and Attached Water. 109 



As the flask had to be left for half an hour in the balance- 

 case every time before weighing, the above experiments occu- 

 pied a few weeks. At the end of this time the flask was found 

 to have lost 0*001 7 gram. 



Examining the final result as a direct determination of the 

 whole of the water, I found that the 4*1605 grams had lost 

 0*6087 gram, or 14*63 per cent., instead of the theoretical 

 amount 14*75. 



The actual weight of water which a given w T eight of the salt 

 lost at a given temperature in a given time has little interest, 

 because it is conditioned by the attitude of the salt to the air- 

 current in the flask, and it is also governed by the rate. But 

 points of very great interest are nevertheless presented when the 

 above numbers are compared. The first loss, at 17° C, may pro- 

 bably be attributable to the more complete drying of the salt, 

 since no further loss was experienced at 25°. Starting at 40°, a 

 considerable loss was experienced, which continued with very 

 considerable regularity until the residue approached in com- 

 position to the one-atom hydrate BaCl 2 + H 2 O ; the loss then 

 suddenly diminished and abruptly stopped. The total loss ex- 

 perienced by the 4*1605 grams of salt when this point was 

 reached is 0*3009 — that is, 7*21 per cent. As the percentage 

 of water in BaCl 2 + H 2 is 7*37, there can be no question 

 that there is a difference in the strength of union of the two 

 water molecules to the salt, or, more exactly, that it requires 

 different physical conditions to separate « H 2 O from BaCl 2 

 than are sufficient to separate /3 H 2 O from « H 2 O, BaCl 2 . 



At what minimum temperature the /S molecule begins to be 

 separated is missed in this Table : it lies somewhere between 

 25° and 40°. But it appears that when the temperature is 

 such that one molecule begins to be stirred, the whole of that 

 molecule is removed if the current of dry air be continued. 

 In the case of the chloride of barium, there is a range of tem- 

 perature below 60° and reaching down to the above-mentioned 

 minimum, in which the « molecule is fixed while the /8 molecule 

 is removable. 



The minimum temperature required to disconnect the « mo- 

 lecule is well marked. The salt having ceased to lose weight 

 at 40° C, lost weight distinctly at 60°, and still more rapidly 

 at 70°. On reducing the temperature to 50° no loss could be 

 detected ; at 60° the same as before, and at 70° the same as 

 before. At 55° there was no loss, nor at 58°, nor at 59°; but 

 at 60° the original loss was reestablished. And until the salt 

 is becoming anhydrous, for each temperature there is a pretty 

 constant loss. 



The anhydrous BaCl 2 thus obtained, when mixed with water, 



