ACTION OP WATER ON IRON. 301 



briskly to dryness, avoiding spattering, redissolved in water, 

 and precipitate of ZnO. separated, ignited, and weighed warm. 



6th. The sohition tested for remains of zinc by bihydro-sul- 

 phuret ammonia. 



118. 24*80 grains of the brass received from Liverpool through 

 Mr. Gilbert Cummins, analyzed in this way, gave the following 

 results, reduced to per cent. : 



Tin = 12-012 

 Lead = 0*266 

 Copper = 79*750 

 Iron = 3*137 

 Zinc = 4*786 

 Loss = 0*049 



100-000 



119. 42*085 grains of the brass from the Dublin Docks gave 

 the following composition, also reduced to per cent. : 



Tin = 0-807 



Lead = 4*062 



Iron = 0-879 



Copper = 65-890 = 2 atoms Cu 



Ziiic = 28-288 = 1 atom Z'n 



Loss = 00-074 



100-000 



120. I analyzed 41-705 grains of tie specimen of Mr. Hart- 

 ley's brass given us by Dr. Kane, with the following results, 

 Miiich present a larger amount of loss than I could have wished, 

 arising from my having been several times delayed in comple- 

 ting the process by unavoidable business. — It gave, reduced to 

 per cent.. 



Tin = 4-524 



Lead = 13 051 



Iron =1-743 



Zinc = 8-639 



Copper = 67-233 



Loss = 4-810 



100-000 



It is hence obvious that all these brasses are chance mixtures, 

 and that the Dublin Dock brass contains the most zinc, and 

 comes nearest to atomic constitution 5 and hence might have 

 been expected, on Mr. Hartley's hypothesis, to have most ef- 



