310 Mr Clark on the Arseniate of Soda. 



The oxygen in the water expelled by a red heat is 1.03, 

 1.05, 1.03. This is as nearly 1. as can be expected in such 

 experiments. The oxygen in the water expelled by the sand- 

 bath is as follows : 



From crystals, - - - - 15.47 



From crystals pounded and wiped in a handkerchief, 14.56 

 From pounded crystals, dried in blotting-paper, 13.98 



In the last experiment, there is little chance of deficiency of 

 water, because the salt endures a long exposure to the air 

 without experiencing Ipss of water. I conclude, therefore, 

 that, exclusive of mechanical water, this arseniate loses 14 pro- 

 portions by a sand-bath heat, and 1 proportion by a red heat. 



Dr Gmelin gives two experiments ; in one of which the to- 

 tal water turns out 15.17 proportions and in the other 15.65. 

 He conceives that the salt contains 16 proportions ; but, from 

 his experiments and my own, I think it much more probable 

 that 15 is the true quantity. 



There seems, indeed, to be among chemists, a most unwar- 

 rantable hai)it of carelessness in deducing the water of crystal- 

 lization. The following is rather a droll example, taken from 

 Professor Mitscherlich''s valuMe Essai/ 07i the Phosphates and 

 Arseniates, He has under examination the other arseniate of 

 soda. 



" 6.789 grammes of arseniate of soda being ignited gave 

 3. of residue. Consequently 100 parts of arseniate of soda 

 combine with 126.3 parts of water. 100 parts of arseniate 

 of soda contain 35.18 parts of soda, according to the atomic 

 weights of Berzelius ; and these contain 9.00 parts of oxygen. 

 126.3 parts of water contain 112.3 parts of oxygen. 



"But 9.00: 112.3: : 1 : 12.5 



" Theeefore, the oxygen in the soda is to that of the wa- 

 ter as 1 is to 12. Consequently the crystallized salt consists 

 of, Sfcr 



This may do in Chemistry, where we are apt to pique our- 

 selves upon accuracy ; but how would it do in Astronomy to 

 read " By observation, it was half past 12. Therefore^ it 

 was precisely 12. Consequently^ &c. .?" 



I select this instance from Professor Mitscherlich's essay, 



