4io 
Psyche 
[December 
may result in lines of extraordinary length, up to 2 m. The opposing 
lines from the upper side of the frame into the tangle often form 
a sheet in front of the lampshade part of the web. The spider oc- 
cupies the center of the lampshade in a typical posture (Fig. 5) and 
holds the sides of the lampshade near the base with its first 2 pairs 
of legs. Leg pairs III and IV hold the silk sheeting spun against 
the substrate. 
The distribution of dry and sticky silk in the web was determined 
by powdering the web with fine cornstarch (Fig. 1). Sticky silk is 
limited to the shade and its support lines, that is, while the frame 
line is not sticky, all the lines enclosed by it are. The sticky silk of 
H. thorelli has been described by Comstock (1940) as a hackled 
band consisting of a warp of four threads, the two outer ones much 
curled, and a broad woof with undulating edges, composed of sticky 
silk. This type of thread was found in a sample from the web of H. 
gertschi which also included single smooth lines on which an irregular 
band of sticky silk had been laid. 
Comstock (1940) observed some stages of web construction in the 
laboratory. He found that the first part of the web was the filmy 
disc of silk against the substrate. This was followed by construction 
of the lampshade. I attempted to investigate web construction by 
destroying established webs and observing the manner in which the 
spider replaced them. Webs destroyed in late afternoon were replaced 
the following morning, while those destroyed in the morning were 
not replaced for about 24 hours. When the web was removed, the 
spider ran a, short distance away on the rock surface. In all of the 1 1 
cases observed, the spider returned to the old web site to build a 
new web. Unfortunately, web construction takes place at night, and 
any light on the spiders halts their activities. 
Predatory behavior 
Remains from undisturbed webs indicate that the main items of 
prey are small Diptera. of various families, Tipulidae, and phalangids 
of the genus Leiobunum (primarily smaller, immature individuals). 
In addition to these animals, abundant at the study site, experimental 
prey were obtained by sweeping in an adjacent field. Most of the 
prey obtained by this method consisted of immature Hemiptera, Or- 
thoptera, and small beetles. 
The spiders were first tested for response to prey in the non-sticky 
tangle below the lampshade. In all cases, the response was the same 
as to that of a mild threat (see below). At no time did any of the 
individuals studied leave the lampshade to investigate potential prey 
