THE VALENCY AND SPECIFIC HEAT OF THE METALS 599 



The nearest analogues of magnesium are able to give exactly similar 

 double salts, both in crystalline form (monoclinic system) and com- 



double cyanides are not separated by diffusion, which in all probability depends both on 

 the relative rate of the diffusion of the component salts and on the degree of affinity 

 acting between them. Those complex states of equilibrium which exist between water, 

 the individual salts MX and NY, and the double salt MNXY.have been already partially 

 analysed (as will be shown hereafter) in that case when the system is heterogeneous 

 (that is, when something separates out in a solid state from the liquid solution), bufr in 

 the case of equilibria in a homogeneous liquid medium (in a solution) the phenomenon id 

 not so clear, because it concerns that very theory of solution which cannot yet be 

 considered as established (Chapter I., Note 9, and others). As regards the heterogeneous 

 decomposition -of double salts, it has long been known that such, salts as carnallite and 

 K 2 Mg(S0 4 ) 2 give up the more soluble salt if an insufficient quantity of water for their 

 complete solution be taken. The complete saturation of 100 parts of water requires at 

 <J 14'1, at 20 25, and at 60 50'2 parts of the latter double salt (anhydrous), while 100 

 parts of water dissolve 27 parts of magnesium sulphate at 0, 86 parts at 20, and 55 

 parts at 60, of the anhydrous salt taken. Of all the states of equilibrium exhibited by 

 double salts the most fully investigated as yet is the system containing water, sodium 

 sulphate, magnesium sulphate, and their double salt, Na2Mg(SO 4 ) 2 , which crystallises 

 with 4 and 6 mol. OH 2 . The first crystallo-hydrate, MgNa 2 (SO 4 ) 2 ,4H 2 O, occurs at 

 Stassfurt, and as a sedimentary deposit in many of the salt lakes near Astrakhan,- and is. 

 therefore called astrakhanite. The specific gravity pf the monoclinic prisms, of this salt 

 is 2-22. If this salt, in a finely divided state, bo mixed with the necessary quantity ot wfcter 

 (according to the equation MgN^(SO 4 ) 2) 4H 2 O + 13H 2 O = Na^SO4 t 10H 2 O + MgSO 4 ,7H2O), 

 the mixture solidifies like plaster of Paris into a homogeneous mass if the temperature 

 be below 22 (Van't Hoff und Van Deventer, 1886 ; Bakhuis Boozeboom, 1887) ; but 

 if the temperature be above this transition-point the water and double salt do 

 not react on each other; that is, they do not solidify or give a mixture of sodium 

 and magnesium sulphates. If a mixture (in equivalent quantities) of solutions of these 

 salts be evaporated, and crystals of astrakhanite and of the individual salts capable of 

 proceeding from it be added to the concentrated solution to avoid the possibility of a 

 supersaturated solution, then at temperatures above 22 astrakhanite is exclusively 

 formed (this is the method of its production), but at lower temperatures the individual 

 salts are alone produced. If equivalent amounts of Glauber's salt and magnesium 

 sulphate be mixed together in a solid state, there is no change at temperatures below 

 22, but at higher temperatures astrakhanite .and water are formed. The volume 

 occupied by Na2SO 4 ,10H 2 O in grams =322/1 -46 = 220'5 cubic centimetres, and by 

 .MgSO 4 ,7H 2 = 246/l'68 = 146'4 c.c.; hence their mixture in equivalent quantities occupies a 

 volume of 366'9 c.c. The volume of astrakhanite = 834/2'22 = 150'5 c.c., and the volume of 

 18H 2 O = 234 c.c., hence their sum = S80'5 c.c., and therefore it is easy to follow the formation 

 of the astrakhanite in a suitable apparatus (a kind of thermometer containing oil and a 

 powdered mixture of sodium and magnesium sulphates), and to see by the variation in 

 volume that below 22 it remains unchanged, and at higher temperatures proceeds the 

 more quickly the higher the temperature. At the transition temperature the solubility 

 of astrakhanite and of the mixture of the component salts is one and the same, whilst at 

 higher temperatures a solution wThich is saturated for a mixture* of the individual salt* 

 would be supersaturated for astrakhanite, and at lower temperatures the solution of 

 astrakhanite will be supersaturated for the component salts, as has been -shown with 

 especial detail by Karsten, Deacon, and others. Roozeboom showed thafc there are two 

 limits to the composition of the solutions which can exist for a double salt ; these limits 

 are respectively obtained by dissolving a mixture of the double salt with each of its 

 component simple salts. Vau't Hoff demonstrated, besides this, that the tendency 

 towards the formation of double salts has a distinct influence on the progress of double 

 decomposition, for at temperatures above 31 the mixture 2MgSO 4 J7H 2 O + 2NaCl 

 into MgNa2(SO 4 ) 2 ,4H 2 O + MgCl 2 ,CH 2 O + 4HO, whilst below 31 there is not this 



