174 THE LESSON OF EVOLUTION 



thus trace out an improvement from the primitive 

 Psychodidae to the more recent, but still very 

 ancient, Tipulidae and Bibionidae. But after the 

 Bibionidse we seen no sign of improvement. No 

 doubt the later forms, such as the Syrphidae and 

 the Muscidse (restricted) are stronger flyers than the 

 earlier Empidae and Therevidae, but this seems to 

 depend more on the development of the muscles 

 than on the branching of the veins. This branching 

 looks like the result of accident or chance ; but each 

 change must have been caused by some agent, either 

 external or internal. 



If it is thought that even in the Diptera we must 

 suppose the existence of recognition-marks, we will 

 go to plants and thus eliminate all kinds of preferen- 

 tial selection. If the reader will take a descriptive 

 flora of any country and select any moderately large 

 genus, and then examine the characters which dis- 

 tinguish the species, I feel sure that he will come 

 to the conclusion that a large proportion of these 

 characters are not specially useful ; that is, one is not 

 more useful than another. And if this is so, these 

 characters could not have been formed by natural 

 selection. But these are the very characters relied 

 upon by botanists for separating the species, and 

 they have been chosen because they are the most 

 constant that can be found. 



Next, as to no two species having the same environ- 

 ment. The lower marine organisms, which float on 

 the sea, such as the Eadiolarians, and the Forami- 

 nifera, are known to be very variable. But the 

 Medusae, the Ctenophora, the Heteropoda, and the 



