30 THE SOLUBLE CONSTITUENTS. 



by alteration in rotation is intra -molecular. It is probably 

 caused by the migration of the water of hydration from one 

 carbon atom to another. 



The anhydrous modification of the a-modification is obtained 

 by heating the hydrated modification to 130 C. It is hydro- 

 scopic and dissolves in water with evolution of heat ; the solu- 

 bility is much greater than that of the hydrated modification. 

 The optical properties are stated by Schmoeger to be the same 

 as those of the hydrated modification. 



There exists also a /^-modification, which is obtained in the 

 anhydrous form by the rapid evaporation of aqueous solutions 

 in metallic vessels. It has a specific rotatory power of 32-7 

 at the moment of solution. Schmoeger states that it has a 



bi-rotation ratio of =-^- which Tanret confirms. The rate of 



' 



change of the a- and jS-modifications is the same for the same 



temperature. 



Both the a- and /^-modifications are converted on dissolving 

 in water into a stable equilibrium form. Schmoeger gives the 

 specific rotatory power [a] D as 52-53 at 20 C. 5 diminishing 

 O075 for each degree C. above and increasing for lower tem- 

 peratures. The author can confirm these numbers absolutely. 

 It has never been prepared pure in the solid state, though 

 considerable evidence of its existence, both in the hydrated 

 and anhydrous modifications, has been obtained by the author. 



By the addition of alcohol or, better, ether to a very highly 

 supersaturated hot solution of milk-sugar, it sets to a solid mass, 

 which may be dried in vacun, and does not then lose weight at 

 100 C., but which contains, however, a certain amount 2 to 4 

 per cent. of water of hydration which is lost at 130 C. There 

 is no appreciable change in rotation on dissolving this product 

 in water and taking readings at intervals ; as some of the read- 

 ings obtained have been above, and some below, that ultimately 

 obtained, it is probable that the very slight differences noticed 

 were due to errors of observation. 



By precipitating less strong solutions of milk-sugar by alcohol, 

 products can be obtained which contain very nearly, if not quite, 

 the same percentage of water as the hydrated a-modification, 

 but which have a much smaller, but not constant, bi-rotation 

 ratio. These also give a constant rotation for a few minutes on 

 dissolving in water and behave as mixtures of the a- and the s table - 

 modifications. They have a less density than the a-modification, 

 but it is not certain whether this may not be due to the very 

 imperfect crystallisation which takes place, the products appearing 

 nearly amorphous. 



By evaporating aqueous solutions of milk-sugar on the water 



