ISOLATION , 173 



derate over the other. Neither can they be disad- 

 vantageous, for their owners are very abundant. 

 These hairs, therefore, must be indifferent, and could 

 not have been developed by natural selection. 



Now as to the variability of useless characters. 



In the vertebrates it is often difficult to feel sure 

 that apparently useless characters may not be recog- 

 nition marks of ancestors. So I will take the vena- 

 tion of the wing in the Diptera as an example. 

 Here we find in the numbers and branchings of 

 the veins some of the characters which separate the 

 families, genera, and sometimes even the species. 

 These are excellent characters from the systema- 

 tist's point of view, for they are remarkably con- 

 stant ; and yet we cannot explain them by the 

 principle of utility. For one kind of branching 

 does not seem to be more useful than another. In 

 this respect the wings of the Diptera are quite 

 different from those of birds, in which we easily 

 recognise a connection between shape and use. We 

 may find a character in a fly's wing which we 

 think to be of some special use, such as the 

 strengthening of the posterior margin in the Syrph- 

 idae, but we find that the house-fly manages 

 equally well without it. Or if we point to the up- 

 ward bend of the fourth longitudinal vein in the 

 house-fly as strengthening the tip of the wing, we 

 find that this is absent in the very common Muscina, 

 which seems to fly quite as well as the others. 

 Again, so constant and yet so varied is the venation 

 that we can construct a tolerably accurate genea- 

 logical tree of the Diptera by its means. We can 



