go FLEAS, FLUKES AND CUCKOOS 



deduced from the following facts. Only a small number of species of 

 fleas are involved and these are widely scattered throughout the 

 families comprising the order as a whole. Host-specificity is less 

 marked than in the mammal fleas, and there are relatively few cases in 

 which geographical variation or the formation of subspecies has 

 occurred. There is also a lack of specialisation in the fleas themselves. 

 By this it is meant that bird fleas as a whole have not, from the struc- 

 tural point of view at any rate, diverged very far from the mammal 

 fleas from which they are derived. In all cases except the genera 

 Dasypsyllus and Mioctenopsylla one can point with confidence to the group 

 of mammal fleas from which they originated. This is one reason why 

 bird fleas are exceptionally interesting objects of study. The change 

 over from mammals to birds seems a difl^icult one and few species have 

 been able to take advantage of this large mass of potential hosts. There 

 are certain conditions which appear particularly important if success 

 is to be achieved in this direction. Out of the 55 or so bird fleas described 

 27 are from birds which return to the same nesting sites year after year, 

 19 are from ground- or hole-nesting birds, 9 are known only from 

 islands, and of the remaining species a large proportion parasitise 

 birds which use mud in the construction of their nests. A combination 

 of the first three conditions is of course the most favourable. The 

 opportunity for straggling from a mammal to a bird host, occurs more 

 frequently on the ground (Plate XXXIVa) and the conditions in these 

 nests are more suitable for the development of the larvae. Colonies of 

 sea birds return year after year to the same site and thus give any fleas 

 which may have succeeded in living on them for one season another 

 chance, and again another chance, to consolidate their position and 

 multiply. Islands — particularly oceanic islands — afford the degree of 

 isolation which favours species formation. Thus we find that a rabbit 

 flea has succeeded in establishing itself twice on sea birds, once on 

 puffins and shearwaters on the rocky islands off* the west coast of 

 Britain (Plate XI and Map 4), and again on the other side of the 

 world on an auklet {Ptychorhamphus aleuticus) on Goronados Isle, Gulf of 

 Galifornia. On the mainland of Britain shelduck, for example, nest in 

 burrows, and are frequently attacked by hungry rabbit fleas which may 

 even be found attached to their heads, yet no shelduck flea has been 

 evolved. The factor which is missing is almost certainly prolonged 

 isolation in a relatively restricted area. On the Ganary and Pityuse 

 Isles a shearwater has acquired a flea of the genus Xenopsylla — the most 



