NOVEMBEK 23, 1917] 



SCIENCE 



501 



these two substances are fundamentally 

 different. This is easilj^ demonstrated by 

 a study of their solutions. When a water 

 solution of trimethylmethoxyammonium 

 hydroxide (A) was heated, it decomposed 

 quantitatively in accordance with the fol- 

 lowing equation : 



(B) (CH3*)3sN - + 



+ H (4) _ 

 + CH,(5) — 

 '^^t.O + CHa -{ O h H. 



(A) (CS,),-N[ 



-OH (5)_ 

 -OCH3(4) — 



(CH3)3N + CH:0 + H,0. 



While trimethylhj'droxyammonium methyl- 

 ate (B) showed a totally different behavior 



(B) (CH3)3NZ8cH3(5) = 



(CH3)3N = + CH30H. 



In addition to these compounds, Meisen- 

 heimer prepared a number of isomeric 

 mixed dialkjd compoimds, with methyl, 

 ethji and propyl radicals, e. g., 



(CH3)3N-^^|.; and ( CH3) 3N Z ocl^' . 



In every case, water decomposed com- 

 pounds of this type to give a tertiary 

 amine, an alcohol and an aldehyde ; but, in- 

 variably, the radical which was eliminated 

 as aldehyde was the radical which occupied 

 "position four (4)" and the group elimi- 

 nated as alcohol alwaj'S occupied "position 

 five (5)." Meisenheimer stated that he 

 never obtained even recognizable traces of 

 the aldehj'de which should have resulted if 

 the group attached in position five had sep- 

 arated in that form. His conclusions may 

 be stated in his own words : 



Durch diese Beaktion ist bewiesen dass die 

 beideu Alkoxygruppen nicht in gleieher Weise an 

 das Stickstoffatom gebunden sind. 



The key to these disputed relations is 

 easily furnished in terms of the electronic 

 conception, by assuming that the one hy- 

 droxyl (or alkoxyl group) is positive and 

 the other negative; thus: 



(A-) CCHM =N— + — +CH.(4)_ 

 (4; CCJI3 ;a-JN ^_0__i.H (5) — 



(CH3-), = N_|__Q__^g -H H,C;:0, 



(CH,*). 



It will be observed that the two groups, 

 or, in the final analyses, the two oxygen 

 atoms, upon which the electromerism de- 

 pends, are not linked directly, but through 

 a third atom, nitrogen. 



EO-I NH OH and HO-1 N H OE. 



This is undoubtedly responsible for the 

 relative stability of these eleetromers as 

 compared with others in which the atoms of 

 different polarity are directly connected; 

 e. g., 



A-\ B and A \- B. 



Here, again, the analogy to structure tau- 

 tomers appears. Chemists have failed to 

 prepare desmotrops of prussic acid, and of 

 many other compounds in which the wan- 

 dering atom passes from one atom in the 

 molecule to another directly linked to it. 

 HON ?=± CNH. 



The majority of successful separations of 

 desmotrops lie among compounds in which, 

 similar to the eleetromers described above, 

 tautomeric changes involve two atoms not 

 directly linked, but connected by a third 

 atom. Thus, in the keto-enol and in the 

 nitro-nitronic acid desmotrops. 



the wandering hydrogen atom passes from 

 carbon to oxygen not directly linked. 



2. Stereomers. — Meisenheimer-^ was the 

 first to observe that amine oxides with three 

 different radicals R R' R"N ^ 0, could be 

 resolved into enantiomorphous modifica- 

 tions. Amine oxides of this kind were pre- 

 pared by oxidizing tertiary amines with 

 hydrogen peroxide, or Caro's acid. Meis- 



23 Meisenheimer, Ber., 41, 3973 (1908) ; Ann., 

 385, 117 (1911); 399, 371 (1913). 



