EVOLUTION xxxvii 



instead of hair. When a bird flies, it fills its lungs with air 

 " in order to increase their volume and make itself Hghter." 

 But the lung being warm causes the air to expand, and 

 piercing through the lung, penetrate every region of the 

 body, including the bones and hair. The result of the 

 expansion of air within the bird's hair is to make it blossom 

 forth into feathers. Lamarck did not fail to see the diffi- 

 culties of his theory. Why have bats not got feathers ? 

 Because they have a complete diaphragm, which prevents 

 the swelling and piercing of the lungs. Why then have 

 flying insects not got feathers ? Each difficulty is met by 

 Lamarck with unshaken confidence in his hypothesis. 



Lamarck did not believe in spontaneous generation, 

 except in the most elementary animals, but he held that 

 among them it is always in progress. Like much else in 

 his work, this is a deduction based on false premises. All 

 infusorians, he says, die in cold weather : they are much 

 too delicate either to survive or to leave any spores or ger- 

 minal material that could last through a winter. Yet, on 

 the return of warm weather, we find them swarming in 

 stagnant water and other places : hence they are spon- 

 taneously generated. ] 



The progress made towards " perfection " of organisa- 

 tion as we advance along the animal scale must be 

 understood to mean a progress in the direction of human 

 organisation. For Lamarck the organisation of man is the 

 type of perfection ; and perfection or imperfection of 

 organisation is judged by its approach towards or depar- 

 ture from the human. Lamarck recognised, however, that 

 a high organisation was characterised by the concentra- 

 tion of organs and functions in special places, whereas in a 

 lower organisation they tend to be more generalised through- 

 out the body. He thus anticipated von Baer's famous law, 

 that development proceeds from the homogeneous to the 

 heterogeneous, upon which was founded one of the most 

 important clauses in Spencer's formula of evolution. 



It must have been quite obvious from the first that the 



