ANTS AND PLANTS 203 



which grew from definite centres or rosettes of an 

 appearance vividly recalling structures found in the 

 well-known fungus-gardens of the South American ants 

 of the genus A tta. An examination of the extremely 

 elaborate nature of these double pitchers, the presence 

 of a sweet substance in the inner pitcher, and the 

 peculiar fungoid growth on the inner wall of the outer 

 pitcher, point strongly to the view that the pitchers 

 are adapted for harbouring ants in addition to their 

 original functions of sheltering the rootlets and stor- 

 age of water. I say "original functions" because it 

 is plain that these pitchers, like those of D. rafflesiana, 

 are developed from the shell-like leaves of the pitcherless 

 species of the genus, but their development has been 

 carried even farther than that of the above-named 

 species. It seems highly probable that in both species 

 the plants benefit to a certain extent by their associa- 

 tion with ants on account of the soil brought into 

 the pitchers by these insects, for particles of soil are 

 found adhering in a perfecily normal manner to the 

 root-hairs. What is the function of the inner pitchers 

 in D. shelfordi ? Mr. Pearson has put forward the 

 ingenious suggestion that they serve as harbours of 

 refuge when the outer pitchers become nearly filled 

 with water after a shower of rain, the mouth of the 

 inner pitcher being so arranged that water cannot 

 enter into it until the outer pitcher is three-quarters 

 full. This theory could easily be put to the test, and 

 I did so by immersing a plant with several pitchers 

 growing from its stem in a bowl of water. As I 

 expected, the ants instantly came swarming out, and 

 by reason of their natural buoyancy came floating to 

 the surface of the water. The risk of ants drowning 



