442 TRANSACTIONS OF SECTION B. 
This compound, when hydrolysed, was converted into a dimethyl mannitol (m.p. 
93°, [a] —8°-8 in alcohol) ‘giving, on oxidation, a dimethyl mannonic acid which 
was identified by analyses of the calcium salt. 
On the other hand, when the hydrolysis of mannitol triacetone, under the condi- 
tions described above, was continued for four hours, the main product was mannitol 
monoacetone (m.p. 85°, [a]p ++ 23°-2). This was converted into tetramethyl mannito 
monoacetone (b.p. 137° to 140° /1l mm., [a]p’”+ 32°-2), and finally into tetramethy 
mannitol (b.p. 167° to 169°/13 m.m. [a]p’ —12°-5). The latter, on oxidation with 
nitric acid, gave a tetramethyl mannonic acid (b.p. 180° to 182° /12 mm. [a]p” + 10°-1) 
which failed to give a lactone on repeated distillation. For the purposcs of com- 
parison, the isomeric tetramethyl mannonic acid was prepared from methylmannoside 
and found to be entirely different, as, on distillation, it was completely converted into 
tetramethyl mannono-lactone (b.p. 174° /11mm. [a]iy in dilute alcohol +-78°-8->27°-5). 
A review of all the above results shows that in dimethyl and tetramethyl- 
mannitol the alkyl groups are present in positions 5:6 and 3:4: 5:6 respectively, 
so that the structure of mannitol triacetone and its partial hydrolysis may be ex- 
pressed by the scheme :— 
CH,-O CH,-OH CH,-OH 
Pe Se (Mc), l l 
i H:O CH:OH CH-OH 
CHO 'H-O CH:OH 
| >C Me), ——> | Se (Me, ——> | 
Tae ‘0 CH:OH 
| 
CH-O CH:-O CH-O 
|e (Me), | Ye (Me), 1 So (Me), 
CH,-0 H,-0 CH,-07% 
The linkage of each of the condensed groups is thus in the a-position throughout, 
but the order in which the acetone residues are removed is quite unexpected and 
involves complex stereochemical considerations. 
Our results are supported by experiments recently made by Mr. J. L. A. Mac- 
donald who, in similar work, converted glycerol acetone into a monomethyl glycerol 
(b.p. 109° /12 mm.) which was identical with the compound obtained on decomposition 
of a: B dibromo—a-methoxy—propane with silver acetate, and hydrolysis of the 
roduct. 
It should however be mentioned that evidence was obtained that mannitol tri- 
acetone is either a mixture, or is capable of reacting in move than one form. When 
tetramethyl mannitol was oxidised by Fenton’s reagent, the bulk of the material 
remained unaltered, but a small quantity of 2: 3: 5 : 6 tetramethyl mannose was formed 
which was identified by conversion into the anilide. A small quantity of material 
in which at least one acetone residue connected y-carbon atoms must therefore have 
been present in the original specimen of the triacetone compound. 
3. The Rolatory Powers of partially Methylated Glucoses. 
By Professor J. C. Irvine and J. P. Scorr, M.A., D.Sc. 
During the past four years a number of partially alkylated sugars have been 
obtained by a variation of the method originally adopted by Purdie and Irvine ! in the 
preparation of tetramethyl glucose, and we are now in a position to add a series of 
partially methylated glucoses to the list of compounds of this type. 
The method of preparation adopted can be expressed in general terms. Glucose 
was condensed with various residues which could subsequently be removed by hydro- 
lysis; the unsubstituted hydroxyl groups were then methylated by the silver oxide 
reaction, and, on hydrolysis, a partially alkylated glucose was obtained. It was 
thus possible to protect selected groups from methylation, as shown in the following 
examples :— 
Monomethyl Glucose. 
Glucose diacetone, when methylated by the modified method described in the 
1Chem. Soc. Trans., 1903, 88, 1021. 
