Paulo,

I assume the pulp is available to you as dry and thin "cornflakes". If so, performing transmittance measurements will be more challenging in terms of consistent layer thickness and light penetration through the sample. I would say the most practical way to start is to perform the measurements in diffuse reflectance (or better: remission) mode. However, be aware that your measurements now mainy represent reflection and disperion of light as a function of your sample characteristics. So if you want to investigate the "consistency" of pulp, it is hopefully captured by this type of measurement (which might be perfectly the case). If not, you can always adjust to transmission mode. In both cases, you have to think carefully about your flake size, the number of flakes, the packing density etc., since all these factors will contribute to your spectral signal as sources of variation. Maybe you can start with only one flake, and use a piece of spectralon to strech it out and press it on the glass window of the spectrometer (then perform the diffuse reflection measurement). If the spectral variation of your consistency classes (1/2/3%) is larger compared to sampling handling, you're in business!