Abstract
In a previous paper, we presented evidence for a large apparent day-to-day variation of the outside-atmosphere deflections of two Volz sunphotometers and discussed the consequences for Volz sunphotometry should such variation prove to be typical. Upon further examination of our data, we have found that much—and perhaps all—of this variation can be attributed to true short-term variation of outside-atmosphere deflection, atmospheric optical depth, or both; we cannot distinguish between the two possibilities. If either or both of these variations occur at a given site at which Volz sunphotometry is done, and if lnJ0 is periodically redetermined on-site by the observer using the Beer-Lambert law, errors up to the size we gave in Paper I may be present in the resulting data. We discuss ways of assessing the size of the error (for existing data) and avoiding it (for future data).
