Why not simply measure the two samples individually and superpose the spectra?

I could see trying to measure the two samples at the same time leading to some potential problems.

That is to say you may measure a different absorption value for the second sample because there are fewer photons left to absorb by the second sample after it has gone through the first sample. 

If i were you I would either measure individual samples or if you need to have a spectrum of the two species together i would make an equal molar mixture of the two compounds in the same cell and measure it that way. 

What Adam says is good advice, although if your concentrations are not too high there shouldn't be issues with sequential absorption.

If you can't mix the two compounds in the same solution, you could always measure independently and add the spectra mathematically

I agree with Hugh about the ow concentrations (for most compounds) but in the case of some organic dyes even low concentrations can have high absorptivity coefficients.  Porphyin for example his a soret band with extinction coefficients that exceed 500 000.

If you absolutely wish to measure your spectra with two cuvettes I would make sure that they are low concentration and that the do not absorb at the same wavelengths (otherwise one may act as a filter for the other)