Nitrogen and an Asymmetric Carbon Atom. 4G7 



The union with allyl iodide was next studied, and it was found 

 that the salt produced had a decided rotatory power [a]^ = 1*5°, 

 and on recrystallisation from alcohol the value of [a] D became 

 163° in alcohol and 21 - 8° in chloroform. 



It is therefore probable that two compounds, one of which 

 is dextrorotatory and the other laevorotatory, have been produced, 

 and that the former is the less soluble in alcohol. The more 

 soluble salt could not be isolated in a pure form, since, however 

 carefully the allyl iodide had been purified, the crude product was 

 always somewhat coloured and the more soluble portions were 

 strongly coloured. 



That the rotatory power is due to the nitrogen atom and not 

 to the carbon is proved by the behaviour of the salt in chloroform 

 solution. The rotatory power falls at first rapidly, then more 

 slowly, until it reaches a constant value of *38° for a 6'1 per cent, 

 solution, or '22° for 3"44 per cent, solution in a 2 dcm. tube 

 corresponding to [aJ^^S'l . The change here is most probably 

 a transformation of one isomer into the other by dissociation into 

 tertiary amine and alkyl iodide, and subsequent recombination 

 until equilibrium is attained. (Cp. the autoracemisation of active 

 nitrogen compounds ; Pope and Harvey, Trans. Chem. Soc. 79, 830, 

 and Jones, Trans. Chem. Soc. 85, 229.) 



The union with benzyl iodide was then studied, and the re- 

 action, which takes place rapidly and gives a practically colourless 

 product, was found to be more suitable for the present investigation 

 than the reaction with allyl iodide. 



The reaction was allowed to proceed in different solvents at 

 the ordinary temperature and the crude product examined. All 

 the products were dextrorotatory. The salt deposited from chloro- 

 form solution had a much greater rotatory power ([a] i) = 23 , 3 in 

 23 per cent, solution in chloroform) than that obtained from al- 

 cohol, ether or ethyl acetate solution ([«]# = 3"5° — 5° in 2*3 per 

 cent, solution in chloroform). 



Repeated crystallisation from hot alcohol or from warm alcohol 

 did not cause any very great change in the rotatory power of the 

 salt ; so that this method does not readily effect the desired 

 separation of the two compounds. 



The d- and Z-camphorsulphonates were therefore prepared, 

 crystallised several times from ethyl acetate, their rotatory powers 

 determined, and the iodide recovered from the various fractions 

 and examined. 



The less soluble fraction of d-camphorsulphonate gave 

 [a] D = 7'18° and [M] D = 31"1° in aqueous solution, and hence 

 must contain a laevorotatory basic radicle. The iodide recovered 

 from it gave [a]^ = — lOO in alcohol. 



The less soluble fraction of £-camphorsulphonate was also 



31—2 



