Nitrogen Compounds and Weak Organic Acids. 197 
of the normal tartrate. As already stated the substance for 
analysis was obtained from the mother liquors from the third 
crystallization. 
•1014 grs. of the tartrate, three times recrystallized from alcohol 
and petroleum ether, dissolved in 13'9 c.c. of water, gave a rotation 
of + 1533° in a two decimeter tube. 
Hence [a] D = + 105° and [M] D — + 427'6°, 
[M] d for acid tartrates of strong bases = + 24 - 5°; therefore [JfJ^for 
the basic ion = + 403'1°. 
This value is slightly above that of the ion of the pure 
ammonium salt, which is 398°. The rotatory power of the iodide 
given below shews that the salt has been only partially resolved. 
The high value for the molecular rotation of the salt is probably 
due to contamination of the acid tartrate by the normal tartrate 
which would naturally raise the value of [ M ] n . The iodide was 
precipitated from the solution of the tartrate in water by means 
of potassium iodide. 
T030 grs. of the iodide in 130 c.c. alcohol in a two decimeter 
tube gave a rotation of +1T7 0 ; therefore [M] n for the iodide 
is 4- 275°. 
It is therefore evident that the substituted ammonium com- 
pound used forms a definite crystalline compound with ( d ) tartaric 
acid, and that from this the (d) base (d) tartrate can be separated 
from the ( l ) base (d) tartrate by crystallization. One would there- 
fore expect an optically active nitrogen compound to resolve an 
inactive organic acid. To test this assumption the author tried 
to resolve inactive mixtures of ( d ) and ( l ) mandelic acids, and of 
(d) and ( l ) valeric acids by means of active phenylmethylbenzyl- 
isopropyl ammonium iodide (Thomas and Jones, J. C. S. 1906, 
p. 280). A calculated quantity of the iodide was dissolved in 
fifty per cent, alcohol and converted into the optically active 
ammonium hydroxide as already described. 
To the solution of the hydroxide was added the calculated 
quantity of the acid. It was found that evaporating the solution 
on the water bath caused complete racemisation to take place. 
However, if the solutions were allowed to evaporate slowly in 
vacuum desiccators, crystalline products were obtained. These 
were recrystallized several times from acetone. 
After the third recrystallization of the ammonium mandelate 
J258 grs. in 16 , 5 c.c. alcohol in a two decimeter tube gave a 
rotation of — , 364°. [ M] D = — 93'8°. 
It is probable, since the difference between [M \ D of the salt 
and the basic ion is greater than that of the pure mandelic 
ion, that racemisation has occurred while the solutions were 
evaporating at the laboratory temperature. Before the first crop 
