176 REPORT—1904. 
gave B benzyl-hydroxylamine on hydrolysis, and was therefore a nitrogen 
ester. 
Later Werner and Buso ! found that the liquid oxygen ester just re- 
ferred to, on treatment with hydrochloric acid underwent isomeric change 
during its partial hydrolysis, and gave rise to a solid oxygenester. There 
are consequently two isomeric oxygen esters, the existence of which can 
only be explained by different steric relations, and a nitrogen ester. 
The phenomena exhibited by the methyl] esters are complementary to 
those observed in the benzyl esters. Petraczek ? prepared a methyl ester 
by the action of sodium and methyl iodide on a benzaldoxime, which ester 
when hydrolysed gave a methyl-hydroxylamine, H. Goldschmidt and 
Kjellin® isolated an ester from the products of the interaction of 
2 benzaldoxime, methy] iodide, and sodium methylate, which gave /3 methyl- 
hydroxylamine on hydrolysis, and was therefore a nitrogen ester ; at 
the same time they observed the odour characteristic of the oxygen esters 
and concluded that the oxygen ester of the syn oxime was formed at the 
same time. 
Finally Dr. Luxmoore?* observed that by the action of methyl bromide 
and hydrobromic acid on £ benzaldoxime the hydrobromide of a new 
nitrogen ester was produced. This ester differed from the nitrogen ester 
already known in being very readily hydrolysed by water to form 
B methyl-hydroxylamine ; it was also labile, and on standing changed into 
the stable syn N ester. 
Thus there is enough evidence to justify the conclusion that two benzyl- 
oxygen esters and two methyl-nitrogen esters exist, and the isomerism in 
these cases can only be explained on the hypothesis of stereoisomerism— 
thus, 
. | and So yo 
wis 
C,;H,ON NOC,H, CH,N N—OH,. 
This conclusion is supported by Goldschmidt’s work on isomeric oxygen 
esters of anisaldoxime and nitro-benzaldoxime.° 
The isomerism of the nitrogen esters is analogous to that of tropine 
and wy tropine. 
In addition to the Beckmann transformation a second criterion for 
the determination of configuration is applicable to the aldoximes ; one of 
the oximes on treatment with acetic anhydride loses water and gives a 
nitrile, whereas the other either remains unchanged or gives an acetyl 
a—C—H 
derivative, the former must therefore be the syn-oxime, (| , and 
NOH 
a—C—H. 
the latter the anti-oxime, ll 
HON 
The hypothesis of Hantzsch and Werner, which has, as we have seen, 
accounted in a satisfactory manner for the isomerism observed among the 
oximes, requires that isomerism should exist also in other compounds with 
1 Ber., 1895, 28, 1278. 2 Ber., 1882, 16, 827. 3 Ber., 1891, 24, 2812. 
Trans. Chem. Soe., 1896, 69, 177. 5 Ber., 1890, 28, 2178. 
