It's a question on Fresnel-Fizeau light drag effect. Long pillar (section is square) made of glass is set horizontally. To the upper face of the pillar, light waves of a star are entering horizontally. These light waves reach bottom face horizontally also (it must be so geometrically). The pillar is supposed to move from the right to the reft because of the motion of the earth. Whether light rays go straight or are bent (from view of the frame of star) ? If light rays go straight, light rays may not be at right angle to the light waves (at frame of glass). And speed of both in the medium may differ. It's unreasonable (according to the "extinction" view, the speed of light waves is constant). Therefore light rays may be bent (at frame of star). And light rays must be at right angle to the light waves (at frame of glass), then, light drag may not be partial.
In a book "Teoria della relativita" By W.Pauli 1970, it's written as follows. An experiment (to measure Fresnel drag coefficient) was done with the light ray that is emitted at right angle to the direction of the motion of medium. About the result, it's not written clearly.