ON THE PRESENT POSITION OF THE CHEMISTRY OF RUBBER. 255 



He proposed the following structure as indicating the arrangement of the 

 carbon atoms : — 



^\c = CH . CHj . CH., . C = CH . C 



CH, 



which would give on oxidation 



OHC . CHj . CHj . CO . CHj, 



i.e., levulinic aldehyde. 



In a subsequent paper ' Harries explained that the substance melting 

 at 195' was not an acid, as at first supposed, but a peroxide of levulinic 

 aldehyde having the formula — 



O = CCCHj) . CHj- CH2 - CH : 



II II 



O =0 



Its formation was due to a method of splitting up of the ozonides, not 

 previously described : — 



>C = CHR + 03= >C — CHR 

 r/ W \ \ 



0—0—0 



R. 



>C CHR=R. 



R/ I I >C< I + OCHR 



0-0—0 R/ \0 



Thus the caoutchouc ozonide is decomposed by water as follows : — 



C,„H,„0, = CHj-CO . CH^ . CHj . CH0^0 = C(CH3)CH, . CH^ . CH : 



^ II II 



0-^ o 



in about equal portions. The peroxide decomposes on long boiling with 

 water into levulinic aldehyde and H2O2, giving subsequently levulinic 

 acid. 



These were the only products of the decomposition, and Harries 

 pointed out that as levulinic aldehyde is the only product formed in the 

 splitting up of the ozonide, the caoutchouc molecule must be formed of a 

 carbon ring, and not an open chain. The molecular weight of the ozonide 

 was found to agree with the formula CigHigOg. 



The chemically reacting molecule of the hydrocarbon itself must be 

 looked upon as a derivative of an eight-carbon- ringed body, i.e., 1 : 5 

 dimethyl-cyclo-octadiene : — 



CHj-C— CH„— CH.,— CH 



II ' II 



CH — C Hj — CH2 . — C — CH3 



' Jler., 38 (1905), 1195. 



