THE GEOMETRICAL SNARE 119 



duced by an external strain. Never does a flaccid 

 filament hang loosely from the spinnerets. May not 

 the tension which is ever present be present of 

 necessity ? 



I have dwelt so long on the method of construction 

 that I will say but little on the destruction of the snare. 

 Yet nothing in the architecture of the spider so sur- 

 prised me as the manner in which it disappears. 



I first observed the process in a true home of the 

 Epeirida. Where the water falls down over a ledge 

 of rock and splashes in a transparent pool the spiders 

 love to construct their snares. Amidst the noise of 

 the falling waters the smaller flies revel in the cascade. 

 Above the pool they hover and dash at intervals 

 amongst the dancing drops that leap from the quiver- 

 ing pool high into the air. It is strange that they 

 should choose the commotion of the waters. It may 

 be for pure pleasure that they dart through the glitter- 

 ing drops and flash over the tiny foam. Where the 

 flies hover, there the spider will weave its snare. At 

 the very edge of the fall the slender lines extend, and 

 look as though they would be torn to pieces in the 

 spray. But the silken threads are strong, and, as the 

 sparkling drops ascend, each line in the slender fabric 

 is spangled with a row of pearls. 



I often visited this pool at sunset to watch the colony 

 at work. The snares were then renewed. After twenty- 

 four hours only the tattered fragments remain ; a 

 new snare was spun every evening. One evening on 

 coming to the pool, I happened to look on a tattered 

 snare with the spider resting at the centre. The snare 

 must have done duty for twenty-four hours ; it was 

 very much broken, and attached to its meshes were 



