THE UNUSUAL WEB OF SPILASMA TUBULOFACIENS. WITH TAXONOMIC NOTES ON THE SPECIES (ARANEAE ARANEIDAE) *

During December 972 I had the opportunity to visit French Guiana to study some o its little known amblypygid auna. While searching or amblypygids I came across the unusual web of an araneid, Spilasma tubulofaciens, and was able to make some observations in the qeld. I would like to acknowledge the gracious help and encouragement given by Dr. Herbert W. Levi, who also provided the laboratory acilities to complete this work. I would like also to thank Dr. J. J. de Granville for his hospitality and guidance in Cayenne, or the use o ORSTOM facilities in Safil, and for many other avors.

I would like to acknowledge the gracious help and encouragement given by Dr. Herbert W. Levi, who also provided the laboratory acilities to complete this work. I would like also to thank Dr. J. J. de Granville for his hospitality and guidance in Cayenne, or the use o ORSTOM facilities in Safil, and for many other avors.

Web of Spilasma tubulofaciens Figures 6-7
Spilasma tubulofaciens females were found in a forest patch clearing occupied with mixed growth of shrubs (mainly Rubiaceae and Leguminoseae), 2-3 m from the muddy banks of Crique Limonade, French Guiana. The suspension lines of the web were attached between leaves of a rubiaceous shrub, about 1.2-.4 m from the ground. In the middle they inserted at the apex of the conical retreat. The snare region was below, hanging obliquely, and had 6 radii and more than 28 turns, with no distinct frame threads. It showed numerous torn spaces and repairs, giving a general appearance of long use and disorganization. No major changes were introduced by the spiders during the night before they were collected next morning. The suspension lines were quite taut, forming nearly a straight horizontal line (Fig. 6). The outer surface of the conical retreat was rough, made up of diverse materials, predominantly remains of insect prey (small coleoptera, flies, etc.) and vegetable matter (protonema of mosses, small bark fragments, vegetable debris). This debris was irmly fixed to the surface by underlying, ine threads of silk. The inside of the retreat was smooth, lined with silk. The *Manuscript received by the editor June 2, 1974. broad, lower end of the retreat had a transverse, single opening with two lips, the posterior lip being the longer. The lining of the anterior lip appeared thinner and more flexible. The supporting lines were strong, reinforced with silk, and contained numerous moss fragments, particularly close to the insertion on the retreat. There were also remains of insects and vegetable matter. The spider remains inside the retreat, and waits head down or its prey, a characteristic position of the orb-weaving spiders. When I offered living ants, which I placed on the lower corner of the catching web, the emale dashe.d out ot? the retreat rapidly, wrapped its prey ( Fig. 9), bit it, and promptly returned to the retreat, without eating the prey.
Discussion. By changing the position ot the hub (coincident in this case with the retreat) to a peripheral, uppermost position, an additional load of tension has been placed on the suspension lines. Additional tension has also been brought to these lines by the taut, almost straight line, to which they have been drawn. Consequently the spider has carefully reinforced the lines with extra layers ot silk, enclosing different building materials, and they no longer function in prey capture but to support the rest of the web and the retreat.
It differs rom S. artifex Simon, which Simon, in   suspension line (Fig. 8) and having a catching web parallel to the floor, with an empty segment.
The numerous repairs in the web of S. tubulofaciens, the thickness of the suspension lines, and the elaborate construction o. the retreat all suggest that this web is long lasting (relative to the life span of the spider), and with time is repaired and modified. The retreat of the younger female had more space for occupancy than the retreat of the older female (as determined by body length as a percent of the length of the retreat), the space perhaps serves to accommodate additional growth of the spider without having to enlarge the retreat. Additional collections might verify this observation.
Nielsen (1928, pp. 534-535) has excellent photographs of Achaearanea saxatile (Koch), a theridiid, building a similar type of conical retreat on the ground and later scaffolding it to the desired height.
It is likely that a similar (energy saving?) strategy is used by S. tubulofaciens, instead of making numerous trips to carry and assemble the materials above the ground, although above the ground modifications and repairs might be introduced at a later date.
The adaptive convergence of webs has been discussed recently by several authors (see Kullmann, I972). Strangely, emphasis in their discussions has been placed on considering orb webs as indivisible entities which appear to have evolved as a unit. Although interactions must likely occur among the major structural components of the web (hub, retreat, and capture threads snare), my impression is that each has distinct adaptations to cope with in response to environmental and physiological cues (e.g. the retreat to best hide the spider). The selective forces which might bring about evolutionary changes in these components, by selecting particular motor patterns of the spider, must differ and thus these components have evolved as distinct characters, although not totally independent. Hingston left no type material for his new species and his descriptions were very general, lacking diagnostic characters (Levi, I963, p. 493). Nevertheless, Hingston included descriptions and detailed sketches of the webs of his new species, which may help to validate some ot? his names, in some instances. Lehtinen (I967, p. 2o) considered that the only way to treat the species described by this author would be to propose all of them as "nomina dubia." I am convinced that the specimens which I have collected belong to one species described by Hingston as E. tubulofaciens. I base my assertion mainly on the unique abdominal and leg coloration patterns, and the remarkable architecture of the retreat and the catching web, which matches Hingston's descriptions. Although there is the possibility of sibling species, I have decided for the sake of stability in nomenclature to retain Hingston's name.
Description. Carapace, chelicerae (except claws), pedipalps, and dorsal surface of legs brownish yellow, with the two posterior legs a darker brown. Inner margin of endites white. Sternum and labium reddish brown. Carapace covered with fine-pointed bristles which become longer toward the anterior margin. Carapace with a welldefined cephalic suture. Ventral coloration of the two first legs is reddish orange but this pigmentation washed away at?ter preserving the specimens in alcohol. Abdomen is longer than wide and slightly wider in its anterior half. Dorsum of abdomen with a broad, white H-shaped mark over black background, positioned transversely on the anterior half. Near its middle, the posterior bar of the H mark has an additional smaller triangular white mark pointing anteriorly and two dorsal black sclerites. On its posterior third, latero-ventrally, two elongated white marks occur, one on each side of the abdomen.
Venter with a light colored epigastrium and colulus. Anterior median eyes are largest and closely positioned, o.7 diameter apart. Posterior medians one diameter apart. Laterals closely set on a common raised black-pigmented carapace projection. Total length 4.9 mm.

Psyche
[June is only known from Equatorial Guinea. Hingston (932, p. 366) described a species, E. sacculifaciens, with similar conical retreat and web architecture to 8. tubulofaciens, which, I suspect, will eventually be ascribed to the genus Spilasma when new material is collected. Its reported size, .5 ram, suggests it might be 'an immature specimen as the other known species of Silasma range 'from 3 to, 6 mm in length.