186 H. G. SMITH. 
perfectly white, and crystallised in globular masses of radiating 
crystals, always taking that form when crystallised from water. 
Raffinose crystallised from water. Obtained from the exudation of 
E. punctata. Natural size. 
The determination of the water of crystallisation was found 
to be of some importance as by heating at different temperatures 
the results did not always agree. When heated to 95° C. until 
constant, 14:53 per cent. of water was driven off, (mean of two 
determinations), but when heated to 100° C. (the loss remained 
constant at near 15:1 per cent., four determinations on different 
material giving 15°13, 15-093, 15-091, 15:11); this is very near the 
theoretical amount required by the recognised formula for raffinose 
viz., C,,H;,0,, + 5 H,O which requires 15:15 per cent. of water. 
On raising the temperature to the melting point (118° ©.) the 
weight still remained the same. By this treatment the sugar wa 
