A graduated cylinder is meant to be read with the surface of the liquid at eye level, where the center of the meniscus shows the measurement line. Is it also true for any non-transparent or a colored liquid? There're some diverse information in different sources.
Graduated cylinders are not designed to deliver accurate volumes, so one does not worry about upper or lower meniscus issues with transparent or dark liquids. The best way is to actually measure the volume of the dark liquid by mass, if this is a concern. This can be tested by weighing the nominal volume delivered by graduated cylinder. One should know the density of the liquid. Make sure the cylinder has a marking which says TD (=to deliver) or Ex (which also means to deliver).
For example, for a 500 mL good quality cylinder (class B) may have an error of 4 mL. This error is quite larger than what one might expect from meniscus errors. This is pretty large for accurate and analytical chemistry measurements.
The volume of the meniscus can be considered as negligible. So the difference in volume between upper and lower meniscus is negligible. Please refer to one of these conventions and stay there. This is specially true when measuring differences of volumes, for example when a certain amount of liquid had been removed from a burette for a titration.
