In this paper we analyse the representation of the visual arts on the Internet through an examination of three main questions:

• How can multimedia technologies (re)present visual art in innovative ways that effectively and appropriately communicate meanings to the viewer/user?
• How can the language or languages of multimedia technologies be best used to communicate the language or languages of the target visual art works? Is it possible to identify a generic rhetoric of the web?
• What is the added value of representation through hypermedia over such traditional media as gallery space, art books, or non-digital media?

The first step consists in finding, analysing and reviewing a representative range of digital exhibition 'spaces' on the web. These might be both digitalised traditional art works such as are represented by classical art museum web sites (for example, Nicolas Pioch's WebLouvre) or art databases, and digitally generated art as represented in digital artist web sites. We attempt to identify a site typology and generic languages or codes through which the art works communicate meanings to the viewer/user, and to identify a rhetoric implicit in the media technologies themselves, through which hypermedia communicates/(re)presents the art. The second step has been to design and publish a web-based questionnaire (Annex A) targeted at a pilot user group of both experienced and naive users from a diverse range of backgrounds; the results are expected within 4 months. An analysis of the results will enable us to describe and document the ways in which naive and experienced web users conceptualise Cyberspace; specifically, what kind(s) of space Cyberspace is broadly conceived to be. This will result in a summary report useful as a reference guide for approaching questions (a) and (b) with a sound body of prior understandings. The final outcome will be a set of recommendations on the basis of questions (b) and (c), on how digital galleries/museums might be constructed on the web.