478 Intelligence and Miscellaneous Articles. 



Although in this equation 4 equivs. sulphate of soda are given, the 

 decomposition takes place with 2 and even with 1 equiv., but a larger 

 quantity of water is then necessary. In this case acid is set free ; 

 however, with a sufficiency of water this does not permanently hinder 

 the decomposition (in which the tendency of the hydrated oxide of 

 tin to separate certainly plays an important part), especially at an 

 elevated temperature. The influence of this latter condition was 

 indeed to be expected, since Fremy has shown that the solution of 

 hydrated oxide of tin in dilute sulphuric acid is decomposed by 

 boiling. 



It now remained to prove that by this means all the tin present 

 as perchloride was precipitated, and that the separation of the tin 

 and chlorine was complete. The decomposition was effected with 

 sulphate of soda (A) and nitrate of ammonia (B). 100 parts of a 

 solution of perchloride of tin gave the following results : — 



A. 

 Tin. Chlorine. 



1. 4-964 5-976 



B. 



Tin. Chlorine. 



1. 4-935 5-962 



2. 4-975 5-956 I 2. 4-948 5-971 



Consequently 100 parts anhydrous perchloride of tin would contain — 

 Calculated. Found. 



A. B. 



1. 2. 1. 2. 



Sneqt. wt 45-30 45-37 45-51 45-38 45-32 



2C1 eqt. wt 54-70 54-63 54-49 54-62 54-68 



100-00 100-00 100-00 100-00 100-00 

 These results show that the separation is effected with great 

 accuracy. 



This behaviour of oxide of tin admits of several useful applications ; 

 among others — 



1. The detection of tin in almost any liquid. 



2. It affords an exact and simple means of quantitatively determi- 

 ning tin, as -well as an easy and perfect means of separating it from 

 chlorine and other halogens, from alkalies, &c. 



It is of especial worth to manufacturers in determining the com- 

 mercial value of tin salts. In the absence of delicate balances they 

 generally have recourse to the nitric acid test, which always gives 

 an incorrect result, in consequence of the volatilization of the 

 chloride of tin. 



In some experiments on the quantitative separation of oxide from 

 binoxide of tin, I obtained on mixing 9-0723 of the above solution 

 to which a few drops of hydrochloric acid had been added with 

 10-6196 of a solution of protochloride corresponding to 16 per cent, 

 of protoxide, 0-6792 protoxide of tin corresponding to 0-5339 grm., or 

 calculated for 100 parts of the solution 5-002, consequently only 0-042 

 per cent, more than the average of the above results, which amounts 

 to 4-960. 



3. It aflbrds a very suitable means of combining oxide of tin with 

 cotton, woollen and silk stuffs in dyeing. Lciwenstein considers that 

 the experiments made by him justify the opinion that this method, 



