I have been trying to carry out the Stewart assay on lipid extraction to detect the phospholipid content. I have used the Bligh/Dyer technique to isolate lipids. I tend to do this in triplicates. The problem I am finding is that the first reading is always higher than the second reading, and the second reading is always higher than the third reading. I am using a quartz cuvette to take the readings as I am working with chloroform, I wash the cuvette with acetone. I am trying to create a standard curve, but the readings are not very linear. So, I am wondering if time is an important factor in this assay?