Authors’ response to T Goodman

therefore very high. It might be more appropriate to consider an approach using a portable spectroradiometer. Second, it is not clear how such a meter will be used. Lamp manufacturers will probably calculate the S=P ratio from spectroradiometric measurements, rather than by direct measurement of the scotopic and photopic luminous ̄ ux. Practitioners, on the other hand, will generally be concerned with: (a) choosing the `best’ lamp for a given application; or (b) ensuring that the installation meets the required speci® cations. For the former they will be guided by information from manufacturers regarding the (photopic) luminous ̄ ux and the S=P ratio for the lamp in question, together with knowledge of the illuminance level which they are trying to achieve Ð they are unlikely to make measurements themselves. For the latter a meter will be required, but does this need to include anything more than a V(l)-® ltered detector and appropriate software, together with the capability for the user to enter the S=P ratio for the lamps being used? A third class of potential users of a measuring instrument are `enforcers’ i.e., those responsible for ensuring safety guidelines, etc, are met. This is a much more complex situation, since guidelines currently specify photometric values and thus measurements should be in these terms as well. Unless and until recommendations are couched in terms of values determined using a de® ned uni® ed system of photometry, measurements associated with the demonstration of compliance with these recommendations will continue to be carried out only in terms of photopic values.