V. J. HORGAN AND J. ST. L. PHILPOT 



Anaerobic alkaline leuco-brilliant cresyl blue reacts significantly more 

 with extracts of irradiated than of unirradiated mice ; but with cuproiodide 

 catalyst there is little evidence of unreactive peroxide since the reaction is 

 almost complete in 10 minutes at room temperature. 



Experiments with brilliant cresyl blue using less or no cuprous catalyst 

 show the presence of unreactive peroxide which is increased by irradiation. 



If the maceration of the mouse is performed in nitrogen rather than in air 

 the ionic yield is somewhat less. This suggests that some of the increased 

 peroxide is formed dui-ing the maceration if air is present, as a result of 

 either destruction of antioxidant by X-rays or production of some very 

 easily autoxidizable material. 



One of our troubles with leuco-brilliant cresyl blue was that small quan- 

 tities of peroxide turned it pink instead of blue. This was due to an impurity 

 extractable with butyl acetate. Although the impurity is of unknown 

 structure it has advantages over brilliant cresyl blue as a peroxide reagent, 

 being more stable in alkaline solution and being apparently better at dis- 

 tinguishing unreactive from reactive peroxide. 



We have searched for model substances with properties resembling 

 'unreactive peroxide', and the results are summarized in Table I, which 

 shows the approximate percentage reaction in ten minutes at room tempera- 

 ture and 100° C. With peroxides of unknown concentration contained in 

 autoxidized substances the peixentage is arbitrarily based on the extent of 

 reaction at 100° with aerobic leuco-2,6-dichlorophenol-indophenol by our 

 modification of the Hartmann-Glavind method ^ We thought that 'unre- 

 active peroxides' might be merely aldehydes, but these showed no reac- 

 tivity whatever under the conditions used. All the hydroperoxides tried 

 and also benzoyl peroxide and hydrogen peroxide were too reactive, going 

 to completion in 10 minutes at room temperature. Diethyl peroxide was 

 about right. In our previous report we excluded this type of peroxide 

 because it does not react with stannous chloride even at 100°C ; but since 

 the mouse extract contains reactive as well as unreactive peroxide its reac- 

 tion with stannous chloride may have been due solely to the former, and 

 the latter can still be classed with diethyl peroxide. 



Disubstituted peroxides like diethyl peroxide are not supposed to be 

 formed when simple unsaturated fats like oleates or linoleates are autoxi- 

 dized. Consistently with this Table /shows that the peroxide in autoxidized 

 ethyl oleate is wholely of the reactive type. It is known that conjugated 

 double bonds can give disubstituted peroxides by 1,4 addition, and con- 

 sistently with this Table I shows some unreactive peroxide among the autoxi- 

 dation products of isoprene and diacetyl. Neither of these substances has, 

 however, been found in mammals and preliminary experiments with the 

 polyconjugated natural substances carotene and vitamin A gave discourag- 

 ing results. We have also found 'unreactive peroxide' in autoxidized 

 morpholine and triethylamine (used as buffers in our peroxide reagents). 

 The nature of these is not yet clear. 



Very recently our attention has been directed to Squalene (dihydro- 

 hexaisoprene) by some work of Sobel and Marmorston-, who have shown 

 that it induces the autoxidation of carcinogens. Bolland and Hughes^ 

 have shown that squalene undergoes chain autoxidation whereby one 



27 



