This study has been based on the previous research and exploration by numerous individuals. Although there have been several conclusions reached, they do not indicate that research in this field is final. More study needs to be undertaken in the following areas.

Multiple Senses
This thesis and the project which exemplified it relied on only the visual display of virtual architecture. In the physical world, one’s perception of architecture is strongly influenced by sound and touch, and less so by smell and taste. Interfaces and peripherals which display such stimuli are being developed in addition to visual display. It is suggested that one’s perception of architecture, and therefore the design of architecture would be affected by the added display to other human senses. More research needs to be done in this area.

Language
The use of language, as textual signage, as voice communication with the environment, or otherwise, is beyond the scope of this study. This is because language barriers among cultures who can access the project on-line would inhibit the success of this study if it depended on language to communicate itself to participants. Also, text does not display legibly in current head-mounted display technology. However, these barriers are likely to be overcome with continued development of software and display technology. It is believed that the use of oral and written language, while it serves a similar function in communicating ideas as does architecture, would enrich the experience of the virtual participant.

"Irrational" Hyperlinks
Hyperlink technology has been used in only a limited way in this project: there is a one-to-one mapping of the links and the worlds they connect. Hyperlink technology does not limit links between worlds such that a link from one world to the next must lead back to the first if entered again from the other side. In fact, link in a world could lead to a different virtual environment each time it is accessed. As well, a volume of a certain size or shape, when entered by a link, may yield an "interior" volume of a different size or shape. Thus, complex webs can be constructed with hyperlinks that cannot be modeled in physical, Cartesian space. These possibilities were explored in the development of this study, but it was quickly determined that participants become disoriented when the visual representation of spatial volumes does not map directly to buildable Cartesian space. More research needs to be done to understand the limits of human cognition in non-Cartesian spaces as made possible by hyperlink technology.

Customization
The project developed to explore this thesis has been constructed in a specific scene-describing language and intended for simulation on a limited number of rendering packages (see Appendix A). Although the output rendered and seen with this varying software may yield different results, the environment as a whole is not configurable or customizable. It is thought that in the near future, customization of software, interfaces, and even content of the virtual realm can and will take place, enabling the participant to configure the environment to one’s one personal taste. The project developed for this study was not designed with such capabilities in mind. Further study needs to be done to understand how and why virtual architecture should be developed under personally-configurable conditions.

Dynamic Environments
The virtual realm offers the opportunity to deny the physics of time, space, light, gravity, and solids; all of these have been addressed to one degree or another in the scope of this thesis. However, the virtual realm enables designers to break free of more constraints, such as the static nature of physical architecture as solid matter in and among a void. Study needs to be done into the expression of liquid, dynamic architecture which responds to a participant’s movements and interactions with other participants or with the environment itself (Best, Virtual Environments 25). "A liquid architecture in cyberspace is clearly a dematerialized architecture. It is an architecture that is no longer satisfied with only space and form and light and all the aspects of the real world. It is an architecture of fluctuating relations between abstract elements. It is an architecture that tends to music" (Novak, "Liquid Architectures" 251). Virtual architecture can be as responsive to the individual’s needs and actions as the complexity of the technology will allow.

Multiple Participants
Although the architecture developed for this study is intended for use by multiple participants, it has not yet been experienced in this way. It is suspected that the expression of virtual architecture, including the design offered here, will change as multiple participants become commonplace and the on-line culture matures. Issues related to virtual communities will need further study, including how their demands and wishes will influence virtual architecture (Rheingold). The representation of multiple participants of varying scales, moving through spaces in different orientations, has not yet been addressed. It is critical to consider these issues as the field of virtual architecture continues to grow.