the Coloration of Eggs. 569 



lemur's preferences are relatively unimportant, and tlie rat's 

 might be explained (with a little inconsistency) as having 

 been, very naturally, for those eggs that he was most 

 accustomed to receive. And so on. And yet it is in- 

 teresting, and in harmony with the theory, that eggs which 

 possess for their first line of defence relative inaccessibility, 

 and particularly Sitagra's with the long-necked nest, should 

 have been placed quite high by both animals, while amongst 

 the eggs least liked by each, were some that are very 

 accessible in nature, as those of Macronyx (and often Pycno- 

 notus, &c.) to the rat, and Pycnonotus to the lemur. What 

 is really required, of course, whether for or against the 

 theory, is careful and critical field-observation on the lines 

 of the excellent work of some of our well-known bird- 

 watchers of to-day. 



Variability. — Variation in eggs is sometimes local. Thus 

 Sitagra ocularia lays only grey-spotted eggs at Chirinda, 

 but red-spotted eggs in some other South African localities. 

 West African Weavers lay some form of eggs that do not 

 seem to figure amongst the eggs of their South African 

 representatives (not that this is necessarily purely a geogra- 

 phical matter), and I was interested lately to see that the 

 forms of Prinia mystaced's, egg that are rarest at Chirinda 

 are apparently common in British East Africa. Instances 

 could be added from the Catalogue of Eggs. 



Variation in the same locality is mainly of two kinds : — 

 (1) definite, as already described for the eggs of many 

 Warblers and Weavers ; (2) indefinite. Selectionist expla- 

 nation would also be of two kinds. Variability may have 

 been selected because useful and necessary ; or (2) it 

 may merely have been tolerated as harmless, through a 

 slackening of the factors that make for uniformity. 



Definite selection factors making for variability might be 

 the need for the baflBing of Cuckoos (possibly an important 

 and somewhat widespread factor), the need for differen- 

 tiation by parents of their own eggs from those of their 

 neighbours, and procryptic adaptation in varying directions. 

 The existence of the second necessity is doubtful in the case 



