14 BULLETIN OF THE BUSSEY INSTITUTION. 
provided with a reflux condenser; then neutralizing the acid 
liquor with dilute caustic soda, using phenolphthalein as an indi- 
cator if the solution is clear enough, and adding to the neutralized 
liquor either a mixture of phenylhydrazin and glacial acetic 
acid or of phenylhydrazin hydrochlorate and sodium acetate, or 
simply, though on the whole less conveniently, a solution of, 
phenylhydrazin in water. Ordinarily, the reagent first named 
has been prepared by shaking together 1 cc. of phenylhydrazin, 
2 ce. of glacial acetic acid, and 10 cc. water; and the second 
solution has been made by mixing the two constituents in mole- 
cular proportions. Usually 0.5 grm. of phenylhydrazin hydro- 
chlorate is dissolved in 5 cc. of water and 0.3245 grm. of 
sodium acetate is added. 
The yellowish-red precipitate that soon falls on the addition of 
the phenylhydrazin reagent should be examined under the micro- 
scope from time to time, and, subsequently, for the sake of 
obtaining more perfect crystals, it should be washed with water 
and recrystallized from alcohol. As was noticed long ago by 
Fischer,* the yellowish color of the original precipitate is not in 
the least an indication of the presence-of mannose, for mannose- 
hydrazone is almost absolutely colorless when pure. Usually 
the yellow precipitate will be seen under the microscope to contain 
considerable quantities of amorphous matter. At the best, the 
mannose-hydrazone will appear in the form of bunches of radiat- 
ing prismatic crystals with blunt ends, —that is to say, the 
terminations of the crystals are chamfered or cut off at a bevel; 
they have no sharp needle-like points nor are the ends of the 
crystals actually square or rectangular. But it often happens 
that mannose-hydrazone, when first precipitated from the neu- 
tralized product of the hydrolysis of vegetable matters, takes the 
form not of crystals but of smooth globules of greenish-yellow or 
brownish-yellow color. These balls are not perfectly round, as a 
general rule, and they are apt to be joined one to another more or 
less closely. They resemble somewhat in appearance the large, 
round cells and clusters of cells of some varieties of yeast. Not 
infrequently the globules are no longer smooth but beset with 
numberless very short points, as of needles about to shoot forth. 
Sometimes mixtures of smooth globules, roughened globules, 
* Tollens, Handbuch der Kohlenhydrate, 2, 115. 
