xii Preface. 



adhering to it in numlDers, and that these insects might 

 then be digested and absorbed. But this I soon saw 

 could not be the case. The viscous matter was without 

 taste or smell, and clearly had no attractiveness to 

 insects, for none were ever to be seen buzzing about it. 

 Moreover, a few days' observation showed that the 

 adhering insects were in fact not digested, nor even 

 dissolved. Might not its use then be to keep insects 

 that were too small to effect cross-fertilisation, and 

 especially to keep ants, from crawling up the stem to 

 the flowers ? 



That it did as a matter of fact keep them off was 

 plain. The bank swarmed with ants, which were to be 

 seen running up and down many of the other flowers in 

 numbers. But not a single ant did I ever find climbing 

 up a lychnis-stem. If one was put on a sticky ring it 

 adhered there, and never succeeded in getting free. The 

 ants seemed to have found out by previous experience 

 how useless it was to try to mount this plant, and rarely 

 was one to be seen even on the part of the stem which 

 was below the undermost viscid ring. That the stem 

 should only be viscid in part and not throughout was 

 intelligible enough, on the principle of economy. But 

 why were not the viscid rings reduced on the same 

 principle to a single one ? Seeing how efBcacious and 

 utterly impassable the viscid matter was, it seemed as 

 though a single ring would do as well as a series. 



The following appeared to me to be a possible ex- 



