44 r. ir. caitpbell — obseevations on spidees. 



touched the first spiral. The web is now finished,* unless, as is 

 often the case, a thread is carried from its centre to a leaf or other 

 fixed object, concealed in which the owner may wait in ease, with 

 one foot on the communicating line, ready to run out on its slightest 

 movement. If now we touch the permanent spiral, we find it very 

 adhesive and elastic, which is not the case with the radii and the 

 other supports of the web. The difference between the two on 

 microscopic examination is seen to be the presence in the former of 

 many small globules, which are found to be very viscid. The 

 spider invariably runs along the radii to its prey, and the cause of 

 its leaving in the centre of the web a few turns of the first spiral, 

 which is non-adhesive, seems to be that it makes that part a genei'al 

 resting-place, and does not wish to be hampered in its movements. 

 In some cases the spiral, as it approaches the circumference, is 

 much more open than at other parts, and this occurs more often iu 

 large webs. It would appear that this is an error, for it occurs 

 where the radii are the farthest apart, and at the weakest points of 

 the whole structure. This objection is however removed if we 

 consider that it would be to the advantage of the spider to allow a 

 powerful insect to break through the web rather than it should be 

 destroyed by the captive's struggles for freedom, and, were it 

 stronger at the circumference, such would be the result in a shorter 

 time than would allow the spider a chance of a personal encounter. 

 If large insects are caught, it is only near the centre, within easy 

 striking distance, and even then the Epeiridae will sometimes bite 

 asunder the threads which hold them rather than expose their web 

 to any risk. 



Adhesiveness is, however, not a property of the webs of all 

 genera. In these cases the prey is secured by entanglement and 

 immediate attack. The snares of the Ciniflonidae, which have on 

 the metatarsal joint of each posterior leg two rows of curved 

 movable spines, with which they crimp the thread as they leave 

 the spinnerets (calamistrum), are adhesive when newly made, and 

 there are none of the viscid globules already referred to. Mr. 

 Blackwall has treated of the matter, f and points out that the 

 application of a polished surface does not derange the web. This, 

 as well as other considerations, leads to the conjecture that the 

 detention of insects is due to some of the threads of the webs of 

 this genus being fibrous. Many spiders' webs are in connexion 

 with holes and crevices in the earth, and there the spiders hyber- 

 nate. Some of the Lycosidse dig an irregular oval cavity (about 

 one inch by half an inch), close it with their thread, and eai-th, 

 and remain there with a cocoon. I kept one in confinement, and 

 twice destroyed her retreat, only to find another made the fol- 

 lowing morning. The top was covered with granular pieces of 

 earth, such as might be surmised to have been raked over the silken 

 lining with her feet. A similar burrowing habit, but one that is 



* For the sake of brevity I omit here the very many stays and side -threads of 

 the web. 

 t ' Trans. Linnean Society,' vol. xvi. 



