BULLETIN OF THE BUSSEY INSTITUTION. TD 
and the radiating crystals with blunt ends appear simultaneously. 
Speaking in very general terms, it may be said that the ‘‘ original 
precipitate’ which appears on adding the phenylhydrazin reagent 
to solutions of mannose consists more frequently of the smooth 
globules than of the crystalline forms, though the latter may be 
exhibited by dissolving the globules in alcohol and allowing the 
solution to crystallize. As will be stated below, there are several 
kinds of globules which may appear in the ‘‘ original precipitate,” 
but it is none the less true that the globules of mannose-hydrazone 
exhibit certain peculiarities which appeal more or less strongly to 
an experienced eye; they are apt, moreover, to appear soon after 
the addition of the reagent, unless the mannose solution was very 
dilute. It is essential, however, in every instance to convert the 
globules to crystals. 
The thought that the prismatic crystals might perhaps result 
from the crystallization of the globules as they lie quietly in the 
‘midst of the original precipitate, — that is, from a rearrangement 
of the particles of matter of which the globules are composed, 
even though no re-solution of them had occurred, — has not been 
borne out by experience. This idea seemed at first to be strength- 
ened in some measure by the occasional formation upon the 
globules of the fringe of crystalline points, above mentioned, 
which gives the elobules a roughened appearance. Sometimes, 
it has almost seemed as if the contents of the smooth globules 
might possibly be in a condition akin to that of a supersaturated 
saline solution almost but not quite ready to crystallize, but on 
crushing some of the globules under the microscope they fell to 
an amorphous powder; though, on the other hand, on crushing 
some of the roughened globules, on which short points or spines 
of projecting crystals had appeared, it has been noticed once or 
twice that they exhibited a crystalline structure internally, in that 
they consisted of a bundle of minute fibrous crystals radiating 
from a point at the centre of the ball. Occasionally the short, 
blunt spines springing from the yeast-like globules show distinctly 
that they are really the bevelled terminations of prismatic or plate- 
like crystals lying in bundles within the ball. Generally speak- 
ing, the appearance of mannose-hydrazone in either of the forms 
above described is sufficiently characteristic, especially to any 
one who has had some experience in testing for this substance, 
