SUNDEW AND BUTTEKWOET. 133 



live flies ? Clearly, there must be some good reason for 

 the practice : the more so as all other insect- eating 

 plants Venus's fly-traps, side-saddle flowers, pitcher- 

 plants, bladderworts, and so forth are invariably deni- 

 zens of damp watery places, rooting as a rule in moist 

 moss or decaying loose vegetation. Now, in such situa- 

 tions it is difficult or impossible for them to obtain those 

 materials from the soil which are usually supplied by 

 constant relays of animal manure ; and under such cir- 

 cumstances, where the roots have no access to decaying 

 animal matter, those plants would flourish best which 

 most utilized every scrap of such matter that happened 

 to fall upon their open leaves. At first, we may feel 

 pretty sure, the leaves would only catch dead flies which 

 accidentally dropped upon their surface ; or they might 

 begin by being descended from slightly viscid ancestors, 

 which had acquired their stickiness to prevent ants and 

 other intruders from climbing up the stalk an explana- 

 tion especially probable in the case of the sundew, seeing 

 that its parent form was almost certainly a saxifrage like 

 the common little London pride ; and these saxifrages are 

 all noticeable for their very sticky glandular stems and 

 dotted leaves. If any such plant, growing in peaty 

 spots, occasionally by mere accident caught flies, which 

 decayed on the surface of its leaves and so supplied it 

 with a little stock of manure, it would benefit by the 

 habit thus initiated ; and natural selection would tend to 

 increase and specialize that habit in the future. So 

 there would slowly be evolved the long glandular tenta- 

 cles, followed by the actual development of a true diges- 

 tive absorbent system, and at last of something closely 

 resembling a set of nerves, to enable the arms to close in 

 immediately upon the struggling prey. 



Butterwort, on the other hand, began by being a sort 



