The basic arguments for a mental image model of thought are based on neuropsychological evidence. Farah [1] found that the same brain regions are activated during both mental representation and actual perception, while Bishiah [2] found that brain traumas that affected perception, also affected the ability to create mental images. Pylyshyn [3] on the other hand, argues that all mental images are guided by ‘’cognitive penetrability’’, thus on their very basis, are manipulated by certain propositional elements. Given this difficulty, Kargopoulos [4] hinted towards further research, featuring shapes and solid objects, for which subjects have no prior extensive knowledge. This would force subjects to use non-semantic strategies of representation, meaning mental imagery. Hinton’s [5] cube problem conforms to these requirements. Hinton’s problem aligns with the idea that spatial tasks (especially tasks with cubes that change layout) are guided by propositional cues (our knowledge about squares) and supports Pylyshyn’s position. Using one of the simplest objects, a cube, Hinton showed that as soon as this shape changes its mental arrangement in space, even suspicious -as to the nature of the experiment- participants will make mistakes that are not present when they manipulate a mental image of the cube sitting on its typical array. Aim: Our goal was to investigate the relationship between spatio-visual skill and the ability for mental partitioning in healthy subjects. Methodology: We used 2 groups (344 participants) a control and an experimental one. In the control group, we presented a Moebius’ strip, in the experimental group, we presented the same Moebius’ strip and asked them to mentally represent it. All participants asked to mentally partition the strip. Results: Of the 344 participants, only 31 managed to give the correct number of vertices in space. Though people had a hard time manipulating the cube’s mental image, their success rates were much higher for the Hinton 1 task in which propositional representation was more accessible. Only 9 of the 344 participants could find the correct answer for the Moebius strip task in which mental manipulation of the strip image was impossible. Conclusions: We come to the conclusion that the relationship between ‘’seeing’’ and ‘’knowing’’ is more complex, not just on the level of the mental image level but also on the level of perception. Our findings bring back to the scientific background the idea that the mind’s selective attention to previous experience and cognitive schemas will decidedly affect human thought.