976 
In this connection it is worth while mentioning that stable super- 
saturated Solutions of al most all substances could be prepared by 
this simple method (e.g. snpersaturated Solutions of such sparingly 
soluble substances as potassium perchlorate, potassium chlorate, 
benzoic acid, salicylic acid etc. have been prepared). 
The tube containing the snpersaturated solution was cooled and 
carefully sealed off at the constricted portion and kept in an un- 
disturbed position. 
Usually after the lapse of two hours the solution was vigorously 
shaken for some minutes in order to observe the effect of mechanical 
shaking on crystallization, but in almost no cases did crystallization 
begin on shaking. 
The bulb containing the solution as well as the sealed-off portion 
of the tube were weighed fcuj and from these we can calculate 
the amount of supersaturation. 
Now the mouth of the bulb was broken and a small amount of 
the isomorphous substance was added and vigorously agita*ed for 
about 15 minutes. If no crystallization sets in, a small crystal of 
the dissolved solid was added and shaken; in this case, copious 
crystallization immediately begins. 
In some cases, instead of using a weighed amount of the substance 
each time, a concentrated solution of the substance is prepared and 
standardised. In each observation 5 c.c. of the standardised solution 
are used and thus the amount of the substance taken is known. 
In all cases Merck’s pure reagents were recrystallised and used 
in the present investigation. In order to avoid contamination of the 
isomorphous substance with germ crystals of the original substance, 
the isomorphous body was kept in a different room. 
Extreme care should be taken against the inoculation of the 
Solutions with germ crystals and in sealing off the tube care should 
be taken that no crystals are formed due to the evaporation of water. 
In all cases no crystallisation was produced with either calcium 
fluoride or strontium chloride (SrCl 2 , 6 H 2 0) though the amount of 
supersaturation was fairly high. 
Retgers (loc.cit) says, ,,CaF 2 and CaCl 2 are both cubic, the former 
being nearly insol uble, the latter extraordinarily deliquescent. Mixed 
crystals, therefore, cannot be formed from an aqueous solution, yet 
the two compounds are isomorphous. lt is quite possible that a 
common solvent rnight exist from which mixed crystals would be 
obtainable or the latter might be obtained from the fused mixture”. 
SrCl 2 , 6H 2 0 is well known to be strictly isomorphous with 
CaCl 2 , 6H 2 0 (cf. Roscoe’s Chemistry, Vol. II (1913) p. 596). Yet 
