I just wanted to add that I see you have the k2006 kit, which I assume you're testing with, but reported reading to the decimal point. When slamming, and anytime really, you only need to use the 10 mil sample size, not 25 mil. This will save you on reagents 