918 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. XXVIH. 
to the long spokes. I watch her in the ingenious work. She is moving across 
hoi lower and longer radii anchoring her filament as she goes. She reaches the 
shorter radii ; she reverses ; she comes back over her long radii again. Soon she 
approaches the short radii on the other side of the snare. But she will go no 
further ; again she reverses ; again she works back over hei long radii 
fi-ving her filament to the spokes. Thus she, so to speak, oscillates over 
her long radii; she largely neglects the upper and shorter ones; only on 
rare occasions does she make a complete circuit round her snare. So seldom 
indeed that, in this instance, there were 130 turns of the viscid spiral attached 
to the lower and longer radii, and only 3 turns to the upper and shorter spokes. 
Moreover, the Nephila employs this mechanism not only when dealing with her 
viscid spiral, but she introduces the same operation into her temporary .spiral 
too. And as with the viscid filament, she may sometimes do so to an extreme 
degree. In thi.s same .snare, for example, she attached 44 turns of her temporary 
spiral to the longer, and only 7 to the shorter spokes. Thus she is a spider which 
works with judgment and system. She can modifv her various lines to suit 
even the most eccentiic of eccentric snares. 
There is another structure in the arehitcctuiw of the Nephila w'hich I have 
not observed in ordinary snares. Xot only does she spread an extensive sheet, 
but she also constructs a special barrier in order to drive her %ictim more surely 
into the toils. She places it so as to face one surface of her .sheet ; it is at a slightly 
higher level than the main snare, and a little distance away. There is nothing 
precise or geometrical in its workmanship, it is merely a tangled maze of lines. 
They are poweidul filaments ; they are non-viseid ,• they m^et at all angles; they 
are spread about in every direction ; they form an entangled barrier through 
which no insect can pass. It is a strong and imixnetrahle silken wall, about 
two feet or more in depth, and secured by an anchorage of powerful cables above, 
below, and at the .sides. In some places its meshes are wide and ojxm, and 
strange insignificant spiders will then sometimes .seek a I’duge there. They 
rvill spin their own miniature but perfect webs amidst the entanglement of 
coar.ser lines. 
The object of the barrier is, I think, this. The main snare, though a vast 
sheet, is in texture rather fragile and thin. It is spread chiefly for the moths 
and flies that emerge from their concealment by night. It will not secure the 
larger prey. A locust will tear itself from the viscid grasp and easily break 
aw'ay. Thus without the aid of some additional resource many of the captures 
would certainly be lost. By means of the bax-rier this purpose is fulfilled. 
The victims w'hich break from the main snare strike the entanglement of 
lines ; they carmot advance ; they strive to escape, but are only driven back again 
into the snare. 
There is another feature in connection with this snare which it is worth our 
while to observe. The Araneus possesses the remarkable habit of swallowing 
completely the remains of her architecture as soon as it is worn out. It usually 
serves for twenty-four hours, after which time it is a tattered web. Then at 
evening we may see the spider demolishing all that is left. She eats up her 
successive radii, her viscid attachments, every little tag and shred, and all the 
tiny insects that happen to bo entangled on her lines. In fact she systematically 
destroys and takes back into he.’ owm substance all that remains of her imservice- 
able snare. 
I have little doubt that the Nephila behaves in a similar w'ay. Her architec- 
ture, of course, lasts a much longer time, and I have not actually seen the final 
desti-uction that occurs. But an experiment will indicate with sufficient certainty 
the manner in which the structure disappears. I sever the strong foundation 
lines all round an immense sheet. I leave just one attachment to the upper 
hoiizontal line. The snare collapses, its elastic sides fall in, its innumerable 
filaments gather themselves together, and, in the form of a sticky rope, it swings 
