28 BULLETIN OF THE BUSSEY INSTITUTION. 
Remarks on other Globules which might be confounded with those 
of Mannose-hydrazone. 
Beside the bare possibility that drops of free phenylhydrazin 
might contaminate a precipitate, as has been suggested by Pavy, 
and mentioned above on page 17, there are some other kinds of 
globules that need to be considered. In the products of the 
hydrolysis of birch wood, in particular, and in those from wood 
of the poplar, the willow, and the sugar-maple, also, though to a 
less degree, the precipitate produced by phenylhydrazin is found 
to contain numerous small, flat, pale globules, like minute drops 
of oil, of yellowish or reddish-yellow color. But from_ these 
globules no crystalline product has as yet been obtained. These 
globules have a very different appearance from the globules of 
mannose-hydrazone. As seen under the microscope they seem 
to be thin, lean, and flat, while the mannose-hydrazone disks 
appear to be thick and plump, and to be built upon a generous 
scale. Though somewhat soluble in glacial acetic acid, the oil- 
like globules are not so readily dissolved by this acid as are the 
spiny balls of dextrose-osazone. The fact that in the case of the 
maple wood oil-like globules appear simultaneously with those of 
mannose-hydrazone in the precipitate produced by the phenyl- 
hydrazin reagents, while in the case of birch wood an abundance 
of the oil-like drops are present, though no mannose-hydrazone | 
could be detected, has made it imperative to try again and again 
to detect mannose in the products of hydrolysis from birch wood, 
though thus far without any success. Many samples of the birch 
wood have been hydrolyzed, both with hydrochloric acid and with 
sulphuric acid of varying degrees of concentration, but no man- 
nose has been detected. | 
In one special trial, made on purpose to determine whether it 
was possible to avoid these globules, 22 grm. of powdered birch 
wood, from a tree felled on October 10, were leached thoroughly 
with ether, and with alcohol of 95 per cent., and were subse- 
quently washed with water and partially dried before proceeding 
to boil the material for three hours in 220 cc. of hydrochloric acid 
of 5 percent. After neutralizing the product of the hydrolysis 
with sodium hydroxide, it was tested with the phenylhydrazin 
reagents. Yellow precipitates, soon becoming red colored, were 
