INTRODUCTION. 



This investigation is the direct outcome of an observation that was made 

 in the chemical laboratory of Johns Hopkins University five or six years ago. 

 It was found by Jones and Ota,* and by Jones and Knightf in connection 

 with their work on the condition of double sulphates, chlorides, nitrates 

 cyanides, etc., in aqueous solution, that concentrated solutions of these sub- 

 stances frequently show abnormally great depression of the freezing-point 

 of water. It was further observed that the molecular depression of the 

 freezing-point often increases with the concentration from a certain definite 

 concentration. 



Work 0} Chambers. The first systematic study of this problem was made 

 by Jones and Chambers. t They worked with calcium chloride, strontium 

 chloride, barium chloride, magnesium chloride, and cadmium chloride; and 

 wdth calcium bromide, strontium bromide, barium bromide, magnesium bro- 

 mide, and cadmium bromide. They measured the freezing-point lowerings 

 produced by these substances over a considerable range of concentration. 

 The general character of the results that they obtained is shown by the 

 data for a few substances given in table 1, on the following page. 



Column I gives the concentration of the solutions in terms of gram- 

 molecular normal; column II, the corrected lowering of the freezing-point 

 of water; and column III, the molecular lowering of the freezing-point of 

 water produced by the dissolved substance at the concentration in question. 



An examination of column III for all of the above substances shows that 

 the molecular lowering of the freezing-point passes through a minimum. 

 This was true for all the compounds investigated by Jones and Chambers, 

 except cadmium chloride and bromide. As the concentration increased 

 from the point at which the minimum in the molecular lowering occurred, 

 which was usually about 0.2 normal, the molecular lowering increased. 

 For the most concentrated solutions studied by Jones and Chambers, the 

 molecular lowering of the freezing-point of water became larger than the 

 theoretical lowering produced by a ternary electrolyte when it was com- 

 pletely dissociated; and it is well known that at these concentrations the 

 compounds under investigation were very far from being completely dis- 

 sociated. This was obviously a remarkable and important phenomenon, 

 meriting careful consideration. 



*Amer. Chem. Journ., 22, 5 (1890). 

 t/6id., 22, 110 (1899). 

 %lUd., 23,89 (1900). 



