132 Dr. Thomson's Experiments to determine the [Aug. 



little from 2-75 that I was induced to conclude that 2-75 repre- 

 sents the true weight of an atom of boracic acid. The object 

 of the remaining experiments which I have to relate was to verify 

 this opinion. 



3. 100 grains of crystallized borax were exposed in a platinum 

 crucible to a heat sufficient to melt the salt. This heat was con- 

 tinued till the borax was reduced to a white mass ; the crucible 

 was then exposed to a strong red heat till the borax was melted 

 into a transparent colourless glass. The loss of weight which 

 the salt sustained was 47*2 grs. It appears from this experi- 

 ment that crystallized borax contains 47-2 grs. of water of 

 crystallization. 



Twenty grains of anhydrous glass of borax were dissolved in 

 distilled water by the assistance of heat, the solution was mixed 

 with an excess of muriatic acid ; and by repeated evaporations, 

 the whole, or almost the whole, of the boracic acid was extracted 

 from it. The weight of this acid (fused into a glass) was 5*1 

 grs. Now 100 grains of crystallized borax contain, as we have 

 seen, 52*8 grains of the anhydrous salt. And 20 : 52*8 :: 5*1 : 

 13*464. It follows from this that 100 grs. of crystallized borax 

 contain 13-464 grains of acid. Borax, therefore, must be com- 

 posed as follows : 



Acid 13-464 



Soda 39-336 



Water 47-200 



100-000 

 Now this is equivalent to 



Acid t 2-75 



Soda 8-034 



Water 9-640 



Two atoms of soda are 8, which differs very little from 8-034. 

 Eight atoms of water weigh 9. This is a little less than 9*64. 

 The reason probably is, that when borax is exposed to a red 

 heat, it is not the water alone which sublimes, but a portion of 

 the salt passes off along with it. Indeed I have verified this 

 suspicion by a direct experiment, and ascertained that the water 

 driven off' from borax by heat carries along with it a sensible 

 quantity of the salt. Had I obtained from the 20 grs. of anhy- 

 drous borax 5-11 grs. of boracic acid instead of 5-1 grs. the 

 quantity actually obtained, in that case the constituents of the 

 salt would have borne to each other the exact ratio of 2*75 acid 

 and 8 soda. Now when a substance is separated by a filter, 

 and afterwards exposed to a red heat, I do not consider myself 

 as capable of coming nearer the truth by the utmost attention 

 which I can pay than l-500th part of the whole. The deficiency 

 of 0-01 gr. being greatly within that limit, 1 consider myself as 

 entitled to infer, from the above analysis, that borax is a com- 



