1010 
0,050 = 10,05 0,050 + 0,02 NaCl 18,382 0,5472 3,356 
0,100 17,64 0,100 + 0,05 __,, 38,55 04575 2,877 
0,200 26,71 0,200 + 0,13 _ ,, 80,58 0,38315 3,376 
Similar measurements with solid oxime salts containing less than 
three and more than one mol. of oxime in one mol. of hydrogen 
chloride yielded constants lying between the values found for /o. 
The results of the inversion measurements showed a satisfactory 
agreement with those of the saponification. 
From these values of 4, it appears that when the solid salt contains 
less acid, the power of the base decreases considerably, whilst the 
base of the liquid salt is by far the weakest of all. Hence, under 
the influence of hydrochloric acid, a shifting of atoms appears to 
set in, thus causing the basic properties to diminish. 
In the ease of the unimolecular oxime thus rearrangement might 
be caused by the formation of formamide 
CH,: NOH~O:CH.NH,. 
Of formamide, however, no well-defined hydrochloric compounds 
are known. I am thinking of undertaking a further investigation of 
the same. 
This shifting, however does not seem to me as being very likely 
because the acetyl- and the benzoyl derivative, also the formad- 
oxime themselves are trimolecular. Moreover, the existence of a salt 
(CH,: NOH), . HCI is difficult to tally with the said shifting. 
The facts may be better explained by assuming a cyclic formation 
of three molecules of oxime. The structure of the base of the primarily 
formed chloride would then be: 
(1\ 
31g ‘T) 
One mol. of this base is capable of combining with three mols. 
of hydrogen chloride; its dissociation constant is about 4, = 23 .10-!3. 
For acetoxime 4} = 6,5 Xx 1013 at 25°, whereas the higher homo- 
logues scarcely possess any basic properties. Hence, for formaldoxime 
we may expect a similar high value for fy. 
If now under the influence of hydrochloric acid a BECKMANN 
rearrangement sets in, the OH-group would change place with the 
H-atoms attached to C. 
