284 
Psyche 
[December 
2. Disguise can be rejected because the structure is by no means 
conspicuous and, in any case, is separated from the spider by a 
wide gap. 
3. Deflection can be rejected because of this same lack of con- 
spicuousness (resulting from the wide dispersal of the opaque silk 
deposits). The same argument applies to advertisement. 
All these forms of defense operate visually and it is necessary to be 
cautious about conclusions concerning the visual acuity of predators. 
One of the lessons of ethological studies is that in many situations 
animals actually attend to relatively simple stimuli even when they 
are quite capable of discriminating more complex ones. There is an 
excellent treatment of this subject in Hinde (1970: 57-80). Despite 
this qualification, we feel that a visually-operating function for the 
skeleton-web stabilimenta of N. clavipes is highly improbable. 
The defensive function of reinforcing the hub against penetration 
by predators that strike through the web can be rejected because the 
silk is simply not deposited in the hub region. The structure could 
conceivably form a barrier to spider-hunting insects that might other- 
wise fly through the skeleton web and attack the spider from below. 
This function is improbable in view of the limited area covered by 
the device. 
A function as an insect attractant is highly improbable, not only 
because the device is inconspicuous but also because the web has no 
viscid spiral and cannot, therefore, trap prey. (In addition, the 
spiders do not feed in the period immediately prior to moulting.) 
Web protection is probably achieved by the barrier webs and this 
function of the stabilimentum can be rejected for this reason. The 
structure is not sufficiently dense to act as an effective sunshield and, 
furthermore, is in the wrong place. (Note that some araneids adopt 
complex postures, at times, in order to minimize heat absorption, 
Krakauer 1972, Robinson & Robinson 1973.) 
The fact that the stabilimentum occurs most commonly in skeleton 
webs could be correlated with a mechanical function. The laying 
down of zig-zags on those radii that are immediately below the point 
at which the moulting spider attaches itself to the hub may stabilize 
this attachment. This would ensure a secure moulting base within 
an area protected by the barrier webs. From this base the spider 
could safely go through the fairly vigorous process of withdrawing 
from the old cuticle. The moulting process involves much pulling 
and jerking and if the spider loses connection with the web, and 
drops to the ground, the result may be fatal. The reinforcement of 
