The first step to transform our hypothesis into a working experiment was the cleaning process. This allowed us to proceed with a more thorough system of data readings that would provide an essential clarity for our methodology.
After careful analysis of the space, working recorder-plans were established in order to provide us with a constant and reliable source to which would be used for our efforts in establishing an initial and final temperature of the radiant floor (as well as intermediate plans for our analysis of the transfer of heat throughout the space). A temperature grid was then established in order to provide accurate temperature records during each of our stages of documentation.

The initial thermostat setting as well as "zeroed pressure gauge" (quotes facilities manager) marked the original state of the "cold-run."

Raytek guns were then used to document all stages of our study, and seemed to be the most efficient (and least hazardly) way to record the data.

After all initial temperatures were recorded, we were ready to power up the flooring system by way of setting the thermostat to our second control temperature, then turning the pressure knobs completely clockwise on the distribution unit. We would estimate a documentation of time in ten minute increments . . .
. . . which would quickly change to a more reasonable documentation of half-hour increments in sixteen series.

We returned to the site three days later to observe and check to see if our final results had varied from our last record; luckily, they remained the same. In the end, our experiment would start at 6:30 pm and run until 2:30 am, whereupon the glass artists were still producing their art at a tireless speed.