GEOMETRICAL SPIDERS 95 



obvious that it is feeling with its leg for the radius, 

 but finds nothing in its place. !t recognizes the loss, 

 ceases its work on the hub, runs out a new radius and 

 again resumes the spinning of the hub. The experi- 

 ment is repeated and more radii are divided. But the 

 spider replaces the lost lines so long as it is engaged 

 on the hub. It is thus clear that the time of hub- 

 formation is the time for testing a previous stage 

 in the construction ; the time to examine if every 

 radius is secure. 



This process of testing the radii is one of great 

 importance, for the absence of a radius means a loss of 

 symmetry in the web. Thus the spider is most dili- 

 gent in this stage of its duty. I will mention an experi- 

 ment to indicate its persistence in this important task. 



A snare of twenty-four radii was being reconstructed. 

 The old snare had been battered to fraofments ; and 

 the spider, having cleared the remnants, was building 

 a new structure in its place. It was laying out its 

 radii. Now as fast as the spider secured a radius, 

 I severed one that it had previously laid down. 

 As the spider worked on one side, I divided on the 

 other side. My destruction kept pace with its con- 

 struction of the radii. The spider continually found 

 the gap that I had made, a vacant interval where 

 there should have been a radius. But it was equal 

 to the emergency. It ran out a new radius. As fast 

 as the spider discovered the injury it remedied the 

 defect. In all I divided a radius twenty-five times 

 and the spider constructed new ones in their place. 

 Sometimes the same radius was severed three or four 

 times, and each time the structure was replaced. 

 After the twenty-five divisions the spider refused to 



