It has been hypothesized that in phylogeny the encounter between potential signalling molecules and the continously changing cell membrane could result in the formation of a ligand specific receptor. This chemical (hormonal) imprinting is then transmitted to the progeny generations. It is, however, very difficult to know whether the selection of cells with receptor-like patterns or amplification of complete receptor-like patterns led to the formation of the receptor-hormone complex. The new technique of ‘chemotactic selection’ provides a physiological response-guided selection of cells. It also enables the testing of subpopulations with the characteristic selector ligand. We show here that of three chemotactic ligands (histamine, di-iodotyrosine (T 2 ) and human insulin), insulin and T 2 selected subpopulations express a significantly high chemotactic response. Since the control medium has a selector capacity itself, we introduced a chemotactic selection coefficient (Ch sel ) which facilitates the comparison of all groups. Using this factor we found that insulin (Ch sel =1.57), functions as a strong selector and T 2 (Ch sel =0.98), was a weak selector. Morphometric evaluation of the cells showed a good correlation between chemotactic responsiveness and morphometric characteristics of subpopulations selected with insulin and histamine. T 2 data suggest that the long lasting responsiveness is not general, but might be subpopulation specific.