500 



SCIENCE 



[N. S. Vol. XL VI. No. 1195 



group N — OR, should show different phys- 

 ical and chemical properties, depending 

 upon the nature of the hydroxyl, or alkoxyl 

 group; i. e., whether it is negative (I.) or 

 positive (11.). 



N H O h E, 



(I.) 



N — + O — + E. 

 (II.) 



Then, again, provided one of the hydroxyl 

 groups is positive and the other negative, 

 compounds of the type, (R)3N(OH)2 

 should exist in two isomeric forms (elec- 

 tromers) when one of the hydrogen atoms 

 is replaced by a single radical R ; and, fur- 

 thermore, there should be two distinct 

 isomers (electromers) if two of the hydro- 

 gen atoms are replaced simultaneously by 

 dissimilar radicals, R and R'. 



(R*)3 



+— O— +H 

 h O h K ■ 



(R+)3 = N , ; 



--HE" 



— -FO — +H 



-f — O — -hE ■■ 



+ 0- + W 

 '■^^ -1 O h E 



In the second place, compounds of the 

 formula R, R', R", N(0H)2, in which there 

 are three different alkyl (or aryl) radicals 

 linked to the nitrogen atom, should exist in 

 stereoisomeric modificatmis, provided one 

 hydroxyl group is negative and the other 

 one is positive. By the action of an opti- 

 cally active acid, e. g., d-bromocamphorsul- 

 fonic acid, or d-tartaric acid, a racemic 

 compound obtained by synthesis should be 

 resolved into a dextro- and a levo-modifica- 

 tion. Optical activity might even persist 

 in the corresponding amine oxides R, R', 

 R", N;! 0. Although the two hydroxyl 

 groups are structurally alike, they are 

 totally different electronically. Conse- 

 quently, the nitrogen atom is linked to five 

 different radicals, and, in this respect, com- 

 pounds of these types may be compared to 

 substituted ammonium derivations of the 

 formula R, R', R", R'", N — X, which have 



been resolved into optically active forms^^ 

 repeatedly. Experimental evidence sup- 

 porting both of these deductions has been 

 presented quite recently. 



1. Electromers. — In an article concern- 

 ing the "Non-equivalence of the Five 

 Valences of Nitrogen," Meisenheimer^^ de- 

 scribes the preparation of two isomeric 

 compounds of the type 



(E)3N(OCH3)(OH). 



The iirst isomer was obtained by the action 

 of methyl iodide upon trimethylamine 

 oxide, and the subsequent replacement of 

 iodine by hydroxyl. Thus: 



1. (CH3)3]Sr = O -I- CHsI = (CH,,)3N = °*^^=' 



2. (CH3)^~°^^' + NaOH=: 



(CH3)aN — g^^^-1-NaI. 



The second isomer was secured by the ac- 

 tion of sodium methylate upon the salt ob- 

 tained by treating trimethylamine oxide 

 with hydrogen chloride. 



1. (CH3)3N = + HC1: 



2. (CHs)3N ~ ^ + NaOCH3 = 



(CH3)3N-OH^ 



(CH3).N_°^jj^ -l-NaCl. 

 (-S) 

 The two forms, {A) and {B), are identical 

 except for the order in which the hydroxyl 

 groups and the methoxyl groups were in- 

 troduced. In (B), as Meisenheimer said, 

 the methoxyl group is linked to the "fifth 

 valence, ' ' or the one which usually engages 

 the acid radical; while it is linked to the 

 "fourth valence" in formula (A). But 



21 Le Bel, Compt. rend., 112, 724 (1891); 129, 

 548 (1899); Ber., 33, 1003 (1900); Wedekind, 

 Ber., 32, 517, 3561 (1899); 35, 766 (1902); 36, 

 3791 (1903); 38, 1838 (1905); Wedekind and 

 Oberheide, ibid., 37, 2712, 3894 (1904) ; Wedekind 

 and Froelich, ibid., 38, 3438 (1905) ; Pope and 

 Peachey, J. Chem. Soc, 75, 1127 (1899) ; Pope and 

 Harvey, ibid., 79, 828 (1901). 



22 Ann., 397, 273 (1912). 



