xxvi] HERBERT SPENCER 25 



had had a little private conversation, we would go into the 

 drawing-room where there was music, and Spencer would 

 sometimes play on his flute. On remarking to him one day 

 that I wondered he could live among such unintellectual 

 people, he said that he had purposely chosen such a home in 

 order to avoid the mental excitement of too much interesting 

 conversation ; that he suffered greatly from insomnia, and 

 that he found that when his evenings were spent in common- 

 place conversation, hearing the news of the day or taking 

 part in a little music, he had a better chance of sleeping. 



In the autumn of 1867 I read the Duke of Argyll's 



"Reign of Law," and though I found much that was 



erroneous and weak in argument, I thought his discussion 



of the mode of flight in birds, founded largely on personal 



observation, was very good ; in fact, the best I had seen. 



Spencer had also read this, and differed from me, thinking 



that important parts of the duke's theory of flight was not 



true, because they would not apply equally to bats ; and we 



had quite a discussion on the subject. The next day, after 



thinking the matter over, I wrote him a long letter of eight 



pages, trying to show that the general principles of flight 



in birds, bats, and insects were the same ; but that in birds 



there were additional special adaptations that render their 



flight more perfect, and their power of motion through the 



air, under adverse conditions, more varied and more complete. 



The duke, dealing with birds only, had dwelt most on these 



ipecial adaptations (chiefly, if I remember, the beautiful 



)verlapping and movements of the separate feathers increas- 



ng resistance during the downward, and decreasing it during 



:he upward stroke) which did not exist in bats or in insects. 



,' also showed that although this adaptation was absent in 



he wings of insects, the general form and movements of the 



zings were similar and produced similar, but not identical 



esults. In his reply he admitted the accuracy of my 



1 description of the flight of insects, but made the following 



I emark in furtherance of his former objection as regards the 



f duke's account of the flight of birds : " If you will move 



\n outstretched wing backwards and forwards with equal 



