thread ; it is largely liquid, but when it is highly magnified 

 irregular threads can be seen in it. 



It is easy to infer the function of these four kinds of silk : 

 the supporting line not only supports the parts fitted for entangling 

 the prey, but communicates to the centre of the web, where 

 the spider is lying in wait, any disturbance of the web ; the 

 primary looped threads also have two functions — they support 

 the secondary looped threads, and by their elasticity allow an 

 entangled insect to become involved in other threads (I have 

 seen these threads stretch to fifty times their first length) ; the 

 secondary looped threads support the viscid silk ; and the viscid 

 silk clings to anything that touches it. 



This remarkable hackled band does not diner so greatly 

 from that of Amaiirohins as would seem at first sight. The 

 supporting thread is a distinct structure, which is spun before 

 the band is made ; it merely supports the band proper, as the 

 dry threads in a dyctinid web support the hackled band of 

 those spiders. Omitting this supporting thread, there remain 

 the four strands of the warp and the viscid silk constituting the 

 woof. The primary looped strands of the warp correspond to 

 the straight strands in the band of Amanrobius, and the secondary 

 looped strands to the curled strands of Amaurohius. 



I at first inferred that the loops of the secondary looped 

 thread were made by the calamistrum ; but this now seems 

 very improbable to me. It is more probable that these loops, 

 like those of the curled thread in the band of Amaurohius, are 

 caused by the curly nature of the thread, and as each loop is 

 formed it is held in position by being attached to the corre- 

 sponding primary looped thread, and by being embedded in 

 viscid silk. 



The loops in the primary looped thread are not regular, and 

 are probably merely a result of this thread being slack. 



At regular intervals the hackled band is bunched in masses 

 upon the supporting thread. Three of these masses arc shown 

 in the figure. The formation of these masses I now believe to 

 be the result of the combing action of the calamistrum. By 

 the formation of these masses a greatly increased quantity of 

 the viscid silk is made available for trapping the prey of the 

 spider. 



2 



