Search the CONFOCAL archive at
http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal

We have the Solent chamber on one of our systems. CO2 level and especially volume for long term 
time lapse experiments have been a problem for us too. After extensive trial and error, so far we 
have solved it two ways:

1) sealed culture flask with a syringe needle through the cap to change medium every day or two 
(tape the flask to the stage to prevent movement when changing medium; use a long piece of 
tubing between the needle and the syringe itself so that the syringe itself is resting on the floor of 
the chamber keeping warm and not exerting any tension on the flask).

2) 12 or 24-well plate sealed with parafilm, water in wells you are not using, syringe needle 
through the lid on one side bubbling 5% CO2 (premix) into the water of one well, syringe needle 
through the lid at the other side of the lid to vent.

Method 1 works for two weeks of continuous timelapse (acquiring an image every thirty minutes), 
Method 2 was suggested by the Solent operations manual. We have used it for four to five days, 
but haven't tried longer.  You can use glass bottom flasks/well plates for oil immersion lenses.

Cells continue to divide and thrive for the whole time lapse. At 10x we have followed the same 
clone of cells for two weeks through multiple doublings. We are not using HEPES, just bicarb.