142 BACTERIA AND THE OUM OF IIAKEA, 



removed by filtration through a hot, wetted, double filter, and 

 the osazones that formed upon the cooling of the solution were 

 filtered off, dried on porous porcelain and examined. As a rule 

 these osazones formed as voluminous yellow masses which con- 

 sisted of interlacing, indefinite, feathery, cr3^stalline tufts. Some- 

 times the mass appeared as a flabby yellow jelly, which shrank 

 to small volume on the filter. The solutions were treated twice 

 with phenylhydrazine solution, a third treatment being unneces- 

 sary. The individual precipitates were kept separate until the 

 melting points showed that the fractions were similar. 



The osazones of the first fraction of the first hydrolysis (0-5 % 

 acid) dried on porcelain as brownish skins like arabinosazone, but 

 they were largely soluble in ether, which enabled four fractions 

 to be obtained; these melted at 118°, 118-120°, 122°, and 132°. 

 They appeared to consist of the same osazone or mixture of osa- 

 zones; the ethereal solutions upon evaporation left reddish-yellow 

 crusts which, when dissolved in water and evaporated, gave a 

 yellow amorphous powder and a reddish-brown vitreous film. 

 Evaporation with dilute acetic acid converted everything into a 

 brownish tar. The second fraction of the first hydrolysis was 

 separated by ether into portions melting at 118°, 123-126°, 130°, 

 and 142° with the same characters as the portions of the first 

 fraction. 



The fractions of the second hydrolysis (1 % acid for I hour) 

 gave osazones which also dried on porcelain as brown skins, but 

 which melted at 158-159° and at 164-166°. These were added 

 together and treated with water on the water-bath. A part dis- 

 solved and separated out upon cooling as stellate clusters of 

 delicate needles mixed with small jagged spheres. On porcelain, 

 the precipitate formed a brown skin which melted at 114°. The 

 residue was again heated with water and filtered. A precipitate 

 of indefinite cr3^stals separated out upon cooling. These melted 

 between 162° and 166°. The residue, insoluble in the quantities 

 of hot water that were used, was suspended in a small volume of 

 water and carefully treated with alcohol while on the water-bath, 

 until the whole of it dissolved. The alcoholic solution was cooled 



