters, for the most part, 

 prefer the theory of a knowledge of landmarks. In this view 

 they are in general agreement with most of the workers on 

 insects. On the other hand, some of the most distinguished 

 students of avian migration have maintained the inadequacy of 

 the theory to account for the long oversea journeys of migrating 

 birds. And, it may be remarked, a sufficient number of trials 

 has been recorded to show that Carrier Pigeons are able to 

 return home across the open sea, that is, without the help of a 

 knowledge of landmarks. The second theory also presents 

 difficulties under conditions in which the objective, or a land- 

 mark on the way to that point, is visible from the beginning of 

 the flight, and it is not a complete explanation of the mechanism 

 of long oversea journeys. 



It is convenient at this stage to propose a division of the 

 phenomena of homing flight into three classes on a negative 

 basis produced by the various considerations which have just 

 been resumed. By the allocation of a given flight to a particular 

 class, the subsequent search for the directive factors will be 

 facilitated. The divisions are as follows : — 



Class 1. In which at the time of departure visual perception 

 of the objective cannot be excluded. 



Class 2. In which visual perception of the objective may be 

 regarded as impossible, and in which the guiding influence of 

 land dispositions cannot be excluded. 



Class 3. In which there is no visible means of guidance of 

 any kind. 



Instances assignable to each class will occur to everyone. 

 The first and second classes invite the very evident explanatory 

 theory of a memory of localities. For the amazing flights 

 performed by birds, and coming within the third class, this 

 theory is obviously invalid. 



The directive mechanism readily admits of analysis into two 

 component parts. The first component is the means which 



