Clarification about the Brown & Twiss experiment setting

Question

Based on "Optical coherence and quantum optics" by Leonard Mandel & Wolf, page 458:

it is said that one of the photo-multipliers (P₂) was "mounted on a horizontal slide which could be traversed in a direction normal to the direction pf propagation of the incident light. thus the cathode apertures, as viewed from the pinhole, could be superimposed or separated ".

up to this point, it makes sense. later on, though, it is said that "the normalized correlation coefficient C(d) of the photo-current fluctuations was measured as a function of the effective separation d=|r₁-r₂| between the photo-cathodes"

this inconsistency (to my understanding) makes me a little confused. how is the motion of P₂ on an axis normal to the direction of propagation of the incident light equivalent to the change in the effective separation d?

thanks in advance for anyone who would take the time and help me resolve the puzzle!