370 MR WATT ON THE HORI^TONTAL 
feet in diameter, and many of the spiders could not be less 
than half an inch broad across the abdomen. 
The long lines, which supported the centres of the webs, 
were fixed to the branches of the trees which were opposite 
to each other, and these hnes radiated regularly from the 
centre of the open spaces, at acute angles to one another. 
Now, I apprehend that it is impossible to conceive how the 
spiders disposed the long threads that sustained the central 
parts of the webs with any degree of regularity, without 
admitting that they must have the power of directing the 
lines they eject to a given spot. 
The manner of forming the ground- work or woof of their 
nets which I have observed, is this : After having made 
choice of a place (say between two trees), and in a calm 
evening, they dart a long thread from their spinners right 
across, from the tree they rest on, till it strikes the oppo- 
site tree. They then turn round and fix the other end of 
the line and run across, strengthening the first thread by 
drawing out a second line after them. Making this line 
sufficiently strong by going twice or thrice over it, they 
take their stand m the centre of it ; and from this point, 
they emit the other threads at acute angles to each other, 
in a similar way to the first ; fixing each of them near the 
centre of the first line, by pressing their spinners against it, 
and running up and down each, as it is formed, they 
tighten and strengthen it. 
The method of weaving the circular part of the web or 
warp has been often accurately described. 
It is well known that, if the web of the garden-spider is 
broken by any contingency, it does not attempt to mend it 
except the air be still, and that it generally chooses the 
twilight for this operation. If the spider does not mend 
its web in a calm evening, the country people consider it a 
1 
