1.In the well-known Stephen method of preparing aldehydes from corresponding nitriles by reducing the latter with anhydrous stannous chloride in the presence of hydrogen chloride in anhydrous ether,two typical nitriles,namely,toluonitrile and a-naphthonitrile,could not be reduced smoothly and gave negligible yields.This was ascribed by previous authors to the steric hindrance in the classical sense of the words.We have now carried out more extensive research,subjecting variously ortho-,meta-,and para-substituted halo-,methoxy-,and carboethoxybenzonitriles to the reduction process under identical conditions and by comparing the yields obtained with various resulting substituted benzaldehydes(see Table Ⅰ),and found that steric hindrance can not be regarded as the sole cause which prevents Stephen reaction to proceed smoothly.In view of the facthat 2,6-dichlorobenzonitrile and 2,6-dimethoxy- benzonitrile failed to be reduced completely,steric hindrance must play an importance part in the process.There must be other factors which share with the steric hindrance in the reaction but their characters are not yet ascertained.2.Must of the substituted benzonitriles used were known and prepared previously by dehydration of the corresponding benzamides or by halogenation of benzonitriles.In both of the processes,the yields were not satisfactory and the products were not always pure enough.We prepared the necessary substituted benzonitriles by Sandmeyer reaction from the corresponding amines and found that this route gave satisfactory yields and purer products.3.Though the substituted benzaldehydes prepared in the present work are alt known compounds,but only a few of these were previously prepared by the Stephen method.4.The attempted reduction of pentaacetyl-D-glucononitrile to pentaacetyl-D-aldehydo-glucose failed;the product was pentaacetyl-D-gluconamide.