BULLETIN OF THE BUSSEY INSTITUTION. dl 
ing the alcoholic solution to crystallize, with proper precautions, 
the characteristic crystals of mannose-hydrazone will appear. If, 
through too rapid evaporation of the alcohol or from any cause, 
no thoroughly satisfactory crystals appear on the first trial, the 
process of recrystallization from alcohol should be repeated until 
all doubts are removed. It is worthy of remark that better 
results have been obtained, as a general rule, when the alcoholic 
solution has not been filtered, but has been simply decanted from 
or even left to stand in the tube in which the solution was made. 
In case neither crystals of mannose-hydrazone nor the charac- 
teristic balls of this substance appear when the unevaporated 
product of the hydrolysis has been tested several times in separate 
portions of the liquor, each of 1 cc. or more, the remainder of the 
liquor is evaporated to dryness on a water-bath, the residue is 
taken up with a small definite quantity of water, usually 10 cc., 
and 1 ce. portions of this solution are tested with the phenyl- 
hydrazin reagents, very much as before, though in this case 
sixteen drops of the reagent may be added instead of the twelve 
drops that are used with the weaker, unevaporated liquor. In 
case no satisfactory precipitate is obtained on testing 1 cc. por- 
tions, then larger quantities are taken: sometimes 3 cc., and 
sometimes the whole of the residual liquor is treated at once with 
the reagent. 
One advantage gained by working methodically with similar 
weights of the woods and definite volumes of the solutions is that 
rough indications may be obtained as to the relative quantities of 
mannan contained in the different woods. Manifestly, a heavy 
precipitate of balls and crystals of mannose-hydrazone obtained 
from 1 cc. of the unevaporated liquor will show that the wood 
under examination contains an abundance of mannan, while a 
meagre precipitate obtained from several cc. of an unevaporated 
liquor indicates that there is less mannan in this wood than there 
was in the other. Similar relative indications are to be had when 
the product of the hydrolysis is evaporated, and 1 or 3 or any 
larger number of the 10 cc. of liquid thus obtained are tested. 
It is easy to demonstrate in this way that while pine wood con- 
tains much mannan, comparatively speaking, the wood of the 
sugar-maple contains less of it, and less at some seasons than at 
others, while in the wood of the gray birch, taken in the quantity 
