On the poor correlation between the inhibition by methotrexate of dihydrofolate reductase and of deoxynucleoside incorporation into DNA.