302 Mr Clark on the Pyrophosphate of Soda. 



cury ; and yet phosphate of soda dried on this bath always 

 gave with nitrate of silver a yellow precipitate. I proposed, 

 therefore, to ascertain first, the weight of water expelled by 

 the sand-bath heat, and then, the alteration in weight produ- 

 ced by a red heat. 



As likely to afford the most uniform results, I made choice 

 of well-formed, unbroken, pure crystals of phosphate of soda. 

 This salt after being dried in a strong sand-bath heat, uni- 

 formly underwent a loss in weight by the farther action of a 

 red heat. The following is the result of three experiments, 

 the quantity of crystals being supposed 1. 



Experiment 



I. II. III.* ]^ean. 



Water, expelled by sand-bath .6168 .6170 .6164 .6167 



Loss, by red heat .0250 .0247 .0247 .0248 



Dry pyrophosphate of soda .3582 .3593 .3589 .3585 



In these experiments, the salt after exposure to the sand- 

 bath heat gave with nitrate of silver a yellow precipitate ; but 

 after exposure to the red heat it gave a white precipitate. 

 These experiments, then, establish that the phosphate of soda, 

 in becoming pyrophosphate, loses weight. What is the mat- 

 ter lost ? 



To determine this important point, I dried a quantity of 

 phosphate of soda, giving it the highest heat of the sand-bath. 

 The phosphate thus dried still tested yellow, with nitrate of 

 silver. I put 23.45 grains of this dried salt into a small re- 

 tort of green glass. This retort was formed of a tube about 

 a finger-length, about the thickness of a quill, and having a 

 bulb at one end about the size of a hazel nut, and a 

 bend a little above this bulb, the other end of the tube ^^ 

 being open. To this small retort, there was adapted a 

 tube of this shape, narrow enough to go within the 

 tube of the retort. The junction was made air-tight 11 

 by oil cement. A small jar filled with mercury was | | 

 inverted in a mercurial trough, and .1 inch of air 



• The experiment marked No. III. is the mean of two experiments, the 

 greatest and the least of the set. The water expelled by the sand-bath 

 was in one, where efflorescence was begun on the surface, .6123, and in the 

 other, whose surface was yet moist, .6204. 



